From 76bc9968ff0ec7d044042f9214fa59efb3146685 Mon Sep 17 00:00:00 2001 From: Adam Ierymenko Date: Wed, 17 Jul 2013 14:10:44 -0400 Subject: New simpler command interface via loopback UDP protected by a crypto token. --- ext/jsoncpp/include/json/autolink.h | 24 - ext/jsoncpp/include/json/config.h | 96 -- ext/jsoncpp/include/json/features.h | 49 - ext/jsoncpp/include/json/forwards.h | 44 - ext/jsoncpp/include/json/json.h | 15 - ext/jsoncpp/include/json/reader.h | 214 ---- ext/jsoncpp/include/json/value.h | 1103 ------------------- ext/jsoncpp/include/json/writer.h | 185 ---- ext/jsoncpp/src/json_batchallocator.h | 130 --- ext/jsoncpp/src/json_internalarray.inl | 456 -------- ext/jsoncpp/src/json_internalmap.inl | 615 ----------- ext/jsoncpp/src/json_reader.cpp | 880 --------------- ext/jsoncpp/src/json_tool.h | 93 -- ext/jsoncpp/src/json_value.cpp | 1829 -------------------------------- ext/jsoncpp/src/json_valueiterator.inl | 299 ------ ext/jsoncpp/src/json_writer.cpp | 838 --------------- 16 files changed, 6870 deletions(-) delete mode 100644 ext/jsoncpp/include/json/autolink.h delete mode 100644 ext/jsoncpp/include/json/config.h delete mode 100644 ext/jsoncpp/include/json/features.h delete mode 100644 ext/jsoncpp/include/json/forwards.h delete mode 100644 ext/jsoncpp/include/json/json.h delete mode 100644 ext/jsoncpp/include/json/reader.h delete mode 100644 ext/jsoncpp/include/json/value.h delete mode 100644 ext/jsoncpp/include/json/writer.h delete mode 100644 ext/jsoncpp/src/json_batchallocator.h delete mode 100644 ext/jsoncpp/src/json_internalarray.inl delete mode 100644 ext/jsoncpp/src/json_internalmap.inl delete mode 100644 ext/jsoncpp/src/json_reader.cpp delete mode 100644 ext/jsoncpp/src/json_tool.h delete mode 100644 ext/jsoncpp/src/json_value.cpp delete mode 100644 ext/jsoncpp/src/json_valueiterator.inl delete mode 100644 ext/jsoncpp/src/json_writer.cpp (limited to 'ext') diff --git a/ext/jsoncpp/include/json/autolink.h b/ext/jsoncpp/include/json/autolink.h deleted file mode 100644 index 02328d1f..00000000 --- a/ext/jsoncpp/include/json/autolink.h +++ /dev/null @@ -1,24 +0,0 @@ -// Copyright 2007-2010 Baptiste Lepilleur -// Distributed under MIT license, or public domain if desired and -// recognized in your jurisdiction. -// See file LICENSE for detail or copy at http://jsoncpp.sourceforge.net/LICENSE - -#ifndef JSON_AUTOLINK_H_INCLUDED -# define JSON_AUTOLINK_H_INCLUDED - -# include "config.h" - -# ifdef JSON_IN_CPPTL -# include -# endif - -# if !defined(JSON_NO_AUTOLINK) && !defined(JSON_DLL_BUILD) && !defined(JSON_IN_CPPTL) -# define CPPTL_AUTOLINK_NAME "json" -# undef CPPTL_AUTOLINK_DLL -# ifdef JSON_DLL -# define CPPTL_AUTOLINK_DLL -# endif -# include "autolink.h" -# endif - -#endif // JSON_AUTOLINK_H_INCLUDED diff --git a/ext/jsoncpp/include/json/config.h b/ext/jsoncpp/include/json/config.h deleted file mode 100644 index 7609d45e..00000000 --- a/ext/jsoncpp/include/json/config.h +++ /dev/null @@ -1,96 +0,0 @@ -// Copyright 2007-2010 Baptiste Lepilleur -// Distributed under MIT license, or public domain if desired and -// recognized in your jurisdiction. -// See file LICENSE for detail or copy at http://jsoncpp.sourceforge.net/LICENSE - -#ifndef JSON_CONFIG_H_INCLUDED -# define JSON_CONFIG_H_INCLUDED - -/// If defined, indicates that json library is embedded in CppTL library. -//# define JSON_IN_CPPTL 1 - -/// If defined, indicates that json may leverage CppTL library -//# define JSON_USE_CPPTL 1 -/// If defined, indicates that cpptl vector based map should be used instead of std::map -/// as Value container. -//# define JSON_USE_CPPTL_SMALLMAP 1 -/// If defined, indicates that Json specific container should be used -/// (hash table & simple deque container with customizable allocator). -/// THIS FEATURE IS STILL EXPERIMENTAL! There is know bugs: See #3177332 -//# define JSON_VALUE_USE_INTERNAL_MAP 1 -/// Force usage of standard new/malloc based allocator instead of memory pool based allocator. -/// The memory pools allocator used optimization (initializing Value and ValueInternalLink -/// as if it was a POD) that may cause some validation tool to report errors. -/// Only has effects if JSON_VALUE_USE_INTERNAL_MAP is defined. -//# define JSON_USE_SIMPLE_INTERNAL_ALLOCATOR 1 - -/// If defined, indicates that Json use exception to report invalid type manipulation -/// instead of C assert macro. -# define JSON_USE_EXCEPTION 1 - -/// If defined, indicates that the source file is amalgated -/// to prevent private header inclusion. -/// Remarks: it is automatically defined in the generated amalgated header. -// #define JSON_IS_AMALGAMATION - - -# ifdef JSON_IN_CPPTL -# include -# ifndef JSON_USE_CPPTL -# define JSON_USE_CPPTL 1 -# endif -# endif - -# ifdef JSON_IN_CPPTL -# define JSON_API CPPTL_API -# elif defined(JSON_DLL_BUILD) -# define JSON_API __declspec(dllexport) -# elif defined(JSON_DLL) -# define JSON_API __declspec(dllimport) -# else -# define JSON_API -# endif - -// If JSON_NO_INT64 is defined, then Json only support C++ "int" type for integer -// Storages, and 64 bits integer support is disabled. -// #define JSON_NO_INT64 1 - -#if defined(_MSC_VER) && _MSC_VER <= 1200 // MSVC 6 -// Microsoft Visual Studio 6 only support conversion from __int64 to double -// (no conversion from unsigned __int64). -#define JSON_USE_INT64_DOUBLE_CONVERSION 1 -#endif // if defined(_MSC_VER) && _MSC_VER < 1200 // MSVC 6 - -#if defined(_MSC_VER) && _MSC_VER >= 1500 // MSVC 2008 -/// Indicates that the following function is deprecated. -# define JSONCPP_DEPRECATED(message) __declspec(deprecated(message)) -#endif - -#if !defined(JSONCPP_DEPRECATED) -# define JSONCPP_DEPRECATED(message) -#endif // if !defined(JSONCPP_DEPRECATED) - -namespace Json { - typedef int Int; - typedef unsigned int UInt; -# if defined(JSON_NO_INT64) - typedef int LargestInt; - typedef unsigned int LargestUInt; -# undef JSON_HAS_INT64 -# else // if defined(JSON_NO_INT64) - // For Microsoft Visual use specific types as long long is not supported -# if defined(_MSC_VER) // Microsoft Visual Studio - typedef __int64 Int64; - typedef unsigned __int64 UInt64; -# else // if defined(_MSC_VER) // Other platforms, use long long - typedef long long int Int64; - typedef unsigned long long int UInt64; -# endif // if defined(_MSC_VER) - typedef Int64 LargestInt; - typedef UInt64 LargestUInt; -# define JSON_HAS_INT64 -# endif // if defined(JSON_NO_INT64) -} // end namespace Json - - -#endif // JSON_CONFIG_H_INCLUDED diff --git a/ext/jsoncpp/include/json/features.h b/ext/jsoncpp/include/json/features.h deleted file mode 100644 index 43532784..00000000 --- a/ext/jsoncpp/include/json/features.h +++ /dev/null @@ -1,49 +0,0 @@ -// Copyright 2007-2010 Baptiste Lepilleur -// Distributed under MIT license, or public domain if desired and -// recognized in your jurisdiction. -// See file LICENSE for detail or copy at http://jsoncpp.sourceforge.net/LICENSE - -#ifndef CPPTL_JSON_FEATURES_H_INCLUDED -# define CPPTL_JSON_FEATURES_H_INCLUDED - -#if !defined(JSON_IS_AMALGAMATION) -# include "forwards.h" -#endif // if !defined(JSON_IS_AMALGAMATION) - -namespace Json { - - /** \brief Configuration passed to reader and writer. - * This configuration object can be used to force the Reader or Writer - * to behave in a standard conforming way. - */ - class JSON_API Features - { - public: - /** \brief A configuration that allows all features and assumes all strings are UTF-8. - * - C & C++ comments are allowed - * - Root object can be any JSON value - * - Assumes Value strings are encoded in UTF-8 - */ - static Features all(); - - /** \brief A configuration that is strictly compatible with the JSON specification. - * - Comments are forbidden. - * - Root object must be either an array or an object value. - * - Assumes Value strings are encoded in UTF-8 - */ - static Features strictMode(); - - /** \brief Initialize the configuration like JsonConfig::allFeatures; - */ - Features(); - - /// \c true if comments are allowed. Default: \c true. - bool allowComments_; - - /// \c true if root must be either an array or an object value. Default: \c false. - bool strictRoot_; - }; - -} // namespace Json - -#endif // CPPTL_JSON_FEATURES_H_INCLUDED diff --git a/ext/jsoncpp/include/json/forwards.h b/ext/jsoncpp/include/json/forwards.h deleted file mode 100644 index ab863da8..00000000 --- a/ext/jsoncpp/include/json/forwards.h +++ /dev/null @@ -1,44 +0,0 @@ -// Copyright 2007-2010 Baptiste Lepilleur -// Distributed under MIT license, or public domain if desired and -// recognized in your jurisdiction. -// See file LICENSE for detail or copy at http://jsoncpp.sourceforge.net/LICENSE - -#ifndef JSON_FORWARDS_H_INCLUDED -# define JSON_FORWARDS_H_INCLUDED - -#if !defined(JSON_IS_AMALGAMATION) -# include "config.h" -#endif // if !defined(JSON_IS_AMALGAMATION) - -namespace Json { - - // writer.h - class FastWriter; - class StyledWriter; - - // reader.h - class Reader; - - // features.h - class Features; - - // value.h - typedef unsigned int ArrayIndex; - class StaticString; - class Path; - class PathArgument; - class Value; - class ValueIteratorBase; - class ValueIterator; - class ValueConstIterator; -#ifdef JSON_VALUE_USE_INTERNAL_MAP - class ValueMapAllocator; - class ValueInternalLink; - class ValueInternalArray; - class ValueInternalMap; -#endif // #ifdef JSON_VALUE_USE_INTERNAL_MAP - -} // namespace Json - - -#endif // JSON_FORWARDS_H_INCLUDED diff --git a/ext/jsoncpp/include/json/json.h b/ext/jsoncpp/include/json/json.h deleted file mode 100644 index da5fc967..00000000 --- a/ext/jsoncpp/include/json/json.h +++ /dev/null @@ -1,15 +0,0 @@ -// Copyright 2007-2010 Baptiste Lepilleur -// Distributed under MIT license, or public domain if desired and -// recognized in your jurisdiction. -// See file LICENSE for detail or copy at http://jsoncpp.sourceforge.net/LICENSE - -#ifndef JSON_JSON_H_INCLUDED -# define JSON_JSON_H_INCLUDED - -# include "autolink.h" -# include "value.h" -# include "reader.h" -# include "writer.h" -# include "features.h" - -#endif // JSON_JSON_H_INCLUDED diff --git a/ext/jsoncpp/include/json/reader.h b/ext/jsoncpp/include/json/reader.h deleted file mode 100644 index 0a324dfc..00000000 --- a/ext/jsoncpp/include/json/reader.h +++ /dev/null @@ -1,214 +0,0 @@ -// Copyright 2007-2010 Baptiste Lepilleur -// Distributed under MIT license, or public domain if desired and -// recognized in your jurisdiction. -// See file LICENSE for detail or copy at http://jsoncpp.sourceforge.net/LICENSE - -#ifndef CPPTL_JSON_READER_H_INCLUDED -# define CPPTL_JSON_READER_H_INCLUDED - -#if !defined(JSON_IS_AMALGAMATION) -# include "features.h" -# include "value.h" -#endif // if !defined(JSON_IS_AMALGAMATION) -# include -# include -# include -# include - -namespace Json { - - /** \brief Unserialize a JSON document into a Value. - * - */ - class JSON_API Reader - { - public: - typedef char Char; - typedef const Char *Location; - - /** \brief Constructs a Reader allowing all features - * for parsing. - */ - Reader(); - - /** \brief Constructs a Reader allowing the specified feature set - * for parsing. - */ - Reader( const Features &features ); - - /** \brief Read a Value from a JSON document. - * \param document UTF-8 encoded string containing the document to read. - * \param root [out] Contains the root value of the document if it was - * successfully parsed. - * \param collectComments \c true to collect comment and allow writing them back during - * serialization, \c false to discard comments. - * This parameter is ignored if Features::allowComments_ - * is \c false. - * \return \c true if the document was successfully parsed, \c false if an error occurred. - */ - bool parse( const std::string &document, - Value &root, - bool collectComments = true ); - - /** \brief Read a Value from a JSON document. - * \param beginDoc Pointer on the beginning of the UTF-8 encoded string of the document to read. - * \param endDoc Pointer on the end of the UTF-8 encoded string of the document to read. - \ Must be >= beginDoc. - * \param root [out] Contains the root value of the document if it was - * successfully parsed. - * \param collectComments \c true to collect comment and allow writing them back during - * serialization, \c false to discard comments. - * This parameter is ignored if Features::allowComments_ - * is \c false. - * \return \c true if the document was successfully parsed, \c false if an error occurred. - */ - bool parse( const char *beginDoc, const char *endDoc, - Value &root, - bool collectComments = true ); - - /// \brief Parse from input stream. - /// \see Json::operator>>(std::istream&, Json::Value&). - bool parse( std::istream &is, - Value &root, - bool collectComments = true ); - - /** \brief Returns a user friendly string that list errors in the parsed document. - * \return Formatted error message with the list of errors with their location in - * the parsed document. An empty string is returned if no error occurred - * during parsing. - * \deprecated Use getFormattedErrorMessages() instead (typo fix). - */ - JSONCPP_DEPRECATED("Use getFormattedErrorMessages instead") - std::string getFormatedErrorMessages() const; - - /** \brief Returns a user friendly string that list errors in the parsed document. - * \return Formatted error message with the list of errors with their location in - * the parsed document. An empty string is returned if no error occurred - * during parsing. - */ - std::string getFormattedErrorMessages() const; - - private: - enum TokenType - { - tokenEndOfStream = 0, - tokenObjectBegin, - tokenObjectEnd, - tokenArrayBegin, - tokenArrayEnd, - tokenString, - tokenNumber, - tokenTrue, - tokenFalse, - tokenNull, - tokenArraySeparator, - tokenMemberSeparator, - tokenComment, - tokenError - }; - - class Token - { - public: - TokenType type_; - Location start_; - Location end_; - }; - - class ErrorInfo - { - public: - Token token_; - std::string message_; - Location extra_; - }; - - typedef std::deque Errors; - - bool expectToken( TokenType type, Token &token, const char *message ); - bool readToken( Token &token ); - void skipSpaces(); - bool match( Location pattern, - int patternLength ); - bool readComment(); - bool readCStyleComment(); - bool readCppStyleComment(); - bool readString(); - void readNumber(); - bool readValue(); - bool readObject( Token &token ); - bool readArray( Token &token ); - bool decodeNumber( Token &token ); - bool decodeString( Token &token ); - bool decodeString( Token &token, std::string &decoded ); - bool decodeDouble( Token &token ); - bool decodeUnicodeCodePoint( Token &token, - Location ¤t, - Location end, - unsigned int &unicode ); - bool decodeUnicodeEscapeSequence( Token &token, - Location ¤t, - Location end, - unsigned int &unicode ); - bool addError( const std::string &message, - Token &token, - Location extra = 0 ); - bool recoverFromError( TokenType skipUntilToken ); - bool addErrorAndRecover( const std::string &message, - Token &token, - TokenType skipUntilToken ); - void skipUntilSpace(); - Value ¤tValue(); - Char getNextChar(); - void getLocationLineAndColumn( Location location, - int &line, - int &column ) const; - std::string getLocationLineAndColumn( Location location ) const; - void addComment( Location begin, - Location end, - CommentPlacement placement ); - void skipCommentTokens( Token &token ); - - typedef std::stack Nodes; - Nodes nodes_; - Errors errors_; - std::string document_; - Location begin_; - Location end_; - Location current_; - Location lastValueEnd_; - Value *lastValue_; - std::string commentsBefore_; - Features features_; - bool collectComments_; - }; - - /** \brief Read from 'sin' into 'root'. - - Always keep comments from the input JSON. - - This can be used to read a file into a particular sub-object. - For example: - \code - Json::Value root; - cin >> root["dir"]["file"]; - cout << root; - \endcode - Result: - \verbatim - { - "dir": { - "file": { - // The input stream JSON would be nested here. - } - } - } - \endverbatim - \throw std::exception on parse error. - \see Json::operator<<() - */ - std::istream& operator>>( std::istream&, Value& ); - -} // namespace Json - -#endif // CPPTL_JSON_READER_H_INCLUDED diff --git a/ext/jsoncpp/include/json/value.h b/ext/jsoncpp/include/json/value.h deleted file mode 100644 index 32e3455e..00000000 --- a/ext/jsoncpp/include/json/value.h +++ /dev/null @@ -1,1103 +0,0 @@ -// Copyright 2007-2010 Baptiste Lepilleur -// Distributed under MIT license, or public domain if desired and -// recognized in your jurisdiction. -// See file LICENSE for detail or copy at http://jsoncpp.sourceforge.net/LICENSE - -#ifndef CPPTL_JSON_H_INCLUDED -# define CPPTL_JSON_H_INCLUDED - -#if !defined(JSON_IS_AMALGAMATION) -# include "forwards.h" -#endif // if !defined(JSON_IS_AMALGAMATION) -# include -# include - -# ifndef JSON_USE_CPPTL_SMALLMAP -# include -# else -# include -# endif -# ifdef JSON_USE_CPPTL -# include -# endif - -/** \brief JSON (JavaScript Object Notation). - */ -namespace Json { - - /** \brief Type of the value held by a Value object. - */ - enum ValueType - { - nullValue = 0, ///< 'null' value - intValue, ///< signed integer value - uintValue, ///< unsigned integer value - realValue, ///< double value - stringValue, ///< UTF-8 string value - booleanValue, ///< bool value - arrayValue, ///< array value (ordered list) - objectValue ///< object value (collection of name/value pairs). - }; - - enum CommentPlacement - { - commentBefore = 0, ///< a comment placed on the line before a value - commentAfterOnSameLine, ///< a comment just after a value on the same line - commentAfter, ///< a comment on the line after a value (only make sense for root value) - numberOfCommentPlacement - }; - -//# ifdef JSON_USE_CPPTL -// typedef CppTL::AnyEnumerator EnumMemberNames; -// typedef CppTL::AnyEnumerator EnumValues; -//# endif - - /** \brief Lightweight wrapper to tag static string. - * - * Value constructor and objectValue member assignement takes advantage of the - * StaticString and avoid the cost of string duplication when storing the - * string or the member name. - * - * Example of usage: - * \code - * Json::Value aValue( StaticString("some text") ); - * Json::Value object; - * static const StaticString code("code"); - * object[code] = 1234; - * \endcode - */ - class JSON_API StaticString - { - public: - explicit StaticString( const char *czstring ) - : str_( czstring ) - { - } - - operator const char *() const - { - return str_; - } - - const char *c_str() const - { - return str_; - } - - private: - const char *str_; - }; - - /** \brief Represents a JSON value. - * - * This class is a discriminated union wrapper that can represents a: - * - signed integer [range: Value::minInt - Value::maxInt] - * - unsigned integer (range: 0 - Value::maxUInt) - * - double - * - UTF-8 string - * - boolean - * - 'null' - * - an ordered list of Value - * - collection of name/value pairs (javascript object) - * - * The type of the held value is represented by a #ValueType and - * can be obtained using type(). - * - * values of an #objectValue or #arrayValue can be accessed using operator[]() methods. - * Non const methods will automatically create the a #nullValue element - * if it does not exist. - * The sequence of an #arrayValue will be automatically resize and initialized - * with #nullValue. resize() can be used to enlarge or truncate an #arrayValue. - * - * The get() methods can be used to obtanis default value in the case the required element - * does not exist. - * - * It is possible to iterate over the list of a #objectValue values using - * the getMemberNames() method. - */ - class JSON_API Value - { - friend class ValueIteratorBase; -# ifdef JSON_VALUE_USE_INTERNAL_MAP - friend class ValueInternalLink; - friend class ValueInternalMap; -# endif - public: - typedef std::vector Members; - typedef ValueIterator iterator; - typedef ValueConstIterator const_iterator; - typedef Json::UInt UInt; - typedef Json::Int Int; -# if defined(JSON_HAS_INT64) - typedef Json::UInt64 UInt64; - typedef Json::Int64 Int64; -#endif // defined(JSON_HAS_INT64) - typedef Json::LargestInt LargestInt; - typedef Json::LargestUInt LargestUInt; - typedef Json::ArrayIndex ArrayIndex; - - static const Value null; - /// Minimum signed integer value that can be stored in a Json::Value. - static const LargestInt minLargestInt; - /// Maximum signed integer value that can be stored in a Json::Value. - static const LargestInt maxLargestInt; - /// Maximum unsigned integer value that can be stored in a Json::Value. - static const LargestUInt maxLargestUInt; - - /// Minimum signed int value that can be stored in a Json::Value. - static const Int minInt; - /// Maximum signed int value that can be stored in a Json::Value. - static const Int maxInt; - /// Maximum unsigned int value that can be stored in a Json::Value. - static const UInt maxUInt; - - /// Minimum signed 64 bits int value that can be stored in a Json::Value. - static const Int64 minInt64; - /// Maximum signed 64 bits int value that can be stored in a Json::Value. - static const Int64 maxInt64; - /// Maximum unsigned 64 bits int value that can be stored in a Json::Value. - static const UInt64 maxUInt64; - - private: -#ifndef JSONCPP_DOC_EXCLUDE_IMPLEMENTATION -# ifndef JSON_VALUE_USE_INTERNAL_MAP - class CZString - { - public: - enum DuplicationPolicy - { - noDuplication = 0, - duplicate, - duplicateOnCopy - }; - CZString( ArrayIndex index ); - CZString( const char *cstr, DuplicationPolicy allocate ); - CZString( const CZString &other ); - ~CZString(); - CZString &operator =( const CZString &other ); - bool operator<( const CZString &other ) const; - bool operator==( const CZString &other ) const; - ArrayIndex index() const; - const char *c_str() const; - bool isStaticString() const; - private: - void swap( CZString &other ); - const char *cstr_; - ArrayIndex index_; - }; - - public: -# ifndef JSON_USE_CPPTL_SMALLMAP - typedef std::map ObjectValues; -# else - typedef CppTL::SmallMap ObjectValues; -# endif // ifndef JSON_USE_CPPTL_SMALLMAP -# endif // ifndef JSON_VALUE_USE_INTERNAL_MAP -#endif // ifndef JSONCPP_DOC_EXCLUDE_IMPLEMENTATION - - public: - /** \brief Create a default Value of the given type. - - This is a very useful constructor. - To create an empty array, pass arrayValue. - To create an empty object, pass objectValue. - Another Value can then be set to this one by assignment. - This is useful since clear() and resize() will not alter types. - - Examples: - \code - Json::Value null_value; // null - Json::Value arr_value(Json::arrayValue); // [] - Json::Value obj_value(Json::objectValue); // {} - \endcode - */ - Value( ValueType type = nullValue ); - Value( Int value ); - Value( UInt value ); -#if defined(JSON_HAS_INT64) - Value( Int64 value ); - Value( UInt64 value ); -#endif // if defined(JSON_HAS_INT64) - Value( double value ); - Value( const char *value ); - Value( const char *beginValue, const char *endValue ); - /** \brief Constructs a value from a static string. - - * Like other value string constructor but do not duplicate the string for - * internal storage. The given string must remain alive after the call to this - * constructor. - * Example of usage: - * \code - * Json::Value aValue( StaticString("some text") ); - * \endcode - */ - Value( const StaticString &value ); - Value( const std::string &value ); -# ifdef JSON_USE_CPPTL - Value( const CppTL::ConstString &value ); -# endif - Value( bool value ); - Value( const Value &other ); - ~Value(); - - Value &operator=( const Value &other ); - /// Swap values. - /// \note Currently, comments are intentionally not swapped, for - /// both logic and efficiency. - void swap( Value &other ); - - ValueType type() const; - - bool operator <( const Value &other ) const; - bool operator <=( const Value &other ) const; - bool operator >=( const Value &other ) const; - bool operator >( const Value &other ) const; - - bool operator ==( const Value &other ) const; - bool operator !=( const Value &other ) const; - - int compare( const Value &other ) const; - - const char *asCString() const; - std::string asString() const; -# ifdef JSON_USE_CPPTL - CppTL::ConstString asConstString() const; -# endif - Int asInt() const; - UInt asUInt() const; - Int64 asInt64() const; - UInt64 asUInt64() const; - LargestInt asLargestInt() const; - LargestUInt asLargestUInt() const; - float asFloat() const; - double asDouble() const; - bool asBool() const; - - bool isNull() const; - bool isBool() const; - bool isInt() const; - bool isUInt() const; - bool isIntegral() const; - bool isDouble() const; - bool isNumeric() const; - bool isString() const; - bool isArray() const; - bool isObject() const; - - bool isConvertibleTo( ValueType other ) const; - - /// Number of values in array or object - ArrayIndex size() const; - - /// \brief Return true if empty array, empty object, or null; - /// otherwise, false. - bool empty() const; - - /// Return isNull() - bool operator!() const; - - /// Remove all object members and array elements. - /// \pre type() is arrayValue, objectValue, or nullValue - /// \post type() is unchanged - void clear(); - - /// Resize the array to size elements. - /// New elements are initialized to null. - /// May only be called on nullValue or arrayValue. - /// \pre type() is arrayValue or nullValue - /// \post type() is arrayValue - void resize( ArrayIndex size ); - - /// Access an array element (zero based index ). - /// If the array contains less than index element, then null value are inserted - /// in the array so that its size is index+1. - /// (You may need to say 'value[0u]' to get your compiler to distinguish - /// this from the operator[] which takes a string.) - Value &operator[]( ArrayIndex index ); - - /// Access an array element (zero based index ). - /// If the array contains less than index element, then null value are inserted - /// in the array so that its size is index+1. - /// (You may need to say 'value[0u]' to get your compiler to distinguish - /// this from the operator[] which takes a string.) - Value &operator[]( int index ); - - /// Access an array element (zero based index ) - /// (You may need to say 'value[0u]' to get your compiler to distinguish - /// this from the operator[] which takes a string.) - const Value &operator[]( ArrayIndex index ) const; - - /// Access an array element (zero based index ) - /// (You may need to say 'value[0u]' to get your compiler to distinguish - /// this from the operator[] which takes a string.) - const Value &operator[]( int index ) const; - - /// If the array contains at least index+1 elements, returns the element value, - /// otherwise returns defaultValue. - Value get( ArrayIndex index, - const Value &defaultValue ) const; - /// Return true if index < size(). - bool isValidIndex( ArrayIndex index ) const; - /// \brief Append value to array at the end. - /// - /// Equivalent to jsonvalue[jsonvalue.size()] = value; - Value &append( const Value &value ); - - /// Access an object value by name, create a null member if it does not exist. - Value &operator[]( const char *key ); - /// Access an object value by name, returns null if there is no member with that name. - const Value &operator[]( const char *key ) const; - /// Access an object value by name, create a null member if it does not exist. - Value &operator[]( const std::string &key ); - /// Access an object value by name, returns null if there is no member with that name. - const Value &operator[]( const std::string &key ) const; - /** \brief Access an object value by name, create a null member if it does not exist. - - * If the object as no entry for that name, then the member name used to store - * the new entry is not duplicated. - * Example of use: - * \code - * Json::Value object; - * static const StaticString code("code"); - * object[code] = 1234; - * \endcode - */ - Value &operator[]( const StaticString &key ); -# ifdef JSON_USE_CPPTL - /// Access an object value by name, create a null member if it does not exist. - Value &operator[]( const CppTL::ConstString &key ); - /// Access an object value by name, returns null if there is no member with that name. - const Value &operator[]( const CppTL::ConstString &key ) const; -# endif - /// Return the member named key if it exist, defaultValue otherwise. - Value get( const char *key, - const Value &defaultValue ) const; - /// Return the member named key if it exist, defaultValue otherwise. - Value get( const std::string &key, - const Value &defaultValue ) const; -# ifdef JSON_USE_CPPTL - /// Return the member named key if it exist, defaultValue otherwise. - Value get( const CppTL::ConstString &key, - const Value &defaultValue ) const; -# endif - /// \brief Remove and return the named member. - /// - /// Do nothing if it did not exist. - /// \return the removed Value, or null. - /// \pre type() is objectValue or nullValue - /// \post type() is unchanged - Value removeMember( const char* key ); - /// Same as removeMember(const char*) - Value removeMember( const std::string &key ); - - /// Return true if the object has a member named key. - bool isMember( const char *key ) const; - /// Return true if the object has a member named key. - bool isMember( const std::string &key ) const; -# ifdef JSON_USE_CPPTL - /// Return true if the object has a member named key. - bool isMember( const CppTL::ConstString &key ) const; -# endif - - /// \brief Return a list of the member names. - /// - /// If null, return an empty list. - /// \pre type() is objectValue or nullValue - /// \post if type() was nullValue, it remains nullValue - Members getMemberNames() const; - -//# ifdef JSON_USE_CPPTL -// EnumMemberNames enumMemberNames() const; -// EnumValues enumValues() const; -//# endif - - /// Comments must be //... or /* ... */ - void setComment( const char *comment, - CommentPlacement placement ); - /// Comments must be //... or /* ... */ - void setComment( const std::string &comment, - CommentPlacement placement ); - bool hasComment( CommentPlacement placement ) const; - /// Include delimiters and embedded newlines. - std::string getComment( CommentPlacement placement ) const; - - std::string toStyledString() const; - - const_iterator begin() const; - const_iterator end() const; - - iterator begin(); - iterator end(); - - private: - Value &resolveReference( const char *key, - bool isStatic ); - -# ifdef JSON_VALUE_USE_INTERNAL_MAP - inline bool isItemAvailable() const - { - return itemIsUsed_ == 0; - } - - inline void setItemUsed( bool isUsed = true ) - { - itemIsUsed_ = isUsed ? 1 : 0; - } - - inline bool isMemberNameStatic() const - { - return memberNameIsStatic_ == 0; - } - - inline void setMemberNameIsStatic( bool isStatic ) - { - memberNameIsStatic_ = isStatic ? 1 : 0; - } -# endif // # ifdef JSON_VALUE_USE_INTERNAL_MAP - - private: - struct CommentInfo - { - CommentInfo(); - ~CommentInfo(); - - void setComment( const char *text ); - - char *comment_; - }; - - //struct MemberNamesTransform - //{ - // typedef const char *result_type; - // const char *operator()( const CZString &name ) const - // { - // return name.c_str(); - // } - //}; - - union ValueHolder - { - LargestInt int_; - LargestUInt uint_; - double real_; - bool bool_; - char *string_; -# ifdef JSON_VALUE_USE_INTERNAL_MAP - ValueInternalArray *array_; - ValueInternalMap *map_; -#else - ObjectValues *map_; -# endif - } value_; - ValueType type_ : 8; - int allocated_ : 1; // Notes: if declared as bool, bitfield is useless. -# ifdef JSON_VALUE_USE_INTERNAL_MAP - unsigned int itemIsUsed_ : 1; // used by the ValueInternalMap container. - int memberNameIsStatic_ : 1; // used by the ValueInternalMap container. -# endif - CommentInfo *comments_; - }; - - - /** \brief Experimental and untested: represents an element of the "path" to access a node. - */ - class PathArgument - { - public: - friend class Path; - - PathArgument(); - PathArgument( ArrayIndex index ); - PathArgument( const char *key ); - PathArgument( const std::string &key ); - - private: - enum Kind - { - kindNone = 0, - kindIndex, - kindKey - }; - std::string key_; - ArrayIndex index_; - Kind kind_; - }; - - /** \brief Experimental and untested: represents a "path" to access a node. - * - * Syntax: - * - "." => root node - * - ".[n]" => elements at index 'n' of root node (an array value) - * - ".name" => member named 'name' of root node (an object value) - * - ".name1.name2.name3" - * - ".[0][1][2].name1[3]" - * - ".%" => member name is provided as parameter - * - ".[%]" => index is provied as parameter - */ - class Path - { - public: - Path( const std::string &path, - const PathArgument &a1 = PathArgument(), - const PathArgument &a2 = PathArgument(), - const PathArgument &a3 = PathArgument(), - const PathArgument &a4 = PathArgument(), - const PathArgument &a5 = PathArgument() ); - - const Value &resolve( const Value &root ) const; - Value resolve( const Value &root, - const Value &defaultValue ) const; - /// Creates the "path" to access the specified node and returns a reference on the node. - Value &make( Value &root ) const; - - private: - typedef std::vector InArgs; - typedef std::vector Args; - - void makePath( const std::string &path, - const InArgs &in ); - void addPathInArg( const std::string &path, - const InArgs &in, - InArgs::const_iterator &itInArg, - PathArgument::Kind kind ); - void invalidPath( const std::string &path, - int location ); - - Args args_; - }; - - - -#ifdef JSON_VALUE_USE_INTERNAL_MAP - /** \brief Allocator to customize Value internal map. - * Below is an example of a simple implementation (default implementation actually - * use memory pool for speed). - * \code - class DefaultValueMapAllocator : public ValueMapAllocator - { - public: // overridden from ValueMapAllocator - virtual ValueInternalMap *newMap() - { - return new ValueInternalMap(); - } - - virtual ValueInternalMap *newMapCopy( const ValueInternalMap &other ) - { - return new ValueInternalMap( other ); - } - - virtual void destructMap( ValueInternalMap *map ) - { - delete map; - } - - virtual ValueInternalLink *allocateMapBuckets( unsigned int size ) - { - return new ValueInternalLink[size]; - } - - virtual void releaseMapBuckets( ValueInternalLink *links ) - { - delete [] links; - } - - virtual ValueInternalLink *allocateMapLink() - { - return new ValueInternalLink(); - } - - virtual void releaseMapLink( ValueInternalLink *link ) - { - delete link; - } - }; - * \endcode - */ - class JSON_API ValueMapAllocator - { - public: - virtual ~ValueMapAllocator(); - virtual ValueInternalMap *newMap() = 0; - virtual ValueInternalMap *newMapCopy( const ValueInternalMap &other ) = 0; - virtual void destructMap( ValueInternalMap *map ) = 0; - virtual ValueInternalLink *allocateMapBuckets( unsigned int size ) = 0; - virtual void releaseMapBuckets( ValueInternalLink *links ) = 0; - virtual ValueInternalLink *allocateMapLink() = 0; - virtual void releaseMapLink( ValueInternalLink *link ) = 0; - }; - - /** \brief ValueInternalMap hash-map bucket chain link (for internal use only). - * \internal previous_ & next_ allows for bidirectional traversal. - */ - class JSON_API ValueInternalLink - { - public: - enum { itemPerLink = 6 }; // sizeof(ValueInternalLink) = 128 on 32 bits architecture. - enum InternalFlags { - flagAvailable = 0, - flagUsed = 1 - }; - - ValueInternalLink(); - - ~ValueInternalLink(); - - Value items_[itemPerLink]; - char *keys_[itemPerLink]; - ValueInternalLink *previous_; - ValueInternalLink *next_; - }; - - - /** \brief A linked page based hash-table implementation used internally by Value. - * \internal ValueInternalMap is a tradional bucket based hash-table, with a linked - * list in each bucket to handle collision. There is an addional twist in that - * each node of the collision linked list is a page containing a fixed amount of - * value. This provides a better compromise between memory usage and speed. - * - * Each bucket is made up of a chained list of ValueInternalLink. The last - * link of a given bucket can be found in the 'previous_' field of the following bucket. - * The last link of the last bucket is stored in tailLink_ as it has no following bucket. - * Only the last link of a bucket may contains 'available' item. The last link always - * contains at least one element unless is it the bucket one very first link. - */ - class JSON_API ValueInternalMap - { - friend class ValueIteratorBase; - friend class Value; - public: - typedef unsigned int HashKey; - typedef unsigned int BucketIndex; - -# ifndef JSONCPP_DOC_EXCLUDE_IMPLEMENTATION - struct IteratorState - { - IteratorState() - : map_(0) - , link_(0) - , itemIndex_(0) - , bucketIndex_(0) - { - } - ValueInternalMap *map_; - ValueInternalLink *link_; - BucketIndex itemIndex_; - BucketIndex bucketIndex_; - }; -# endif // ifndef JSONCPP_DOC_EXCLUDE_IMPLEMENTATION - - ValueInternalMap(); - ValueInternalMap( const ValueInternalMap &other ); - ValueInternalMap &operator =( const ValueInternalMap &other ); - ~ValueInternalMap(); - - void swap( ValueInternalMap &other ); - - BucketIndex size() const; - - void clear(); - - bool reserveDelta( BucketIndex growth ); - - bool reserve( BucketIndex newItemCount ); - - const Value *find( const char *key ) const; - - Value *find( const char *key ); - - Value &resolveReference( const char *key, - bool isStatic ); - - void remove( const char *key ); - - void doActualRemove( ValueInternalLink *link, - BucketIndex index, - BucketIndex bucketIndex ); - - ValueInternalLink *&getLastLinkInBucket( BucketIndex bucketIndex ); - - Value &setNewItem( const char *key, - bool isStatic, - ValueInternalLink *link, - BucketIndex index ); - - Value &unsafeAdd( const char *key, - bool isStatic, - HashKey hashedKey ); - - HashKey hash( const char *key ) const; - - int compare( const ValueInternalMap &other ) const; - - private: - void makeBeginIterator( IteratorState &it ) const; - void makeEndIterator( IteratorState &it ) const; - static bool equals( const IteratorState &x, const IteratorState &other ); - static void increment( IteratorState &iterator ); - static void incrementBucket( IteratorState &iterator ); - static void decrement( IteratorState &iterator ); - static const char *key( const IteratorState &iterator ); - static const char *key( const IteratorState &iterator, bool &isStatic ); - static Value &value( const IteratorState &iterator ); - static int distance( const IteratorState &x, const IteratorState &y ); - - private: - ValueInternalLink *buckets_; - ValueInternalLink *tailLink_; - BucketIndex bucketsSize_; - BucketIndex itemCount_; - }; - - /** \brief A simplified deque implementation used internally by Value. - * \internal - * It is based on a list of fixed "page", each page contains a fixed number of items. - * Instead of using a linked-list, a array of pointer is used for fast item look-up. - * Look-up for an element is as follow: - * - compute page index: pageIndex = itemIndex / itemsPerPage - * - look-up item in page: pages_[pageIndex][itemIndex % itemsPerPage] - * - * Insertion is amortized constant time (only the array containing the index of pointers - * need to be reallocated when items are appended). - */ - class JSON_API ValueInternalArray - { - friend class Value; - friend class ValueIteratorBase; - public: - enum { itemsPerPage = 8 }; // should be a power of 2 for fast divide and modulo. - typedef Value::ArrayIndex ArrayIndex; - typedef unsigned int PageIndex; - -# ifndef JSONCPP_DOC_EXCLUDE_IMPLEMENTATION - struct IteratorState // Must be a POD - { - IteratorState() - : array_(0) - , currentPageIndex_(0) - , currentItemIndex_(0) - { - } - ValueInternalArray *array_; - Value **currentPageIndex_; - unsigned int currentItemIndex_; - }; -# endif // ifndef JSONCPP_DOC_EXCLUDE_IMPLEMENTATION - - ValueInternalArray(); - ValueInternalArray( const ValueInternalArray &other ); - ValueInternalArray &operator =( const ValueInternalArray &other ); - ~ValueInternalArray(); - void swap( ValueInternalArray &other ); - - void clear(); - void resize( ArrayIndex newSize ); - - Value &resolveReference( ArrayIndex index ); - - Value *find( ArrayIndex index ) const; - - ArrayIndex size() const; - - int compare( const ValueInternalArray &other ) const; - - private: - static bool equals( const IteratorState &x, const IteratorState &other ); - static void increment( IteratorState &iterator ); - static void decrement( IteratorState &iterator ); - static Value &dereference( const IteratorState &iterator ); - static Value &unsafeDereference( const IteratorState &iterator ); - static int distance( const IteratorState &x, const IteratorState &y ); - static ArrayIndex indexOf( const IteratorState &iterator ); - void makeBeginIterator( IteratorState &it ) const; - void makeEndIterator( IteratorState &it ) const; - void makeIterator( IteratorState &it, ArrayIndex index ) const; - - void makeIndexValid( ArrayIndex index ); - - Value **pages_; - ArrayIndex size_; - PageIndex pageCount_; - }; - - /** \brief Experimental: do not use. Allocator to customize Value internal array. - * Below is an example of a simple implementation (actual implementation use - * memory pool). - \code -class DefaultValueArrayAllocator : public ValueArrayAllocator -{ -public: // overridden from ValueArrayAllocator - virtual ~DefaultValueArrayAllocator() - { - } - - virtual ValueInternalArray *newArray() - { - return new ValueInternalArray(); - } - - virtual ValueInternalArray *newArrayCopy( const ValueInternalArray &other ) - { - return new ValueInternalArray( other ); - } - - virtual void destruct( ValueInternalArray *array ) - { - delete array; - } - - virtual void reallocateArrayPageIndex( Value **&indexes, - ValueInternalArray::PageIndex &indexCount, - ValueInternalArray::PageIndex minNewIndexCount ) - { - ValueInternalArray::PageIndex newIndexCount = (indexCount*3)/2 + 1; - if ( minNewIndexCount > newIndexCount ) - newIndexCount = minNewIndexCount; - void *newIndexes = realloc( indexes, sizeof(Value*) * newIndexCount ); - if ( !newIndexes ) - throw std::bad_alloc(); - indexCount = newIndexCount; - indexes = static_cast( newIndexes ); - } - virtual void releaseArrayPageIndex( Value **indexes, - ValueInternalArray::PageIndex indexCount ) - { - if ( indexes ) - free( indexes ); - } - - virtual Value *allocateArrayPage() - { - return static_cast( malloc( sizeof(Value) * ValueInternalArray::itemsPerPage ) ); - } - - virtual void releaseArrayPage( Value *value ) - { - if ( value ) - free( value ); - } -}; - \endcode - */ - class JSON_API ValueArrayAllocator - { - public: - virtual ~ValueArrayAllocator(); - virtual ValueInternalArray *newArray() = 0; - virtual ValueInternalArray *newArrayCopy( const ValueInternalArray &other ) = 0; - virtual void destructArray( ValueInternalArray *array ) = 0; - /** \brief Reallocate array page index. - * Reallocates an array of pointer on each page. - * \param indexes [input] pointer on the current index. May be \c NULL. - * [output] pointer on the new index of at least - * \a minNewIndexCount pages. - * \param indexCount [input] current number of pages in the index. - * [output] number of page the reallocated index can handle. - * \b MUST be >= \a minNewIndexCount. - * \param minNewIndexCount Minimum number of page the new index must be able to - * handle. - */ - virtual void reallocateArrayPageIndex( Value **&indexes, - ValueInternalArray::PageIndex &indexCount, - ValueInternalArray::PageIndex minNewIndexCount ) = 0; - virtual void releaseArrayPageIndex( Value **indexes, - ValueInternalArray::PageIndex indexCount ) = 0; - virtual Value *allocateArrayPage() = 0; - virtual void releaseArrayPage( Value *value ) = 0; - }; -#endif // #ifdef JSON_VALUE_USE_INTERNAL_MAP - - - /** \brief base class for Value iterators. - * - */ - class ValueIteratorBase - { - public: - typedef unsigned int size_t; - typedef int difference_type; - typedef ValueIteratorBase SelfType; - - ValueIteratorBase(); -#ifndef JSON_VALUE_USE_INTERNAL_MAP - explicit ValueIteratorBase( const Value::ObjectValues::iterator ¤t ); -#else - ValueIteratorBase( const ValueInternalArray::IteratorState &state ); - ValueIteratorBase( const ValueInternalMap::IteratorState &state ); -#endif - - bool operator ==( const SelfType &other ) const - { - return isEqual( other ); - } - - bool operator !=( const SelfType &other ) const - { - return !isEqual( other ); - } - - difference_type operator -( const SelfType &other ) const - { - return computeDistance( other ); - } - - /// Return either the index or the member name of the referenced value as a Value. - Value key() const; - - /// Return the index of the referenced Value. -1 if it is not an arrayValue. - UInt index() const; - - /// Return the member name of the referenced Value. "" if it is not an objectValue. - const char *memberName() const; - - protected: - Value &deref() const; - - void increment(); - - void decrement(); - - difference_type computeDistance( const SelfType &other ) const; - - bool isEqual( const SelfType &other ) const; - - void copy( const SelfType &other ); - - private: -#ifndef JSON_VALUE_USE_INTERNAL_MAP - Value::ObjectValues::iterator current_; - // Indicates that iterator is for a null value. - bool isNull_; -#else - union - { - ValueInternalArray::IteratorState array_; - ValueInternalMap::IteratorState map_; - } iterator_; - bool isArray_; -#endif - }; - - /** \brief const iterator for object and array value. - * - */ - class ValueConstIterator : public ValueIteratorBase - { - friend class Value; - public: - typedef unsigned int size_t; - typedef int difference_type; - typedef const Value &reference; - typedef const Value *pointer; - typedef ValueConstIterator SelfType; - - ValueConstIterator(); - private: - /*! \internal Use by Value to create an iterator. - */ -#ifndef JSON_VALUE_USE_INTERNAL_MAP - explicit ValueConstIterator( const Value::ObjectValues::iterator ¤t ); -#else - ValueConstIterator( const ValueInternalArray::IteratorState &state ); - ValueConstIterator( const ValueInternalMap::IteratorState &state ); -#endif - public: - SelfType &operator =( const ValueIteratorBase &other ); - - SelfType operator++( int ) - { - SelfType temp( *this ); - ++*this; - return temp; - } - - SelfType operator--( int ) - { - SelfType temp( *this ); - --*this; - return temp; - } - - SelfType &operator--() - { - decrement(); - return *this; - } - - SelfType &operator++() - { - increment(); - return *this; - } - - reference operator *() const - { - return deref(); - } - }; - - - /** \brief Iterator for object and array value. - */ - class ValueIterator : public ValueIteratorBase - { - friend class Value; - public: - typedef unsigned int size_t; - typedef int difference_type; - typedef Value &reference; - typedef Value *pointer; - typedef ValueIterator SelfType; - - ValueIterator(); - ValueIterator( const ValueConstIterator &other ); - ValueIterator( const ValueIterator &other ); - private: - /*! \internal Use by Value to create an iterator. - */ -#ifndef JSON_VALUE_USE_INTERNAL_MAP - explicit ValueIterator( const Value::ObjectValues::iterator ¤t ); -#else - ValueIterator( const ValueInternalArray::IteratorState &state ); - ValueIterator( const ValueInternalMap::IteratorState &state ); -#endif - public: - - SelfType &operator =( const SelfType &other ); - - SelfType operator++( int ) - { - SelfType temp( *this ); - ++*this; - return temp; - } - - SelfType operator--( int ) - { - SelfType temp( *this ); - --*this; - return temp; - } - - SelfType &operator--() - { - decrement(); - return *this; - } - - SelfType &operator++() - { - increment(); - return *this; - } - - reference operator *() const - { - return deref(); - } - }; - - -} // namespace Json - - -#endif // CPPTL_JSON_H_INCLUDED diff --git a/ext/jsoncpp/include/json/writer.h b/ext/jsoncpp/include/json/writer.h deleted file mode 100644 index 47893636..00000000 --- a/ext/jsoncpp/include/json/writer.h +++ /dev/null @@ -1,185 +0,0 @@ -// Copyright 2007-2010 Baptiste Lepilleur -// Distributed under MIT license, or public domain if desired and -// recognized in your jurisdiction. -// See file LICENSE for detail or copy at http://jsoncpp.sourceforge.net/LICENSE - -#ifndef JSON_WRITER_H_INCLUDED -# define JSON_WRITER_H_INCLUDED - -#if !defined(JSON_IS_AMALGAMATION) -# include "value.h" -#endif // if !defined(JSON_IS_AMALGAMATION) -# include -# include -# include - -namespace Json { - - class Value; - - /** \brief Abstract class for writers. - */ - class JSON_API Writer - { - public: - virtual ~Writer(); - - virtual std::string write( const Value &root ) = 0; - }; - - /** \brief Outputs a Value in JSON format without formatting (not human friendly). - * - * The JSON document is written in a single line. It is not intended for 'human' consumption, - * but may be usefull to support feature such as RPC where bandwith is limited. - * \sa Reader, Value - */ - class JSON_API FastWriter : public Writer - { - public: - FastWriter(); - virtual ~FastWriter(){} - - void enableYAMLCompatibility(); - - public: // overridden from Writer - virtual std::string write( const Value &root ); - - private: - void writeValue( const Value &value ); - - std::string document_; - bool yamlCompatiblityEnabled_; - }; - - /** \brief Writes a Value in JSON format in a human friendly way. - * - * The rules for line break and indent are as follow: - * - Object value: - * - if empty then print {} without indent and line break - * - if not empty the print '{', line break & indent, print one value per line - * and then unindent and line break and print '}'. - * - Array value: - * - if empty then print [] without indent and line break - * - if the array contains no object value, empty array or some other value types, - * and all the values fit on one lines, then print the array on a single line. - * - otherwise, it the values do not fit on one line, or the array contains - * object or non empty array, then print one value per line. - * - * If the Value have comments then they are outputed according to their #CommentPlacement. - * - * \sa Reader, Value, Value::setComment() - */ - class JSON_API StyledWriter: public Writer - { - public: - StyledWriter(); - virtual ~StyledWriter(){} - - public: // overridden from Writer - /** \brief Serialize a Value in JSON format. - * \param root Value to serialize. - * \return String containing the JSON document that represents the root value. - */ - virtual std::string write( const Value &root ); - - private: - void writeValue( const Value &value ); - void writeArrayValue( const Value &value ); - bool isMultineArray( const Value &value ); - void pushValue( const std::string &value ); - void writeIndent(); - void writeWithIndent( const std::string &value ); - void indent(); - void unindent(); - void writeCommentBeforeValue( const Value &root ); - void writeCommentAfterValueOnSameLine( const Value &root ); - bool hasCommentForValue( const Value &value ); - static std::string normalizeEOL( const std::string &text ); - - typedef std::vector ChildValues; - - ChildValues childValues_; - std::string document_; - std::string indentString_; - int rightMargin_; - int indentSize_; - bool addChildValues_; - }; - - /** \brief Writes a Value in JSON format in a human friendly way, - to a stream rather than to a string. - * - * The rules for line break and indent are as follow: - * - Object value: - * - if empty then print {} without indent and line break - * - if not empty the print '{', line break & indent, print one value per line - * and then unindent and line break and print '}'. - * - Array value: - * - if empty then print [] without indent and line break - * - if the array contains no object value, empty array or some other value types, - * and all the values fit on one lines, then print the array on a single line. - * - otherwise, it the values do not fit on one line, or the array contains - * object or non empty array, then print one value per line. - * - * If the Value have comments then they are outputed according to their #CommentPlacement. - * - * \param indentation Each level will be indented by this amount extra. - * \sa Reader, Value, Value::setComment() - */ - class JSON_API StyledStreamWriter - { - public: - StyledStreamWriter( std::string indentation="\t" ); - ~StyledStreamWriter(){} - - public: - /** \brief Serialize a Value in JSON format. - * \param out Stream to write to. (Can be ostringstream, e.g.) - * \param root Value to serialize. - * \note There is no point in deriving from Writer, since write() should not return a value. - */ - void write( std::ostream &out, const Value &root ); - - private: - void writeValue( const Value &value ); - void writeArrayValue( const Value &value ); - bool isMultineArray( const Value &value ); - void pushValue( const std::string &value ); - void writeIndent(); - void writeWithIndent( const std::string &value ); - void indent(); - void unindent(); - void writeCommentBeforeValue( const Value &root ); - void writeCommentAfterValueOnSameLine( const Value &root ); - bool hasCommentForValue( const Value &value ); - static std::string normalizeEOL( const std::string &text ); - - typedef std::vector ChildValues; - - ChildValues childValues_; - std::ostream* document_; - std::string indentString_; - int rightMargin_; - std::string indentation_; - bool addChildValues_; - }; - -# if defined(JSON_HAS_INT64) - std::string JSON_API valueToString( Int value ); - std::string JSON_API valueToString( UInt value ); -# endif // if defined(JSON_HAS_INT64) - std::string JSON_API valueToString( LargestInt value ); - std::string JSON_API valueToString( LargestUInt value ); - std::string JSON_API valueToString( double value ); - std::string JSON_API valueToString( bool value ); - std::string JSON_API valueToQuotedString( const char *value ); - - /// \brief Output using the StyledStreamWriter. - /// \see Json::operator>>() - std::ostream& operator<<( std::ostream&, const Value &root ); - -} // namespace Json - - - -#endif // JSON_WRITER_H_INCLUDED diff --git a/ext/jsoncpp/src/json_batchallocator.h b/ext/jsoncpp/src/json_batchallocator.h deleted file mode 100644 index 173e2ed2..00000000 --- a/ext/jsoncpp/src/json_batchallocator.h +++ /dev/null @@ -1,130 +0,0 @@ -// Copyright 2007-2010 Baptiste Lepilleur -// Distributed under MIT license, or public domain if desired and -// recognized in your jurisdiction. -// See file LICENSE for detail or copy at http://jsoncpp.sourceforge.net/LICENSE - -#ifndef JSONCPP_BATCHALLOCATOR_H_INCLUDED -# define JSONCPP_BATCHALLOCATOR_H_INCLUDED - -# include -# include - -# ifndef JSONCPP_DOC_EXCLUDE_IMPLEMENTATION - -namespace Json { - -/* Fast memory allocator. - * - * This memory allocator allocates memory for a batch of object (specified by - * the page size, the number of object in each page). - * - * It does not allow the destruction of a single object. All the allocated objects - * can be destroyed at once. The memory can be either released or reused for future - * allocation. - * - * The in-place new operator must be used to construct the object using the pointer - * returned by allocate. - */ -template -class BatchAllocator -{ -public: - typedef AllocatedType Type; - - BatchAllocator( unsigned int objectsPerPage = 255 ) - : freeHead_( 0 ) - , objectsPerPage_( objectsPerPage ) - { -// printf( "Size: %d => %s\n", sizeof(AllocatedType), typeid(AllocatedType).name() ); - assert( sizeof(AllocatedType) * objectPerAllocation >= sizeof(AllocatedType *) ); // We must be able to store a slist in the object free space. - assert( objectsPerPage >= 16 ); - batches_ = allocateBatch( 0 ); // allocated a dummy page - currentBatch_ = batches_; - } - - ~BatchAllocator() - { - for ( BatchInfo *batch = batches_; batch; ) - { - BatchInfo *nextBatch = batch->next_; - free( batch ); - batch = nextBatch; - } - } - - /// allocate space for an array of objectPerAllocation object. - /// @warning it is the responsability of the caller to call objects constructors. - AllocatedType *allocate() - { - if ( freeHead_ ) // returns node from free list. - { - AllocatedType *object = freeHead_; - freeHead_ = *(AllocatedType **)object; - return object; - } - if ( currentBatch_->used_ == currentBatch_->end_ ) - { - currentBatch_ = currentBatch_->next_; - while ( currentBatch_ && currentBatch_->used_ == currentBatch_->end_ ) - currentBatch_ = currentBatch_->next_; - - if ( !currentBatch_ ) // no free batch found, allocate a new one - { - currentBatch_ = allocateBatch( objectsPerPage_ ); - currentBatch_->next_ = batches_; // insert at the head of the list - batches_ = currentBatch_; - } - } - AllocatedType *allocated = currentBatch_->used_; - currentBatch_->used_ += objectPerAllocation; - return allocated; - } - - /// Release the object. - /// @warning it is the responsability of the caller to actually destruct the object. - void release( AllocatedType *object ) - { - assert( object != 0 ); - *(AllocatedType **)object = freeHead_; - freeHead_ = object; - } - -private: - struct BatchInfo - { - BatchInfo *next_; - AllocatedType *used_; - AllocatedType *end_; - AllocatedType buffer_[objectPerAllocation]; - }; - - // disabled copy constructor and assignement operator. - BatchAllocator( const BatchAllocator & ); - void operator =( const BatchAllocator &); - - static BatchInfo *allocateBatch( unsigned int objectsPerPage ) - { - const unsigned int mallocSize = sizeof(BatchInfo) - sizeof(AllocatedType)* objectPerAllocation - + sizeof(AllocatedType) * objectPerAllocation * objectsPerPage; - BatchInfo *batch = static_cast( malloc( mallocSize ) ); - batch->next_ = 0; - batch->used_ = batch->buffer_; - batch->end_ = batch->buffer_ + objectsPerPage; - return batch; - } - - BatchInfo *batches_; - BatchInfo *currentBatch_; - /// Head of a single linked list within the allocated space of freeed object - AllocatedType *freeHead_; - unsigned int objectsPerPage_; -}; - - -} // namespace Json - -# endif // ifndef JSONCPP_DOC_INCLUDE_IMPLEMENTATION - -#endif // JSONCPP_BATCHALLOCATOR_H_INCLUDED - diff --git a/ext/jsoncpp/src/json_internalarray.inl b/ext/jsoncpp/src/json_internalarray.inl deleted file mode 100644 index 3a532ad7..00000000 --- a/ext/jsoncpp/src/json_internalarray.inl +++ /dev/null @@ -1,456 +0,0 @@ -// Copyright 2007-2010 Baptiste Lepilleur -// Distributed under MIT license, or public domain if desired and -// recognized in your jurisdiction. -// See file LICENSE for detail or copy at http://jsoncpp.sourceforge.net/LICENSE - -// included by json_value.cpp - -namespace Json { - -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// class ValueInternalArray -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// - -ValueArrayAllocator::~ValueArrayAllocator() -{ -} - -// ////////////////////////////////////////////////////////////////// -// class DefaultValueArrayAllocator -// ////////////////////////////////////////////////////////////////// -#ifdef JSON_USE_SIMPLE_INTERNAL_ALLOCATOR -class DefaultValueArrayAllocator : public ValueArrayAllocator -{ -public: // overridden from ValueArrayAllocator - virtual ~DefaultValueArrayAllocator() - { - } - - virtual ValueInternalArray *newArray() - { - return new ValueInternalArray(); - } - - virtual ValueInternalArray *newArrayCopy( const ValueInternalArray &other ) - { - return new ValueInternalArray( other ); - } - - virtual void destructArray( ValueInternalArray *array ) - { - delete array; - } - - virtual void reallocateArrayPageIndex( Value **&indexes, - ValueInternalArray::PageIndex &indexCount, - ValueInternalArray::PageIndex minNewIndexCount ) - { - ValueInternalArray::PageIndex newIndexCount = (indexCount*3)/2 + 1; - if ( minNewIndexCount > newIndexCount ) - newIndexCount = minNewIndexCount; - void *newIndexes = realloc( indexes, sizeof(Value*) * newIndexCount ); - if ( !newIndexes ) - throw std::bad_alloc(); - indexCount = newIndexCount; - indexes = static_cast( newIndexes ); - } - virtual void releaseArrayPageIndex( Value **indexes, - ValueInternalArray::PageIndex indexCount ) - { - if ( indexes ) - free( indexes ); - } - - virtual Value *allocateArrayPage() - { - return static_cast( malloc( sizeof(Value) * ValueInternalArray::itemsPerPage ) ); - } - - virtual void releaseArrayPage( Value *value ) - { - if ( value ) - free( value ); - } -}; - -#else // #ifdef JSON_USE_SIMPLE_INTERNAL_ALLOCATOR -/// @todo make this thread-safe (lock when accessign batch allocator) -class DefaultValueArrayAllocator : public ValueArrayAllocator -{ -public: // overridden from ValueArrayAllocator - virtual ~DefaultValueArrayAllocator() - { - } - - virtual ValueInternalArray *newArray() - { - ValueInternalArray *array = arraysAllocator_.allocate(); - new (array) ValueInternalArray(); // placement new - return array; - } - - virtual ValueInternalArray *newArrayCopy( const ValueInternalArray &other ) - { - ValueInternalArray *array = arraysAllocator_.allocate(); - new (array) ValueInternalArray( other ); // placement new - return array; - } - - virtual void destructArray( ValueInternalArray *array ) - { - if ( array ) - { - array->~ValueInternalArray(); - arraysAllocator_.release( array ); - } - } - - virtual void reallocateArrayPageIndex( Value **&indexes, - ValueInternalArray::PageIndex &indexCount, - ValueInternalArray::PageIndex minNewIndexCount ) - { - ValueInternalArray::PageIndex newIndexCount = (indexCount*3)/2 + 1; - if ( minNewIndexCount > newIndexCount ) - newIndexCount = minNewIndexCount; - void *newIndexes = realloc( indexes, sizeof(Value*) * newIndexCount ); - if ( !newIndexes ) - throw std::bad_alloc(); - indexCount = newIndexCount; - indexes = static_cast( newIndexes ); - } - virtual void releaseArrayPageIndex( Value **indexes, - ValueInternalArray::PageIndex indexCount ) - { - if ( indexes ) - free( indexes ); - } - - virtual Value *allocateArrayPage() - { - return static_cast( pagesAllocator_.allocate() ); - } - - virtual void releaseArrayPage( Value *value ) - { - if ( value ) - pagesAllocator_.release( value ); - } -private: - BatchAllocator arraysAllocator_; - BatchAllocator pagesAllocator_; -}; -#endif // #ifdef JSON_USE_SIMPLE_INTERNAL_ALLOCATOR - -static ValueArrayAllocator *&arrayAllocator() -{ - static DefaultValueArrayAllocator defaultAllocator; - static ValueArrayAllocator *arrayAllocator = &defaultAllocator; - return arrayAllocator; -} - -static struct DummyArrayAllocatorInitializer { - DummyArrayAllocatorInitializer() - { - arrayAllocator(); // ensure arrayAllocator() statics are initialized before main(). - } -} dummyArrayAllocatorInitializer; - -// ////////////////////////////////////////////////////////////////// -// class ValueInternalArray -// ////////////////////////////////////////////////////////////////// -bool -ValueInternalArray::equals( const IteratorState &x, - const IteratorState &other ) -{ - return x.array_ == other.array_ - && x.currentItemIndex_ == other.currentItemIndex_ - && x.currentPageIndex_ == other.currentPageIndex_; -} - - -void -ValueInternalArray::increment( IteratorState &it ) -{ - JSON_ASSERT_MESSAGE( it.array_ && - (it.currentPageIndex_ - it.array_->pages_)*itemsPerPage + it.currentItemIndex_ - != it.array_->size_, - "ValueInternalArray::increment(): moving iterator beyond end" ); - ++(it.currentItemIndex_); - if ( it.currentItemIndex_ == itemsPerPage ) - { - it.currentItemIndex_ = 0; - ++(it.currentPageIndex_); - } -} - - -void -ValueInternalArray::decrement( IteratorState &it ) -{ - JSON_ASSERT_MESSAGE( it.array_ && it.currentPageIndex_ == it.array_->pages_ - && it.currentItemIndex_ == 0, - "ValueInternalArray::decrement(): moving iterator beyond end" ); - if ( it.currentItemIndex_ == 0 ) - { - it.currentItemIndex_ = itemsPerPage-1; - --(it.currentPageIndex_); - } - else - { - --(it.currentItemIndex_); - } -} - - -Value & -ValueInternalArray::unsafeDereference( const IteratorState &it ) -{ - return (*(it.currentPageIndex_))[it.currentItemIndex_]; -} - - -Value & -ValueInternalArray::dereference( const IteratorState &it ) -{ - JSON_ASSERT_MESSAGE( it.array_ && - (it.currentPageIndex_ - it.array_->pages_)*itemsPerPage + it.currentItemIndex_ - < it.array_->size_, - "ValueInternalArray::dereference(): dereferencing invalid iterator" ); - return unsafeDereference( it ); -} - -void -ValueInternalArray::makeBeginIterator( IteratorState &it ) const -{ - it.array_ = const_cast( this ); - it.currentItemIndex_ = 0; - it.currentPageIndex_ = pages_; -} - - -void -ValueInternalArray::makeIterator( IteratorState &it, ArrayIndex index ) const -{ - it.array_ = const_cast( this ); - it.currentItemIndex_ = index % itemsPerPage; - it.currentPageIndex_ = pages_ + index / itemsPerPage; -} - - -void -ValueInternalArray::makeEndIterator( IteratorState &it ) const -{ - makeIterator( it, size_ ); -} - - -ValueInternalArray::ValueInternalArray() - : pages_( 0 ) - , size_( 0 ) - , pageCount_( 0 ) -{ -} - - -ValueInternalArray::ValueInternalArray( const ValueInternalArray &other ) - : pages_( 0 ) - , pageCount_( 0 ) - , size_( other.size_ ) -{ - PageIndex minNewPages = other.size_ / itemsPerPage; - arrayAllocator()->reallocateArrayPageIndex( pages_, pageCount_, minNewPages ); - JSON_ASSERT_MESSAGE( pageCount_ >= minNewPages, - "ValueInternalArray::reserve(): bad reallocation" ); - IteratorState itOther; - other.makeBeginIterator( itOther ); - Value *value; - for ( ArrayIndex index = 0; index < size_; ++index, increment(itOther) ) - { - if ( index % itemsPerPage == 0 ) - { - PageIndex pageIndex = index / itemsPerPage; - value = arrayAllocator()->allocateArrayPage(); - pages_[pageIndex] = value; - } - new (value) Value( dereference( itOther ) ); - } -} - - -ValueInternalArray & -ValueInternalArray::operator =( const ValueInternalArray &other ) -{ - ValueInternalArray temp( other ); - swap( temp ); - return *this; -} - - -ValueInternalArray::~ValueInternalArray() -{ - // destroy all constructed items - IteratorState it; - IteratorState itEnd; - makeBeginIterator( it); - makeEndIterator( itEnd ); - for ( ; !equals(it,itEnd); increment(it) ) - { - Value *value = &dereference(it); - value->~Value(); - } - // release all pages - PageIndex lastPageIndex = size_ / itemsPerPage; - for ( PageIndex pageIndex = 0; pageIndex < lastPageIndex; ++pageIndex ) - arrayAllocator()->releaseArrayPage( pages_[pageIndex] ); - // release pages index - arrayAllocator()->releaseArrayPageIndex( pages_, pageCount_ ); -} - - -void -ValueInternalArray::swap( ValueInternalArray &other ) -{ - Value **tempPages = pages_; - pages_ = other.pages_; - other.pages_ = tempPages; - ArrayIndex tempSize = size_; - size_ = other.size_; - other.size_ = tempSize; - PageIndex tempPageCount = pageCount_; - pageCount_ = other.pageCount_; - other.pageCount_ = tempPageCount; -} - -void -ValueInternalArray::clear() -{ - ValueInternalArray dummy; - swap( dummy ); -} - - -void -ValueInternalArray::resize( ArrayIndex newSize ) -{ - if ( newSize == 0 ) - clear(); - else if ( newSize < size_ ) - { - IteratorState it; - IteratorState itEnd; - makeIterator( it, newSize ); - makeIterator( itEnd, size_ ); - for ( ; !equals(it,itEnd); increment(it) ) - { - Value *value = &dereference(it); - value->~Value(); - } - PageIndex pageIndex = (newSize + itemsPerPage - 1) / itemsPerPage; - PageIndex lastPageIndex = size_ / itemsPerPage; - for ( ; pageIndex < lastPageIndex; ++pageIndex ) - arrayAllocator()->releaseArrayPage( pages_[pageIndex] ); - size_ = newSize; - } - else if ( newSize > size_ ) - resolveReference( newSize ); -} - - -void -ValueInternalArray::makeIndexValid( ArrayIndex index ) -{ - // Need to enlarge page index ? - if ( index >= pageCount_ * itemsPerPage ) - { - PageIndex minNewPages = (index + 1) / itemsPerPage; - arrayAllocator()->reallocateArrayPageIndex( pages_, pageCount_, minNewPages ); - JSON_ASSERT_MESSAGE( pageCount_ >= minNewPages, "ValueInternalArray::reserve(): bad reallocation" ); - } - - // Need to allocate new pages ? - ArrayIndex nextPageIndex = - (size_ % itemsPerPage) != 0 ? size_ - (size_%itemsPerPage) + itemsPerPage - : size_; - if ( nextPageIndex <= index ) - { - PageIndex pageIndex = nextPageIndex / itemsPerPage; - PageIndex pageToAllocate = (index - nextPageIndex) / itemsPerPage + 1; - for ( ; pageToAllocate-- > 0; ++pageIndex ) - pages_[pageIndex] = arrayAllocator()->allocateArrayPage(); - } - - // Initialize all new entries - IteratorState it; - IteratorState itEnd; - makeIterator( it, size_ ); - size_ = index + 1; - makeIterator( itEnd, size_ ); - for ( ; !equals(it,itEnd); increment(it) ) - { - Value *value = &dereference(it); - new (value) Value(); // Construct a default value using placement new - } -} - -Value & -ValueInternalArray::resolveReference( ArrayIndex index ) -{ - if ( index >= size_ ) - makeIndexValid( index ); - return pages_[index/itemsPerPage][index%itemsPerPage]; -} - -Value * -ValueInternalArray::find( ArrayIndex index ) const -{ - if ( index >= size_ ) - return 0; - return &(pages_[index/itemsPerPage][index%itemsPerPage]); -} - -ValueInternalArray::ArrayIndex -ValueInternalArray::size() const -{ - return size_; -} - -int -ValueInternalArray::distance( const IteratorState &x, const IteratorState &y ) -{ - return indexOf(y) - indexOf(x); -} - - -ValueInternalArray::ArrayIndex -ValueInternalArray::indexOf( const IteratorState &iterator ) -{ - if ( !iterator.array_ ) - return ArrayIndex(-1); - return ArrayIndex( - (iterator.currentPageIndex_ - iterator.array_->pages_) * itemsPerPage - + iterator.currentItemIndex_ ); -} - - -int -ValueInternalArray::compare( const ValueInternalArray &other ) const -{ - int sizeDiff( size_ - other.size_ ); - if ( sizeDiff != 0 ) - return sizeDiff; - - for ( ArrayIndex index =0; index < size_; ++index ) - { - int diff = pages_[index/itemsPerPage][index%itemsPerPage].compare( - other.pages_[index/itemsPerPage][index%itemsPerPage] ); - if ( diff != 0 ) - return diff; - } - return 0; -} - -} // namespace Json diff --git a/ext/jsoncpp/src/json_internalmap.inl b/ext/jsoncpp/src/json_internalmap.inl deleted file mode 100644 index f2fa1606..00000000 --- a/ext/jsoncpp/src/json_internalmap.inl +++ /dev/null @@ -1,615 +0,0 @@ -// Copyright 2007-2010 Baptiste Lepilleur -// Distributed under MIT license, or public domain if desired and -// recognized in your jurisdiction. -// See file LICENSE for detail or copy at http://jsoncpp.sourceforge.net/LICENSE - -// included by json_value.cpp - -namespace Json { - -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// class ValueInternalMap -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// - -/** \internal MUST be safely initialized using memset( this, 0, sizeof(ValueInternalLink) ); - * This optimization is used by the fast allocator. - */ -ValueInternalLink::ValueInternalLink() - : previous_( 0 ) - , next_( 0 ) -{ -} - -ValueInternalLink::~ValueInternalLink() -{ - for ( int index =0; index < itemPerLink; ++index ) - { - if ( !items_[index].isItemAvailable() ) - { - if ( !items_[index].isMemberNameStatic() ) - free( keys_[index] ); - } - else - break; - } -} - - - -ValueMapAllocator::~ValueMapAllocator() -{ -} - -#ifdef JSON_USE_SIMPLE_INTERNAL_ALLOCATOR -class DefaultValueMapAllocator : public ValueMapAllocator -{ -public: // overridden from ValueMapAllocator - virtual ValueInternalMap *newMap() - { - return new ValueInternalMap(); - } - - virtual ValueInternalMap *newMapCopy( const ValueInternalMap &other ) - { - return new ValueInternalMap( other ); - } - - virtual void destructMap( ValueInternalMap *map ) - { - delete map; - } - - virtual ValueInternalLink *allocateMapBuckets( unsigned int size ) - { - return new ValueInternalLink[size]; - } - - virtual void releaseMapBuckets( ValueInternalLink *links ) - { - delete [] links; - } - - virtual ValueInternalLink *allocateMapLink() - { - return new ValueInternalLink(); - } - - virtual void releaseMapLink( ValueInternalLink *link ) - { - delete link; - } -}; -#else -/// @todo make this thread-safe (lock when accessign batch allocator) -class DefaultValueMapAllocator : public ValueMapAllocator -{ -public: // overridden from ValueMapAllocator - virtual ValueInternalMap *newMap() - { - ValueInternalMap *map = mapsAllocator_.allocate(); - new (map) ValueInternalMap(); // placement new - return map; - } - - virtual ValueInternalMap *newMapCopy( const ValueInternalMap &other ) - { - ValueInternalMap *map = mapsAllocator_.allocate(); - new (map) ValueInternalMap( other ); // placement new - return map; - } - - virtual void destructMap( ValueInternalMap *map ) - { - if ( map ) - { - map->~ValueInternalMap(); - mapsAllocator_.release( map ); - } - } - - virtual ValueInternalLink *allocateMapBuckets( unsigned int size ) - { - return new ValueInternalLink[size]; - } - - virtual void releaseMapBuckets( ValueInternalLink *links ) - { - delete [] links; - } - - virtual ValueInternalLink *allocateMapLink() - { - ValueInternalLink *link = linksAllocator_.allocate(); - memset( link, 0, sizeof(ValueInternalLink) ); - return link; - } - - virtual void releaseMapLink( ValueInternalLink *link ) - { - link->~ValueInternalLink(); - linksAllocator_.release( link ); - } -private: - BatchAllocator mapsAllocator_; - BatchAllocator linksAllocator_; -}; -#endif - -static ValueMapAllocator *&mapAllocator() -{ - static DefaultValueMapAllocator defaultAllocator; - static ValueMapAllocator *mapAllocator = &defaultAllocator; - return mapAllocator; -} - -static struct DummyMapAllocatorInitializer { - DummyMapAllocatorInitializer() - { - mapAllocator(); // ensure mapAllocator() statics are initialized before main(). - } -} dummyMapAllocatorInitializer; - - - -// h(K) = value * K >> w ; with w = 32 & K prime w.r.t. 2^32. - -/* -use linked list hash map. -buckets array is a container. -linked list element contains 6 key/values. (memory = (16+4) * 6 + 4 = 124) -value have extra state: valid, available, deleted -*/ - - -ValueInternalMap::ValueInternalMap() - : buckets_( 0 ) - , tailLink_( 0 ) - , bucketsSize_( 0 ) - , itemCount_( 0 ) -{ -} - - -ValueInternalMap::ValueInternalMap( const ValueInternalMap &other ) - : buckets_( 0 ) - , tailLink_( 0 ) - , bucketsSize_( 0 ) - , itemCount_( 0 ) -{ - reserve( other.itemCount_ ); - IteratorState it; - IteratorState itEnd; - other.makeBeginIterator( it ); - other.makeEndIterator( itEnd ); - for ( ; !equals(it,itEnd); increment(it) ) - { - bool isStatic; - const char *memberName = key( it, isStatic ); - const Value &aValue = value( it ); - resolveReference(memberName, isStatic) = aValue; - } -} - - -ValueInternalMap & -ValueInternalMap::operator =( const ValueInternalMap &other ) -{ - ValueInternalMap dummy( other ); - swap( dummy ); - return *this; -} - - -ValueInternalMap::~ValueInternalMap() -{ - if ( buckets_ ) - { - for ( BucketIndex bucketIndex =0; bucketIndex < bucketsSize_; ++bucketIndex ) - { - ValueInternalLink *link = buckets_[bucketIndex].next_; - while ( link ) - { - ValueInternalLink *linkToRelease = link; - link = link->next_; - mapAllocator()->releaseMapLink( linkToRelease ); - } - } - mapAllocator()->releaseMapBuckets( buckets_ ); - } -} - - -void -ValueInternalMap::swap( ValueInternalMap &other ) -{ - ValueInternalLink *tempBuckets = buckets_; - buckets_ = other.buckets_; - other.buckets_ = tempBuckets; - ValueInternalLink *tempTailLink = tailLink_; - tailLink_ = other.tailLink_; - other.tailLink_ = tempTailLink; - BucketIndex tempBucketsSize = bucketsSize_; - bucketsSize_ = other.bucketsSize_; - other.bucketsSize_ = tempBucketsSize; - BucketIndex tempItemCount = itemCount_; - itemCount_ = other.itemCount_; - other.itemCount_ = tempItemCount; -} - - -void -ValueInternalMap::clear() -{ - ValueInternalMap dummy; - swap( dummy ); -} - - -ValueInternalMap::BucketIndex -ValueInternalMap::size() const -{ - return itemCount_; -} - -bool -ValueInternalMap::reserveDelta( BucketIndex growth ) -{ - return reserve( itemCount_ + growth ); -} - -bool -ValueInternalMap::reserve( BucketIndex newItemCount ) -{ - if ( !buckets_ && newItemCount > 0 ) - { - buckets_ = mapAllocator()->allocateMapBuckets( 1 ); - bucketsSize_ = 1; - tailLink_ = &buckets_[0]; - } -// BucketIndex idealBucketCount = (newItemCount + ValueInternalLink::itemPerLink) / ValueInternalLink::itemPerLink; - return true; -} - - -const Value * -ValueInternalMap::find( const char *key ) const -{ - if ( !bucketsSize_ ) - return 0; - HashKey hashedKey = hash( key ); - BucketIndex bucketIndex = hashedKey % bucketsSize_; - for ( const ValueInternalLink *current = &buckets_[bucketIndex]; - current != 0; - current = current->next_ ) - { - for ( BucketIndex index=0; index < ValueInternalLink::itemPerLink; ++index ) - { - if ( current->items_[index].isItemAvailable() ) - return 0; - if ( strcmp( key, current->keys_[index] ) == 0 ) - return ¤t->items_[index]; - } - } - return 0; -} - - -Value * -ValueInternalMap::find( const char *key ) -{ - const ValueInternalMap *constThis = this; - return const_cast( constThis->find( key ) ); -} - - -Value & -ValueInternalMap::resolveReference( const char *key, - bool isStatic ) -{ - HashKey hashedKey = hash( key ); - if ( bucketsSize_ ) - { - BucketIndex bucketIndex = hashedKey % bucketsSize_; - ValueInternalLink **previous = 0; - BucketIndex index; - for ( ValueInternalLink *current = &buckets_[bucketIndex]; - current != 0; - previous = ¤t->next_, current = current->next_ ) - { - for ( index=0; index < ValueInternalLink::itemPerLink; ++index ) - { - if ( current->items_[index].isItemAvailable() ) - return setNewItem( key, isStatic, current, index ); - if ( strcmp( key, current->keys_[index] ) == 0 ) - return current->items_[index]; - } - } - } - - reserveDelta( 1 ); - return unsafeAdd( key, isStatic, hashedKey ); -} - - -void -ValueInternalMap::remove( const char *key ) -{ - HashKey hashedKey = hash( key ); - if ( !bucketsSize_ ) - return; - BucketIndex bucketIndex = hashedKey % bucketsSize_; - for ( ValueInternalLink *link = &buckets_[bucketIndex]; - link != 0; - link = link->next_ ) - { - BucketIndex index; - for ( index =0; index < ValueInternalLink::itemPerLink; ++index ) - { - if ( link->items_[index].isItemAvailable() ) - return; - if ( strcmp( key, link->keys_[index] ) == 0 ) - { - doActualRemove( link, index, bucketIndex ); - return; - } - } - } -} - -void -ValueInternalMap::doActualRemove( ValueInternalLink *link, - BucketIndex index, - BucketIndex bucketIndex ) -{ - // find last item of the bucket and swap it with the 'removed' one. - // set removed items flags to 'available'. - // if last page only contains 'available' items, then desallocate it (it's empty) - ValueInternalLink *&lastLink = getLastLinkInBucket( index ); - BucketIndex lastItemIndex = 1; // a link can never be empty, so start at 1 - for ( ; - lastItemIndex < ValueInternalLink::itemPerLink; - ++lastItemIndex ) // may be optimized with dicotomic search - { - if ( lastLink->items_[lastItemIndex].isItemAvailable() ) - break; - } - - BucketIndex lastUsedIndex = lastItemIndex - 1; - Value *valueToDelete = &link->items_[index]; - Value *valueToPreserve = &lastLink->items_[lastUsedIndex]; - if ( valueToDelete != valueToPreserve ) - valueToDelete->swap( *valueToPreserve ); - if ( lastUsedIndex == 0 ) // page is now empty - { // remove it from bucket linked list and delete it. - ValueInternalLink *linkPreviousToLast = lastLink->previous_; - if ( linkPreviousToLast != 0 ) // can not deleted bucket link. - { - mapAllocator()->releaseMapLink( lastLink ); - linkPreviousToLast->next_ = 0; - lastLink = linkPreviousToLast; - } - } - else - { - Value dummy; - valueToPreserve->swap( dummy ); // restore deleted to default Value. - valueToPreserve->setItemUsed( false ); - } - --itemCount_; -} - - -ValueInternalLink *& -ValueInternalMap::getLastLinkInBucket( BucketIndex bucketIndex ) -{ - if ( bucketIndex == bucketsSize_ - 1 ) - return tailLink_; - ValueInternalLink *&previous = buckets_[bucketIndex+1].previous_; - if ( !previous ) - previous = &buckets_[bucketIndex]; - return previous; -} - - -Value & -ValueInternalMap::setNewItem( const char *key, - bool isStatic, - ValueInternalLink *link, - BucketIndex index ) -{ - char *duplicatedKey = makeMemberName( key ); - ++itemCount_; - link->keys_[index] = duplicatedKey; - link->items_[index].setItemUsed(); - link->items_[index].setMemberNameIsStatic( isStatic ); - return link->items_[index]; // items already default constructed. -} - - -Value & -ValueInternalMap::unsafeAdd( const char *key, - bool isStatic, - HashKey hashedKey ) -{ - JSON_ASSERT_MESSAGE( bucketsSize_ > 0, "ValueInternalMap::unsafeAdd(): internal logic error." ); - BucketIndex bucketIndex = hashedKey % bucketsSize_; - ValueInternalLink *&previousLink = getLastLinkInBucket( bucketIndex ); - ValueInternalLink *link = previousLink; - BucketIndex index; - for ( index =0; index < ValueInternalLink::itemPerLink; ++index ) - { - if ( link->items_[index].isItemAvailable() ) - break; - } - if ( index == ValueInternalLink::itemPerLink ) // need to add a new page - { - ValueInternalLink *newLink = mapAllocator()->allocateMapLink(); - index = 0; - link->next_ = newLink; - previousLink = newLink; - link = newLink; - } - return setNewItem( key, isStatic, link, index ); -} - - -ValueInternalMap::HashKey -ValueInternalMap::hash( const char *key ) const -{ - HashKey hash = 0; - while ( *key ) - hash += *key++ * 37; - return hash; -} - - -int -ValueInternalMap::compare( const ValueInternalMap &other ) const -{ - int sizeDiff( itemCount_ - other.itemCount_ ); - if ( sizeDiff != 0 ) - return sizeDiff; - // Strict order guaranty is required. Compare all keys FIRST, then compare values. - IteratorState it; - IteratorState itEnd; - makeBeginIterator( it ); - makeEndIterator( itEnd ); - for ( ; !equals(it,itEnd); increment(it) ) - { - if ( !other.find( key( it ) ) ) - return 1; - } - - // All keys are equals, let's compare values - makeBeginIterator( it ); - for ( ; !equals(it,itEnd); increment(it) ) - { - const Value *otherValue = other.find( key( it ) ); - int valueDiff = value(it).compare( *otherValue ); - if ( valueDiff != 0 ) - return valueDiff; - } - return 0; -} - - -void -ValueInternalMap::makeBeginIterator( IteratorState &it ) const -{ - it.map_ = const_cast( this ); - it.bucketIndex_ = 0; - it.itemIndex_ = 0; - it.link_ = buckets_; -} - - -void -ValueInternalMap::makeEndIterator( IteratorState &it ) const -{ - it.map_ = const_cast( this ); - it.bucketIndex_ = bucketsSize_; - it.itemIndex_ = 0; - it.link_ = 0; -} - - -bool -ValueInternalMap::equals( const IteratorState &x, const IteratorState &other ) -{ - return x.map_ == other.map_ - && x.bucketIndex_ == other.bucketIndex_ - && x.link_ == other.link_ - && x.itemIndex_ == other.itemIndex_; -} - - -void -ValueInternalMap::incrementBucket( IteratorState &iterator ) -{ - ++iterator.bucketIndex_; - JSON_ASSERT_MESSAGE( iterator.bucketIndex_ <= iterator.map_->bucketsSize_, - "ValueInternalMap::increment(): attempting to iterate beyond end." ); - if ( iterator.bucketIndex_ == iterator.map_->bucketsSize_ ) - iterator.link_ = 0; - else - iterator.link_ = &(iterator.map_->buckets_[iterator.bucketIndex_]); - iterator.itemIndex_ = 0; -} - - -void -ValueInternalMap::increment( IteratorState &iterator ) -{ - JSON_ASSERT_MESSAGE( iterator.map_, "Attempting to iterator using invalid iterator." ); - ++iterator.itemIndex_; - if ( iterator.itemIndex_ == ValueInternalLink::itemPerLink ) - { - JSON_ASSERT_MESSAGE( iterator.link_ != 0, - "ValueInternalMap::increment(): attempting to iterate beyond end." ); - iterator.link_ = iterator.link_->next_; - if ( iterator.link_ == 0 ) - incrementBucket( iterator ); - } - else if ( iterator.link_->items_[iterator.itemIndex_].isItemAvailable() ) - { - incrementBucket( iterator ); - } -} - - -void -ValueInternalMap::decrement( IteratorState &iterator ) -{ - if ( iterator.itemIndex_ == 0 ) - { - JSON_ASSERT_MESSAGE( iterator.map_, "Attempting to iterate using invalid iterator." ); - if ( iterator.link_ == &iterator.map_->buckets_[iterator.bucketIndex_] ) - { - JSON_ASSERT_MESSAGE( iterator.bucketIndex_ > 0, "Attempting to iterate beyond beginning." ); - --(iterator.bucketIndex_); - } - iterator.link_ = iterator.link_->previous_; - iterator.itemIndex_ = ValueInternalLink::itemPerLink - 1; - } -} - - -const char * -ValueInternalMap::key( const IteratorState &iterator ) -{ - JSON_ASSERT_MESSAGE( iterator.link_, "Attempting to iterate using invalid iterator." ); - return iterator.link_->keys_[iterator.itemIndex_]; -} - -const char * -ValueInternalMap::key( const IteratorState &iterator, bool &isStatic ) -{ - JSON_ASSERT_MESSAGE( iterator.link_, "Attempting to iterate using invalid iterator." ); - isStatic = iterator.link_->items_[iterator.itemIndex_].isMemberNameStatic(); - return iterator.link_->keys_[iterator.itemIndex_]; -} - - -Value & -ValueInternalMap::value( const IteratorState &iterator ) -{ - JSON_ASSERT_MESSAGE( iterator.link_, "Attempting to iterate using invalid iterator." ); - return iterator.link_->items_[iterator.itemIndex_]; -} - - -int -ValueInternalMap::distance( const IteratorState &x, const IteratorState &y ) -{ - int offset = 0; - IteratorState it = x; - while ( !equals( it, y ) ) - increment( it ); - return offset; -} - -} // namespace Json diff --git a/ext/jsoncpp/src/json_reader.cpp b/ext/jsoncpp/src/json_reader.cpp deleted file mode 100644 index 8bb03042..00000000 --- a/ext/jsoncpp/src/json_reader.cpp +++ /dev/null @@ -1,880 +0,0 @@ -// Copyright 2007-2010 Baptiste Lepilleur -// Distributed under MIT license, or public domain if desired and -// recognized in your jurisdiction. -// See file LICENSE for detail or copy at http://jsoncpp.sourceforge.net/LICENSE - -#if !defined(JSON_IS_AMALGAMATION) -# include -# include -# include "json_tool.h" -#endif // if !defined(JSON_IS_AMALGAMATION) -#include -#include -#include -#include -#include -#include - -#if _MSC_VER >= 1400 // VC++ 8.0 -#pragma warning( disable : 4996 ) // disable warning about strdup being deprecated. -#endif - -namespace Json { - -// Implementation of class Features -// //////////////////////////////// - -Features::Features() - : allowComments_( true ) - , strictRoot_( false ) -{ -} - - -Features -Features::all() -{ - return Features(); -} - - -Features -Features::strictMode() -{ - Features features; - features.allowComments_ = false; - features.strictRoot_ = true; - return features; -} - -// Implementation of class Reader -// //////////////////////////////// - - -static inline bool -in( Reader::Char c, Reader::Char c1, Reader::Char c2, Reader::Char c3, Reader::Char c4 ) -{ - return c == c1 || c == c2 || c == c3 || c == c4; -} - -static inline bool -in( Reader::Char c, Reader::Char c1, Reader::Char c2, Reader::Char c3, Reader::Char c4, Reader::Char c5 ) -{ - return c == c1 || c == c2 || c == c3 || c == c4 || c == c5; -} - - -static bool -containsNewLine( Reader::Location begin, - Reader::Location end ) -{ - for ( ;begin < end; ++begin ) - if ( *begin == '\n' || *begin == '\r' ) - return true; - return false; -} - - -// Class Reader -// ////////////////////////////////////////////////////////////////// - -Reader::Reader() - : features_( Features::all() ) -{ -} - - -Reader::Reader( const Features &features ) - : features_( features ) -{ -} - - -bool -Reader::parse( const std::string &document, - Value &root, - bool collectComments ) -{ - document_ = document; - const char *begin = document_.c_str(); - const char *end = begin + document_.length(); - return parse( begin, end, root, collectComments ); -} - - -bool -Reader::parse( std::istream& sin, - Value &root, - bool collectComments ) -{ - //std::istream_iterator begin(sin); - //std::istream_iterator end; - // Those would allow streamed input from a file, if parse() were a - // template function. - - // Since std::string is reference-counted, this at least does not - // create an extra copy. - std::string doc; - std::getline(sin, doc, (char)EOF); - return parse( doc, root, collectComments ); -} - -bool -Reader::parse( const char *beginDoc, const char *endDoc, - Value &root, - bool collectComments ) -{ - if ( !features_.allowComments_ ) - { - collectComments = false; - } - - begin_ = beginDoc; - end_ = endDoc; - collectComments_ = collectComments; - current_ = begin_; - lastValueEnd_ = 0; - lastValue_ = 0; - commentsBefore_ = ""; - errors_.clear(); - while ( !nodes_.empty() ) - nodes_.pop(); - nodes_.push( &root ); - - bool successful = readValue(); - Token token; - skipCommentTokens( token ); - if ( collectComments_ && !commentsBefore_.empty() ) - root.setComment( commentsBefore_, commentAfter ); - if ( features_.strictRoot_ ) - { - if ( !root.isArray() && !root.isObject() ) - { - // Set error location to start of doc, ideally should be first token found in doc - token.type_ = tokenError; - token.start_ = beginDoc; - token.end_ = endDoc; - addError( "A valid JSON document must be either an array or an object value.", - token ); - return false; - } - } - return successful; -} - - -bool -Reader::readValue() -{ - Token token; - skipCommentTokens( token ); - bool successful = true; - - if ( collectComments_ && !commentsBefore_.empty() ) - { - currentValue().setComment( commentsBefore_, commentBefore ); - commentsBefore_ = ""; - } - - - switch ( token.type_ ) - { - case tokenObjectBegin: - successful = readObject( token ); - break; - case tokenArrayBegin: - successful = readArray( token ); - break; - case tokenNumber: - successful = decodeNumber( token ); - break; - case tokenString: - successful = decodeString( token ); - break; - case tokenTrue: - currentValue() = true; - break; - case tokenFalse: - currentValue() = false; - break; - case tokenNull: - currentValue() = Value(); - break; - default: - return addError( "Syntax error: value, object or array expected.", token ); - } - - if ( collectComments_ ) - { - lastValueEnd_ = current_; - lastValue_ = ¤tValue(); - } - - return successful; -} - - -void -Reader::skipCommentTokens( Token &token ) -{ - if ( features_.allowComments_ ) - { - do - { - readToken( token ); - } - while ( token.type_ == tokenComment ); - } - else - { - readToken( token ); - } -} - - -bool -Reader::expectToken( TokenType type, Token &token, const char *message ) -{ - readToken( token ); - if ( token.type_ != type ) - return addError( message, token ); - return true; -} - - -bool -Reader::readToken( Token &token ) -{ - skipSpaces(); - token.start_ = current_; - Char c = getNextChar(); - bool ok = true; - switch ( c ) - { - case '{': - token.type_ = tokenObjectBegin; - break; - case '}': - token.type_ = tokenObjectEnd; - break; - case '[': - token.type_ = tokenArrayBegin; - break; - case ']': - token.type_ = tokenArrayEnd; - break; - case '"': - token.type_ = tokenString; - ok = readString(); - break; - case '/': - token.type_ = tokenComment; - ok = readComment(); - break; - case '0': - case '1': - case '2': - case '3': - case '4': - case '5': - case '6': - case '7': - case '8': - case '9': - case '-': - token.type_ = tokenNumber; - readNumber(); - break; - case 't': - token.type_ = tokenTrue; - ok = match( "rue", 3 ); - break; - case 'f': - token.type_ = tokenFalse; - ok = match( "alse", 4 ); - break; - case 'n': - token.type_ = tokenNull; - ok = match( "ull", 3 ); - break; - case ',': - token.type_ = tokenArraySeparator; - break; - case ':': - token.type_ = tokenMemberSeparator; - break; - case 0: - token.type_ = tokenEndOfStream; - break; - default: - ok = false; - break; - } - if ( !ok ) - token.type_ = tokenError; - token.end_ = current_; - return true; -} - - -void -Reader::skipSpaces() -{ - while ( current_ != end_ ) - { - Char c = *current_; - if ( c == ' ' || c == '\t' || c == '\r' || c == '\n' ) - ++current_; - else - break; - } -} - - -bool -Reader::match( Location pattern, - int patternLength ) -{ - if ( end_ - current_ < patternLength ) - return false; - int index = patternLength; - while ( index-- ) - if ( current_[index] != pattern[index] ) - return false; - current_ += patternLength; - return true; -} - - -bool -Reader::readComment() -{ - Location commentBegin = current_ - 1; - Char c = getNextChar(); - bool successful = false; - if ( c == '*' ) - successful = readCStyleComment(); - else if ( c == '/' ) - successful = readCppStyleComment(); - if ( !successful ) - return false; - - if ( collectComments_ ) - { - CommentPlacement placement = commentBefore; - if ( lastValueEnd_ && !containsNewLine( lastValueEnd_, commentBegin ) ) - { - if ( c != '*' || !containsNewLine( commentBegin, current_ ) ) - placement = commentAfterOnSameLine; - } - - addComment( commentBegin, current_, placement ); - } - return true; -} - - -void -Reader::addComment( Location begin, - Location end, - CommentPlacement placement ) -{ - assert( collectComments_ ); - if ( placement == commentAfterOnSameLine ) - { - assert( lastValue_ != 0 ); - lastValue_->setComment( std::string( begin, end ), placement ); - } - else - { - if ( !commentsBefore_.empty() ) - commentsBefore_ += "\n"; - commentsBefore_ += std::string( begin, end ); - } -} - - -bool -Reader::readCStyleComment() -{ - while ( current_ != end_ ) - { - Char c = getNextChar(); - if ( c == '*' && *current_ == '/' ) - break; - } - return getNextChar() == '/'; -} - - -bool -Reader::readCppStyleComment() -{ - while ( current_ != end_ ) - { - Char c = getNextChar(); - if ( c == '\r' || c == '\n' ) - break; - } - return true; -} - - -void -Reader::readNumber() -{ - while ( current_ != end_ ) - { - if ( !(*current_ >= '0' && *current_ <= '9') && - !in( *current_, '.', 'e', 'E', '+', '-' ) ) - break; - ++current_; - } -} - -bool -Reader::readString() -{ - Char c = 0; - while ( current_ != end_ ) - { - c = getNextChar(); - if ( c == '\\' ) - getNextChar(); - else if ( c == '"' ) - break; - } - return c == '"'; -} - - -bool -Reader::readObject( Token &/*tokenStart*/ ) -{ - Token tokenName; - std::string name; - currentValue() = Value( objectValue ); - while ( readToken( tokenName ) ) - { - bool initialTokenOk = true; - while ( tokenName.type_ == tokenComment && initialTokenOk ) - initialTokenOk = readToken( tokenName ); - if ( !initialTokenOk ) - break; - if ( tokenName.type_ == tokenObjectEnd && name.empty() ) // empty object - return true; - if ( tokenName.type_ != tokenString ) - break; - - name = ""; - if ( !decodeString( tokenName, name ) ) - return recoverFromError( tokenObjectEnd ); - - Token colon; - if ( !readToken( colon ) || colon.type_ != tokenMemberSeparator ) - { - return addErrorAndRecover( "Missing ':' after object member name", - colon, - tokenObjectEnd ); - } - Value &value = currentValue()[ name ]; - nodes_.push( &value ); - bool ok = readValue(); - nodes_.pop(); - if ( !ok ) // error already set - return recoverFromError( tokenObjectEnd ); - - Token comma; - if ( !readToken( comma ) - || ( comma.type_ != tokenObjectEnd && - comma.type_ != tokenArraySeparator && - comma.type_ != tokenComment ) ) - { - return addErrorAndRecover( "Missing ',' or '}' in object declaration", - comma, - tokenObjectEnd ); - } - bool finalizeTokenOk = true; - while ( comma.type_ == tokenComment && - finalizeTokenOk ) - finalizeTokenOk = readToken( comma ); - if ( comma.type_ == tokenObjectEnd ) - return true; - } - return addErrorAndRecover( "Missing '}' or object member name", - tokenName, - tokenObjectEnd ); -} - - -bool -Reader::readArray( Token &/*tokenStart*/ ) -{ - currentValue() = Value( arrayValue ); - skipSpaces(); - if ( *current_ == ']' ) // empty array - { - Token endArray; - readToken( endArray ); - return true; - } - int index = 0; - for (;;) - { - Value &value = currentValue()[ index++ ]; - nodes_.push( &value ); - bool ok = readValue(); - nodes_.pop(); - if ( !ok ) // error already set - return recoverFromError( tokenArrayEnd ); - - Token token; - // Accept Comment after last item in the array. - ok = readToken( token ); - while ( token.type_ == tokenComment && ok ) - { - ok = readToken( token ); - } - bool badTokenType = ( token.type_ != tokenArraySeparator && - token.type_ != tokenArrayEnd ); - if ( !ok || badTokenType ) - { - return addErrorAndRecover( "Missing ',' or ']' in array declaration", - token, - tokenArrayEnd ); - } - if ( token.type_ == tokenArrayEnd ) - break; - } - return true; -} - - -bool -Reader::decodeNumber( Token &token ) -{ - bool isDouble = false; - for ( Location inspect = token.start_; inspect != token.end_; ++inspect ) - { - isDouble = isDouble - || in( *inspect, '.', 'e', 'E', '+' ) - || ( *inspect == '-' && inspect != token.start_ ); - } - if ( isDouble ) - return decodeDouble( token ); - // Attempts to parse the number as an integer. If the number is - // larger than the maximum supported value of an integer then - // we decode the number as a double. - Location current = token.start_; - bool isNegative = *current == '-'; - if ( isNegative ) - ++current; - Value::LargestUInt maxIntegerValue = isNegative ? Value::LargestUInt(-Value::minLargestInt) - : Value::maxLargestUInt; - Value::LargestUInt threshold = maxIntegerValue / 10; - Value::UInt lastDigitThreshold = Value::UInt( maxIntegerValue % 10 ); - assert( lastDigitThreshold >=0 && lastDigitThreshold <= 9 ); - Value::LargestUInt value = 0; - while ( current < token.end_ ) - { - Char c = *current++; - if ( c < '0' || c > '9' ) - return addError( "'" + std::string( token.start_, token.end_ ) + "' is not a number.", token ); - Value::UInt digit(c - '0'); - if ( value >= threshold ) - { - // If the current digit is not the last one, or if it is - // greater than the last digit of the maximum integer value, - // the parse the number as a double. - if ( current != token.end_ || digit > lastDigitThreshold ) - { - return decodeDouble( token ); - } - } - value = value * 10 + digit; - } - if ( isNegative ) - currentValue() = -Value::LargestInt( value ); - else if ( value <= Value::LargestUInt(Value::maxInt) ) - currentValue() = Value::LargestInt( value ); - else - currentValue() = value; - return true; -} - - -bool -Reader::decodeDouble( Token &token ) -{ - double value = 0; - const int bufferSize = 32; - int count; - int length = int(token.end_ - token.start_); - if ( length <= bufferSize ) - { - Char buffer[bufferSize+1]; - memcpy( buffer, token.start_, length ); - buffer[length] = 0; - count = sscanf( buffer, "%lf", &value ); - } - else - { - std::string buffer( token.start_, token.end_ ); - count = sscanf( buffer.c_str(), "%lf", &value ); - } - - if ( count != 1 ) - return addError( "'" + std::string( token.start_, token.end_ ) + "' is not a number.", token ); - currentValue() = value; - return true; -} - - -bool -Reader::decodeString( Token &token ) -{ - std::string decoded; - if ( !decodeString( token, decoded ) ) - return false; - currentValue() = decoded; - return true; -} - - -bool -Reader::decodeString( Token &token, std::string &decoded ) -{ - decoded.reserve( token.end_ - token.start_ - 2 ); - Location current = token.start_ + 1; // skip '"' - Location end = token.end_ - 1; // do not include '"' - while ( current != end ) - { - Char c = *current++; - if ( c == '"' ) - break; - else if ( c == '\\' ) - { - if ( current == end ) - return addError( "Empty escape sequence in string", token, current ); - Char escape = *current++; - switch ( escape ) - { - case '"': decoded += '"'; break; - case '/': decoded += '/'; break; - case '\\': decoded += '\\'; break; - case 'b': decoded += '\b'; break; - case 'f': decoded += '\f'; break; - case 'n': decoded += '\n'; break; - case 'r': decoded += '\r'; break; - case 't': decoded += '\t'; break; - case 'u': - { - unsigned int unicode; - if ( !decodeUnicodeCodePoint( token, current, end, unicode ) ) - return false; - decoded += codePointToUTF8(unicode); - } - break; - default: - return addError( "Bad escape sequence in string", token, current ); - } - } - else - { - decoded += c; - } - } - return true; -} - -bool -Reader::decodeUnicodeCodePoint( Token &token, - Location ¤t, - Location end, - unsigned int &unicode ) -{ - - if ( !decodeUnicodeEscapeSequence( token, current, end, unicode ) ) - return false; - if (unicode >= 0xD800 && unicode <= 0xDBFF) - { - // surrogate pairs - if (end - current < 6) - return addError( "additional six characters expected to parse unicode surrogate pair.", token, current ); - unsigned int surrogatePair; - if (*(current++) == '\\' && *(current++)== 'u') - { - if (decodeUnicodeEscapeSequence( token, current, end, surrogatePair )) - { - unicode = 0x10000 + ((unicode & 0x3FF) << 10) + (surrogatePair & 0x3FF); - } - else - return false; - } - else - return addError( "expecting another \\u token to begin the second half of a unicode surrogate pair", token, current ); - } - return true; -} - -bool -Reader::decodeUnicodeEscapeSequence( Token &token, - Location ¤t, - Location end, - unsigned int &unicode ) -{ - if ( end - current < 4 ) - return addError( "Bad unicode escape sequence in string: four digits expected.", token, current ); - unicode = 0; - for ( int index =0; index < 4; ++index ) - { - Char c = *current++; - unicode *= 16; - if ( c >= '0' && c <= '9' ) - unicode += c - '0'; - else if ( c >= 'a' && c <= 'f' ) - unicode += c - 'a' + 10; - else if ( c >= 'A' && c <= 'F' ) - unicode += c - 'A' + 10; - else - return addError( "Bad unicode escape sequence in string: hexadecimal digit expected.", token, current ); - } - return true; -} - - -bool -Reader::addError( const std::string &message, - Token &token, - Location extra ) -{ - ErrorInfo info; - info.token_ = token; - info.message_ = message; - info.extra_ = extra; - errors_.push_back( info ); - return false; -} - - -bool -Reader::recoverFromError( TokenType skipUntilToken ) -{ - int errorCount = int(errors_.size()); - Token skip; - for (;;) - { - if ( !readToken(skip) ) - errors_.resize( errorCount ); // discard errors caused by recovery - if ( skip.type_ == skipUntilToken || skip.type_ == tokenEndOfStream ) - break; - } - errors_.resize( errorCount ); - return false; -} - - -bool -Reader::addErrorAndRecover( const std::string &message, - Token &token, - TokenType skipUntilToken ) -{ - addError( message, token ); - return recoverFromError( skipUntilToken ); -} - - -Value & -Reader::currentValue() -{ - return *(nodes_.top()); -} - - -Reader::Char -Reader::getNextChar() -{ - if ( current_ == end_ ) - return 0; - return *current_++; -} - - -void -Reader::getLocationLineAndColumn( Location location, - int &line, - int &column ) const -{ - Location current = begin_; - Location lastLineStart = current; - line = 0; - while ( current < location && current != end_ ) - { - Char c = *current++; - if ( c == '\r' ) - { - if ( *current == '\n' ) - ++current; - lastLineStart = current; - ++line; - } - else if ( c == '\n' ) - { - lastLineStart = current; - ++line; - } - } - // column & line start at 1 - column = int(location - lastLineStart) + 1; - ++line; -} - - -std::string -Reader::getLocationLineAndColumn( Location location ) const -{ - int line, column; - getLocationLineAndColumn( location, line, column ); - char buffer[18+16+16+1]; - sprintf( buffer, "Line %d, Column %d", line, column ); - return buffer; -} - - -// Deprecated. Preserved for backward compatibility -std::string -Reader::getFormatedErrorMessages() const -{ - return getFormattedErrorMessages(); -} - - -std::string -Reader::getFormattedErrorMessages() const -{ - std::string formattedMessage; - for ( Errors::const_iterator itError = errors_.begin(); - itError != errors_.end(); - ++itError ) - { - const ErrorInfo &error = *itError; - formattedMessage += "* " + getLocationLineAndColumn( error.token_.start_ ) + "\n"; - formattedMessage += " " + error.message_ + "\n"; - if ( error.extra_ ) - formattedMessage += "See " + getLocationLineAndColumn( error.extra_ ) + " for detail.\n"; - } - return formattedMessage; -} - - -std::istream& operator>>( std::istream &sin, Value &root ) -{ - Json::Reader reader; - bool ok = reader.parse(sin, root, true); - //JSON_ASSERT( ok ); - if (!ok) throw std::runtime_error(reader.getFormattedErrorMessages()); - return sin; -} - - -} // namespace Json diff --git a/ext/jsoncpp/src/json_tool.h b/ext/jsoncpp/src/json_tool.h deleted file mode 100644 index 658031bb..00000000 --- a/ext/jsoncpp/src/json_tool.h +++ /dev/null @@ -1,93 +0,0 @@ -// Copyright 2007-2010 Baptiste Lepilleur -// Distributed under MIT license, or public domain if desired and -// recognized in your jurisdiction. -// See file LICENSE for detail or copy at http://jsoncpp.sourceforge.net/LICENSE - -#ifndef LIB_JSONCPP_JSON_TOOL_H_INCLUDED -# define LIB_JSONCPP_JSON_TOOL_H_INCLUDED - -/* This header provides common string manipulation support, such as UTF-8, - * portable conversion from/to string... - * - * It is an internal header that must not be exposed. - */ - -namespace Json { - -/// Converts a unicode code-point to UTF-8. -static inline std::string -codePointToUTF8(unsigned int cp) -{ - std::string result; - - // based on description from http://en.wikipedia.org/wiki/UTF-8 - - if (cp <= 0x7f) - { - result.resize(1); - result[0] = static_cast(cp); - } - else if (cp <= 0x7FF) - { - result.resize(2); - result[1] = static_cast(0x80 | (0x3f & cp)); - result[0] = static_cast(0xC0 | (0x1f & (cp >> 6))); - } - else if (cp <= 0xFFFF) - { - result.resize(3); - result[2] = static_cast(0x80 | (0x3f & cp)); - result[1] = 0x80 | static_cast((0x3f & (cp >> 6))); - result[0] = 0xE0 | static_cast((0xf & (cp >> 12))); - } - else if (cp <= 0x10FFFF) - { - result.resize(4); - result[3] = static_cast(0x80 | (0x3f & cp)); - result[2] = static_cast(0x80 | (0x3f & (cp >> 6))); - result[1] = static_cast(0x80 | (0x3f & (cp >> 12))); - result[0] = static_cast(0xF0 | (0x7 & (cp >> 18))); - } - - return result; -} - - -/// Returns true if ch is a control character (in range [0,32[). -static inline bool -isControlCharacter(char ch) -{ - return ch > 0 && ch <= 0x1F; -} - - -enum { - /// Constant that specify the size of the buffer that must be passed to uintToString. - uintToStringBufferSize = 3*sizeof(LargestUInt)+1 -}; - -// Defines a char buffer for use with uintToString(). -typedef char UIntToStringBuffer[uintToStringBufferSize]; - - -/** Converts an unsigned integer to string. - * @param value Unsigned interger to convert to string - * @param current Input/Output string buffer. - * Must have at least uintToStringBufferSize chars free. - */ -static inline void -uintToString( LargestUInt value, - char *¤t ) -{ - *--current = 0; - do - { - *--current = char(value % 10) + '0'; - value /= 10; - } - while ( value != 0 ); -} - -} // namespace Json { - -#endif // LIB_JSONCPP_JSON_TOOL_H_INCLUDED diff --git a/ext/jsoncpp/src/json_value.cpp b/ext/jsoncpp/src/json_value.cpp deleted file mode 100644 index ff98f635..00000000 --- a/ext/jsoncpp/src/json_value.cpp +++ /dev/null @@ -1,1829 +0,0 @@ -// Copyright 2007-2010 Baptiste Lepilleur -// Distributed under MIT license, or public domain if desired and -// recognized in your jurisdiction. -// See file LICENSE for detail or copy at http://jsoncpp.sourceforge.net/LICENSE - -#if !defined(JSON_IS_AMALGAMATION) -# include -# include -# ifndef JSON_USE_SIMPLE_INTERNAL_ALLOCATOR -# include "json_batchallocator.h" -# endif // #ifndef JSON_USE_SIMPLE_INTERNAL_ALLOCATOR -#endif // if !defined(JSON_IS_AMALGAMATION) -#include -#include -#include -#include -#include -#ifdef JSON_USE_CPPTL -# include -#endif -#include // size_t - -#define JSON_ASSERT_UNREACHABLE assert( false ) -#define JSON_ASSERT( condition ) assert( condition ); // @todo <= change this into an exception throw -#define JSON_FAIL_MESSAGE( message ) throw std::runtime_error( message ); -#define JSON_ASSERT_MESSAGE( condition, message ) if (!( condition )) JSON_FAIL_MESSAGE( message ) - -namespace Json { - -const Value Value::null; -const Int Value::minInt = Int( ~(UInt(-1)/2) ); -const Int Value::maxInt = Int( UInt(-1)/2 ); -const UInt Value::maxUInt = UInt(-1); -const Int64 Value::minInt64 = Int64( ~(UInt64(-1)/2) ); -const Int64 Value::maxInt64 = Int64( UInt64(-1)/2 ); -const UInt64 Value::maxUInt64 = UInt64(-1); -const LargestInt Value::minLargestInt = LargestInt( ~(LargestUInt(-1)/2) ); -const LargestInt Value::maxLargestInt = LargestInt( LargestUInt(-1)/2 ); -const LargestUInt Value::maxLargestUInt = LargestUInt(-1); - - -/// Unknown size marker -static const unsigned int unknown = (unsigned)-1; - - -/** Duplicates the specified string value. - * @param value Pointer to the string to duplicate. Must be zero-terminated if - * length is "unknown". - * @param length Length of the value. if equals to unknown, then it will be - * computed using strlen(value). - * @return Pointer on the duplicate instance of string. - */ -static inline char * -duplicateStringValue( const char *value, - unsigned int length = unknown ) -{ - if ( length == unknown ) - length = (unsigned int)strlen(value); - char *newString = static_cast( malloc( length + 1 ) ); - JSON_ASSERT_MESSAGE( newString != 0, "Failed to allocate string value buffer" ); - memcpy( newString, value, length ); - newString[length] = 0; - return newString; -} - - -/** Free the string duplicated by duplicateStringValue(). - */ -static inline void -releaseStringValue( char *value ) -{ - if ( value ) - free( value ); -} - -} // namespace Json - - -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// ValueInternals... -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -#if !defined(JSON_IS_AMALGAMATION) -# ifdef JSON_VALUE_USE_INTERNAL_MAP -# include "json_internalarray.inl" -# include "json_internalmap.inl" -# endif // JSON_VALUE_USE_INTERNAL_MAP - -# include "json_valueiterator.inl" -#endif // if !defined(JSON_IS_AMALGAMATION) - -namespace Json { - -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// class Value::CommentInfo -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// - - -Value::CommentInfo::CommentInfo() - : comment_( 0 ) -{ -} - -Value::CommentInfo::~CommentInfo() -{ - if ( comment_ ) - releaseStringValue( comment_ ); -} - - -void -Value::CommentInfo::setComment( const char *text ) -{ - if ( comment_ ) - releaseStringValue( comment_ ); - JSON_ASSERT( text != 0 ); - JSON_ASSERT_MESSAGE( text[0]=='\0' || text[0]=='/', "Comments must start with /"); - // It seems that /**/ style comments are acceptable as well. - comment_ = duplicateStringValue( text ); -} - - -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// class Value::CZString -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -# ifndef JSON_VALUE_USE_INTERNAL_MAP - -// Notes: index_ indicates if the string was allocated when -// a string is stored. - -Value::CZString::CZString( ArrayIndex index ) - : cstr_( 0 ) - , index_( index ) -{ -} - -Value::CZString::CZString( const char *cstr, DuplicationPolicy allocate ) - : cstr_( allocate == duplicate ? duplicateStringValue(cstr) - : cstr ) - , index_( allocate ) -{ -} - -Value::CZString::CZString( const CZString &other ) -: cstr_( other.index_ != noDuplication && other.cstr_ != 0 - ? duplicateStringValue( other.cstr_ ) - : other.cstr_ ) - , index_( other.cstr_ ? (other.index_ == noDuplication ? noDuplication : duplicate) - : other.index_ ) -{ -} - -Value::CZString::~CZString() -{ - if ( cstr_ && index_ == duplicate ) - releaseStringValue( const_cast( cstr_ ) ); -} - -void -Value::CZString::swap( CZString &other ) -{ - std::swap( cstr_, other.cstr_ ); - std::swap( index_, other.index_ ); -} - -Value::CZString & -Value::CZString::operator =( const CZString &other ) -{ - CZString temp( other ); - swap( temp ); - return *this; -} - -bool -Value::CZString::operator<( const CZString &other ) const -{ - if ( cstr_ ) - return strcmp( cstr_, other.cstr_ ) < 0; - return index_ < other.index_; -} - -bool -Value::CZString::operator==( const CZString &other ) const -{ - if ( cstr_ ) - return strcmp( cstr_, other.cstr_ ) == 0; - return index_ == other.index_; -} - - -ArrayIndex -Value::CZString::index() const -{ - return index_; -} - - -const char * -Value::CZString::c_str() const -{ - return cstr_; -} - -bool -Value::CZString::isStaticString() const -{ - return index_ == noDuplication; -} - -#endif // ifndef JSON_VALUE_USE_INTERNAL_MAP - - -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// class Value::Value -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// - -/*! \internal Default constructor initialization must be equivalent to: - * memset( this, 0, sizeof(Value) ) - * This optimization is used in ValueInternalMap fast allocator. - */ -Value::Value( ValueType type ) - : type_( type ) - , allocated_( 0 ) - , comments_( 0 ) -# ifdef JSON_VALUE_USE_INTERNAL_MAP - , itemIsUsed_( 0 ) -#endif -{ - switch ( type ) - { - case nullValue: - break; - case intValue: - case uintValue: - value_.int_ = 0; - break; - case realValue: - value_.real_ = 0.0; - break; - case stringValue: - value_.string_ = 0; - break; -#ifndef JSON_VALUE_USE_INTERNAL_MAP - case arrayValue: - case objectValue: - value_.map_ = new ObjectValues(); - break; -#else - case arrayValue: - value_.array_ = arrayAllocator()->newArray(); - break; - case objectValue: - value_.map_ = mapAllocator()->newMap(); - break; -#endif - case booleanValue: - value_.bool_ = false; - break; - default: - JSON_ASSERT_UNREACHABLE; - } -} - - -#if defined(JSON_HAS_INT64) -Value::Value( UInt value ) - : type_( uintValue ) - , comments_( 0 ) -# ifdef JSON_VALUE_USE_INTERNAL_MAP - , itemIsUsed_( 0 ) -#endif -{ - value_.uint_ = value; -} - -Value::Value( Int value ) - : type_( intValue ) - , comments_( 0 ) -# ifdef JSON_VALUE_USE_INTERNAL_MAP - , itemIsUsed_( 0 ) -#endif -{ - value_.int_ = value; -} - -#endif // if defined(JSON_HAS_INT64) - - -Value::Value( Int64 value ) - : type_( intValue ) - , comments_( 0 ) -# ifdef JSON_VALUE_USE_INTERNAL_MAP - , itemIsUsed_( 0 ) -#endif -{ - value_.int_ = value; -} - - -Value::Value( UInt64 value ) - : type_( uintValue ) - , comments_( 0 ) -# ifdef JSON_VALUE_USE_INTERNAL_MAP - , itemIsUsed_( 0 ) -#endif -{ - value_.uint_ = value; -} - -Value::Value( double value ) - : type_( realValue ) - , comments_( 0 ) -# ifdef JSON_VALUE_USE_INTERNAL_MAP - , itemIsUsed_( 0 ) -#endif -{ - value_.real_ = value; -} - -Value::Value( const char *value ) - : type_( stringValue ) - , allocated_( true ) - , comments_( 0 ) -# ifdef JSON_VALUE_USE_INTERNAL_MAP - , itemIsUsed_( 0 ) -#endif -{ - value_.string_ = duplicateStringValue( value ); -} - - -Value::Value( const char *beginValue, - const char *endValue ) - : type_( stringValue ) - , allocated_( true ) - , comments_( 0 ) -# ifdef JSON_VALUE_USE_INTERNAL_MAP - , itemIsUsed_( 0 ) -#endif -{ - value_.string_ = duplicateStringValue( beginValue, - (unsigned int)(endValue - beginValue) ); -} - - -Value::Value( const std::string &value ) - : type_( stringValue ) - , allocated_( true ) - , comments_( 0 ) -# ifdef JSON_VALUE_USE_INTERNAL_MAP - , itemIsUsed_( 0 ) -#endif -{ - value_.string_ = duplicateStringValue( value.c_str(), - (unsigned int)value.length() ); - -} - -Value::Value( const StaticString &value ) - : type_( stringValue ) - , allocated_( false ) - , comments_( 0 ) -# ifdef JSON_VALUE_USE_INTERNAL_MAP - , itemIsUsed_( 0 ) -#endif -{ - value_.string_ = const_cast( value.c_str() ); -} - - -# ifdef JSON_USE_CPPTL -Value::Value( const CppTL::ConstString &value ) - : type_( stringValue ) - , allocated_( true ) - , comments_( 0 ) -# ifdef JSON_VALUE_USE_INTERNAL_MAP - , itemIsUsed_( 0 ) -#endif -{ - value_.string_ = duplicateStringValue( value, value.length() ); -} -# endif - -Value::Value( bool value ) - : type_( booleanValue ) - , comments_( 0 ) -# ifdef JSON_VALUE_USE_INTERNAL_MAP - , itemIsUsed_( 0 ) -#endif -{ - value_.bool_ = value; -} - - -Value::Value( const Value &other ) - : type_( other.type_ ) - , comments_( 0 ) -# ifdef JSON_VALUE_USE_INTERNAL_MAP - , itemIsUsed_( 0 ) -#endif -{ - switch ( type_ ) - { - case nullValue: - case intValue: - case uintValue: - case realValue: - case booleanValue: - value_ = other.value_; - break; - case stringValue: - if ( other.value_.string_ ) - { - value_.string_ = duplicateStringValue( other.value_.string_ ); - allocated_ = true; - } - else - value_.string_ = 0; - break; -#ifndef JSON_VALUE_USE_INTERNAL_MAP - case arrayValue: - case objectValue: - value_.map_ = new ObjectValues( *other.value_.map_ ); - break; -#else - case arrayValue: - value_.array_ = arrayAllocator()->newArrayCopy( *other.value_.array_ ); - break; - case objectValue: - value_.map_ = mapAllocator()->newMapCopy( *other.value_.map_ ); - break; -#endif - default: - JSON_ASSERT_UNREACHABLE; - } - if ( other.comments_ ) - { - comments_ = new CommentInfo[numberOfCommentPlacement]; - for ( int comment =0; comment < numberOfCommentPlacement; ++comment ) - { - const CommentInfo &otherComment = other.comments_[comment]; - if ( otherComment.comment_ ) - comments_[comment].setComment( otherComment.comment_ ); - } - } -} - - -Value::~Value() -{ - switch ( type_ ) - { - case nullValue: - case intValue: - case uintValue: - case realValue: - case booleanValue: - break; - case stringValue: - if ( allocated_ ) - releaseStringValue( value_.string_ ); - break; -#ifndef JSON_VALUE_USE_INTERNAL_MAP - case arrayValue: - case objectValue: - delete value_.map_; - break; -#else - case arrayValue: - arrayAllocator()->destructArray( value_.array_ ); - break; - case objectValue: - mapAllocator()->destructMap( value_.map_ ); - break; -#endif - default: - JSON_ASSERT_UNREACHABLE; - } - - if ( comments_ ) - delete[] comments_; -} - -Value & -Value::operator=( const Value &other ) -{ - Value temp( other ); - swap( temp ); - return *this; -} - -void -Value::swap( Value &other ) -{ - ValueType temp = type_; - type_ = other.type_; - other.type_ = temp; - std::swap( value_, other.value_ ); - int temp2 = allocated_; - allocated_ = other.allocated_; - other.allocated_ = temp2; -} - -ValueType -Value::type() const -{ - return type_; -} - - -int -Value::compare( const Value &other ) const -{ - if ( *this < other ) - return -1; - if ( *this > other ) - return 1; - return 0; -} - - -bool -Value::operator <( const Value &other ) const -{ - int typeDelta = type_ - other.type_; - if ( typeDelta ) - return typeDelta < 0 ? true : false; - switch ( type_ ) - { - case nullValue: - return false; - case intValue: - return value_.int_ < other.value_.int_; - case uintValue: - return value_.uint_ < other.value_.uint_; - case realValue: - return value_.real_ < other.value_.real_; - case booleanValue: - return value_.bool_ < other.value_.bool_; - case stringValue: - return ( value_.string_ == 0 && other.value_.string_ ) - || ( other.value_.string_ - && value_.string_ - && strcmp( value_.string_, other.value_.string_ ) < 0 ); -#ifndef JSON_VALUE_USE_INTERNAL_MAP - case arrayValue: - case objectValue: - { - int delta = int( value_.map_->size() - other.value_.map_->size() ); - if ( delta ) - return delta < 0; - return (*value_.map_) < (*other.value_.map_); - } -#else - case arrayValue: - return value_.array_->compare( *(other.value_.array_) ) < 0; - case objectValue: - return value_.map_->compare( *(other.value_.map_) ) < 0; -#endif - default: - JSON_ASSERT_UNREACHABLE; - } - return false; // unreachable -} - -bool -Value::operator <=( const Value &other ) const -{ - return !(other < *this); -} - -bool -Value::operator >=( const Value &other ) const -{ - return !(*this < other); -} - -bool -Value::operator >( const Value &other ) const -{ - return other < *this; -} - -bool -Value::operator ==( const Value &other ) const -{ - //if ( type_ != other.type_ ) - // GCC 2.95.3 says: - // attempt to take address of bit-field structure member `Json::Value::type_' - // Beats me, but a temp solves the problem. - int temp = other.type_; - if ( type_ != temp ) - return false; - switch ( type_ ) - { - case nullValue: - return true; - case intValue: - return value_.int_ == other.value_.int_; - case uintValue: - return value_.uint_ == other.value_.uint_; - case realValue: - return value_.real_ == other.value_.real_; - case booleanValue: - return value_.bool_ == other.value_.bool_; - case stringValue: - return ( value_.string_ == other.value_.string_ ) - || ( other.value_.string_ - && value_.string_ - && strcmp( value_.string_, other.value_.string_ ) == 0 ); -#ifndef JSON_VALUE_USE_INTERNAL_MAP - case arrayValue: - case objectValue: - return value_.map_->size() == other.value_.map_->size() - && (*value_.map_) == (*other.value_.map_); -#else - case arrayValue: - return value_.array_->compare( *(other.value_.array_) ) == 0; - case objectValue: - return value_.map_->compare( *(other.value_.map_) ) == 0; -#endif - default: - JSON_ASSERT_UNREACHABLE; - } - return false; // unreachable -} - -bool -Value::operator !=( const Value &other ) const -{ - return !( *this == other ); -} - -const char * -Value::asCString() const -{ - JSON_ASSERT( type_ == stringValue ); - return value_.string_; -} - - -std::string -Value::asString() const -{ - switch ( type_ ) - { - case nullValue: - return ""; - case stringValue: - return value_.string_ ? value_.string_ : ""; - case booleanValue: - return value_.bool_ ? "true" : "false"; - case intValue: - case uintValue: - case realValue: - case arrayValue: - case objectValue: - JSON_FAIL_MESSAGE( "Type is not convertible to string" ); - default: - JSON_ASSERT_UNREACHABLE; - } - return ""; // unreachable -} - -# ifdef JSON_USE_CPPTL -CppTL::ConstString -Value::asConstString() const -{ - return CppTL::ConstString( asString().c_str() ); -} -# endif - - -Value::Int -Value::asInt() const -{ - switch ( type_ ) - { - case nullValue: - return 0; - case intValue: - JSON_ASSERT_MESSAGE( value_.int_ >= minInt && value_.int_ <= maxInt, "unsigned integer out of signed int range" ); - return Int(value_.int_); - case uintValue: - JSON_ASSERT_MESSAGE( value_.uint_ <= UInt(maxInt), "unsigned integer out of signed int range" ); - return Int(value_.uint_); - case realValue: - JSON_ASSERT_MESSAGE( value_.real_ >= minInt && value_.real_ <= maxInt, "Real out of signed integer range" ); - return Int( value_.real_ ); - case booleanValue: - return value_.bool_ ? 1 : 0; - case stringValue: - case arrayValue: - case objectValue: - JSON_FAIL_MESSAGE( "Type is not convertible to int" ); - default: - JSON_ASSERT_UNREACHABLE; - } - return 0; // unreachable; -} - - -Value::UInt -Value::asUInt() const -{ - switch ( type_ ) - { - case nullValue: - return 0; - case intValue: - JSON_ASSERT_MESSAGE( value_.int_ >= 0, "Negative integer can not be converted to unsigned integer" ); - JSON_ASSERT_MESSAGE( value_.int_ <= maxUInt, "signed integer out of UInt range" ); - return UInt(value_.int_); - case uintValue: - JSON_ASSERT_MESSAGE( value_.uint_ <= maxUInt, "unsigned integer out of UInt range" ); - return UInt(value_.uint_); - case realValue: - JSON_ASSERT_MESSAGE( value_.real_ >= 0 && value_.real_ <= maxUInt, "Real out of unsigned integer range" ); - return UInt( value_.real_ ); - case booleanValue: - return value_.bool_ ? 1 : 0; - case stringValue: - case arrayValue: - case objectValue: - JSON_FAIL_MESSAGE( "Type is not convertible to uint" ); - default: - JSON_ASSERT_UNREACHABLE; - } - return 0; // unreachable; -} - - -# if defined(JSON_HAS_INT64) - -Value::Int64 -Value::asInt64() const -{ - switch ( type_ ) - { - case nullValue: - return 0; - case intValue: - return value_.int_; - case uintValue: - JSON_ASSERT_MESSAGE( value_.uint_ <= UInt64(maxInt64), "unsigned integer out of Int64 range" ); - return value_.uint_; - case realValue: - JSON_ASSERT_MESSAGE( value_.real_ >= minInt64 && value_.real_ <= maxInt64, "Real out of Int64 range" ); - return Int( value_.real_ ); - case booleanValue: - return value_.bool_ ? 1 : 0; - case stringValue: - case arrayValue: - case objectValue: - JSON_FAIL_MESSAGE( "Type is not convertible to Int64" ); - default: - JSON_ASSERT_UNREACHABLE; - } - return 0; // unreachable; -} - - -Value::UInt64 -Value::asUInt64() const -{ - switch ( type_ ) - { - case nullValue: - return 0; - case intValue: - JSON_ASSERT_MESSAGE( value_.int_ >= 0, "Negative integer can not be converted to UInt64" ); - return value_.int_; - case uintValue: - return value_.uint_; - case realValue: - JSON_ASSERT_MESSAGE( value_.real_ >= 0 && value_.real_ <= maxUInt64, "Real out of UInt64 range" ); - return UInt( value_.real_ ); - case booleanValue: - return value_.bool_ ? 1 : 0; - case stringValue: - case arrayValue: - case objectValue: - JSON_FAIL_MESSAGE( "Type is not convertible to UInt64" ); - default: - JSON_ASSERT_UNREACHABLE; - } - return 0; // unreachable; -} -# endif // if defined(JSON_HAS_INT64) - - -LargestInt -Value::asLargestInt() const -{ -#if defined(JSON_NO_INT64) - return asInt(); -#else - return asInt64(); -#endif -} - - -LargestUInt -Value::asLargestUInt() const -{ -#if defined(JSON_NO_INT64) - return asUInt(); -#else - return asUInt64(); -#endif -} - - -double -Value::asDouble() const -{ - switch ( type_ ) - { - case nullValue: - return 0.0; - case intValue: - return static_cast( value_.int_ ); - case uintValue: -#if !defined(JSON_USE_INT64_DOUBLE_CONVERSION) - return static_cast( value_.uint_ ); -#else // if !defined(JSON_USE_INT64_DOUBLE_CONVERSION) - return static_cast( Int(value_.uint_/2) ) * 2 + Int(value_.uint_ & 1); -#endif // if !defined(JSON_USE_INT64_DOUBLE_CONVERSION) - case realValue: - return value_.real_; - case booleanValue: - return value_.bool_ ? 1.0 : 0.0; - case stringValue: - case arrayValue: - case objectValue: - JSON_FAIL_MESSAGE( "Type is not convertible to double" ); - default: - JSON_ASSERT_UNREACHABLE; - } - return 0; // unreachable; -} - -float -Value::asFloat() const -{ - switch ( type_ ) - { - case nullValue: - return 0.0f; - case intValue: - return static_cast( value_.int_ ); - case uintValue: -#if !defined(JSON_USE_INT64_DOUBLE_CONVERSION) - return static_cast( value_.uint_ ); -#else // if !defined(JSON_USE_INT64_DOUBLE_CONVERSION) - return static_cast( Int(value_.uint_/2) ) * 2 + Int(value_.uint_ & 1); -#endif // if !defined(JSON_USE_INT64_DOUBLE_CONVERSION) - case realValue: - return static_cast( value_.real_ ); - case booleanValue: - return value_.bool_ ? 1.0f : 0.0f; - case stringValue: - case arrayValue: - case objectValue: - JSON_FAIL_MESSAGE( "Type is not convertible to float" ); - default: - JSON_ASSERT_UNREACHABLE; - } - return 0.0f; // unreachable; -} - -bool -Value::asBool() const -{ - switch ( type_ ) - { - case nullValue: - return false; - case intValue: - case uintValue: - return value_.int_ != 0; - case realValue: - return value_.real_ != 0.0; - case booleanValue: - return value_.bool_; - case stringValue: - return value_.string_ && value_.string_[0] != 0; - case arrayValue: - case objectValue: - return value_.map_->size() != 0; - default: - JSON_ASSERT_UNREACHABLE; - } - return false; // unreachable; -} - - -bool -Value::isConvertibleTo( ValueType other ) const -{ - switch ( type_ ) - { - case nullValue: - return true; - case intValue: - return ( other == nullValue && value_.int_ == 0 ) - || other == intValue - || ( other == uintValue && value_.int_ >= 0 ) - || other == realValue - || other == stringValue - || other == booleanValue; - case uintValue: - return ( other == nullValue && value_.uint_ == 0 ) - || ( other == intValue && value_.uint_ <= (unsigned)maxInt ) - || other == uintValue - || other == realValue - || other == stringValue - || other == booleanValue; - case realValue: - return ( other == nullValue && value_.real_ == 0.0 ) - || ( other == intValue && value_.real_ >= minInt && value_.real_ <= maxInt ) - || ( other == uintValue && value_.real_ >= 0 && value_.real_ <= maxUInt ) - || other == realValue - || other == stringValue - || other == booleanValue; - case booleanValue: - return ( other == nullValue && value_.bool_ == false ) - || other == intValue - || other == uintValue - || other == realValue - || other == stringValue - || other == booleanValue; - case stringValue: - return other == stringValue - || ( other == nullValue && (!value_.string_ || value_.string_[0] == 0) ); - case arrayValue: - return other == arrayValue - || ( other == nullValue && value_.map_->size() == 0 ); - case objectValue: - return other == objectValue - || ( other == nullValue && value_.map_->size() == 0 ); - default: - JSON_ASSERT_UNREACHABLE; - } - return false; // unreachable; -} - - -/// Number of values in array or object -ArrayIndex -Value::size() const -{ - switch ( type_ ) - { - case nullValue: - case intValue: - case uintValue: - case realValue: - case booleanValue: - case stringValue: - return 0; -#ifndef JSON_VALUE_USE_INTERNAL_MAP - case arrayValue: // size of the array is highest index + 1 - if ( !value_.map_->empty() ) - { - ObjectValues::const_iterator itLast = value_.map_->end(); - --itLast; - return (*itLast).first.index()+1; - } - return 0; - case objectValue: - return ArrayIndex( value_.map_->size() ); -#else - case arrayValue: - return Int( value_.array_->size() ); - case objectValue: - return Int( value_.map_->size() ); -#endif - default: - JSON_ASSERT_UNREACHABLE; - } - return 0; // unreachable; -} - - -bool -Value::empty() const -{ - if ( isNull() || isArray() || isObject() ) - return size() == 0u; - else - return false; -} - - -bool -Value::operator!() const -{ - return isNull(); -} - - -void -Value::clear() -{ - JSON_ASSERT( type_ == nullValue || type_ == arrayValue || type_ == objectValue ); - - switch ( type_ ) - { -#ifndef JSON_VALUE_USE_INTERNAL_MAP - case arrayValue: - case objectValue: - value_.map_->clear(); - break; -#else - case arrayValue: - value_.array_->clear(); - break; - case objectValue: - value_.map_->clear(); - break; -#endif - default: - break; - } -} - -void -Value::resize( ArrayIndex newSize ) -{ - JSON_ASSERT( type_ == nullValue || type_ == arrayValue ); - if ( type_ == nullValue ) - *this = Value( arrayValue ); -#ifndef JSON_VALUE_USE_INTERNAL_MAP - ArrayIndex oldSize = size(); - if ( newSize == 0 ) - clear(); - else if ( newSize > oldSize ) - (*this)[ newSize - 1 ]; - else - { - for ( ArrayIndex index = newSize; index < oldSize; ++index ) - { - value_.map_->erase( index ); - } - assert( size() == newSize ); - } -#else - value_.array_->resize( newSize ); -#endif -} - - -Value & -Value::operator[]( ArrayIndex index ) -{ - JSON_ASSERT( type_ == nullValue || type_ == arrayValue ); - if ( type_ == nullValue ) - *this = Value( arrayValue ); -#ifndef JSON_VALUE_USE_INTERNAL_MAP - CZString key( index ); - ObjectValues::iterator it = value_.map_->lower_bound( key ); - if ( it != value_.map_->end() && (*it).first == key ) - return (*it).second; - - ObjectValues::value_type defaultValue( key, null ); - it = value_.map_->insert( it, defaultValue ); - return (*it).second; -#else - return value_.array_->resolveReference( index ); -#endif -} - - -Value & -Value::operator[]( int index ) -{ - JSON_ASSERT( index >= 0 ); - return (*this)[ ArrayIndex(index) ]; -} - - -const Value & -Value::operator[]( ArrayIndex index ) const -{ - JSON_ASSERT( type_ == nullValue || type_ == arrayValue ); - if ( type_ == nullValue ) - return null; -#ifndef JSON_VALUE_USE_INTERNAL_MAP - CZString key( index ); - ObjectValues::const_iterator it = value_.map_->find( key ); - if ( it == value_.map_->end() ) - return null; - return (*it).second; -#else - Value *value = value_.array_->find( index ); - return value ? *value : null; -#endif -} - - -const Value & -Value::operator[]( int index ) const -{ - JSON_ASSERT( index >= 0 ); - return (*this)[ ArrayIndex(index) ]; -} - - -Value & -Value::operator[]( const char *key ) -{ - return resolveReference( key, false ); -} - - -Value & -Value::resolveReference( const char *key, - bool isStatic ) -{ - JSON_ASSERT( type_ == nullValue || type_ == objectValue ); - if ( type_ == nullValue ) - *this = Value( objectValue ); -#ifndef JSON_VALUE_USE_INTERNAL_MAP - CZString actualKey( key, isStatic ? CZString::noDuplication - : CZString::duplicateOnCopy ); - ObjectValues::iterator it = value_.map_->lower_bound( actualKey ); - if ( it != value_.map_->end() && (*it).first == actualKey ) - return (*it).second; - - ObjectValues::value_type defaultValue( actualKey, null ); - it = value_.map_->insert( it, defaultValue ); - Value &value = (*it).second; - return value; -#else - return value_.map_->resolveReference( key, isStatic ); -#endif -} - - -Value -Value::get( ArrayIndex index, - const Value &defaultValue ) const -{ - const Value *value = &((*this)[index]); - return value == &null ? defaultValue : *value; -} - - -bool -Value::isValidIndex( ArrayIndex index ) const -{ - return index < size(); -} - - - -const Value & -Value::operator[]( const char *key ) const -{ - JSON_ASSERT( type_ == nullValue || type_ == objectValue ); - if ( type_ == nullValue ) - return null; -#ifndef JSON_VALUE_USE_INTERNAL_MAP - CZString actualKey( key, CZString::noDuplication ); - ObjectValues::const_iterator it = value_.map_->find( actualKey ); - if ( it == value_.map_->end() ) - return null; - return (*it).second; -#else - const Value *value = value_.map_->find( key ); - return value ? *value : null; -#endif -} - - -Value & -Value::operator[]( const std::string &key ) -{ - return (*this)[ key.c_str() ]; -} - - -const Value & -Value::operator[]( const std::string &key ) const -{ - return (*this)[ key.c_str() ]; -} - -Value & -Value::operator[]( const StaticString &key ) -{ - return resolveReference( key, true ); -} - - -# ifdef JSON_USE_CPPTL -Value & -Value::operator[]( const CppTL::ConstString &key ) -{ - return (*this)[ key.c_str() ]; -} - - -const Value & -Value::operator[]( const CppTL::ConstString &key ) const -{ - return (*this)[ key.c_str() ]; -} -# endif - - -Value & -Value::append( const Value &value ) -{ - return (*this)[size()] = value; -} - - -Value -Value::get( const char *key, - const Value &defaultValue ) const -{ - const Value *value = &((*this)[key]); - return value == &null ? defaultValue : *value; -} - - -Value -Value::get( const std::string &key, - const Value &defaultValue ) const -{ - return get( key.c_str(), defaultValue ); -} - -Value -Value::removeMember( const char* key ) -{ - JSON_ASSERT( type_ == nullValue || type_ == objectValue ); - if ( type_ == nullValue ) - return null; -#ifndef JSON_VALUE_USE_INTERNAL_MAP - CZString actualKey( key, CZString::noDuplication ); - ObjectValues::iterator it = value_.map_->find( actualKey ); - if ( it == value_.map_->end() ) - return null; - Value old(it->second); - value_.map_->erase(it); - return old; -#else - Value *value = value_.map_->find( key ); - if (value){ - Value old(*value); - value_.map_.remove( key ); - return old; - } else { - return null; - } -#endif -} - -Value -Value::removeMember( const std::string &key ) -{ - return removeMember( key.c_str() ); -} - -# ifdef JSON_USE_CPPTL -Value -Value::get( const CppTL::ConstString &key, - const Value &defaultValue ) const -{ - return get( key.c_str(), defaultValue ); -} -# endif - -bool -Value::isMember( const char *key ) const -{ - const Value *value = &((*this)[key]); - return value != &null; -} - - -bool -Value::isMember( const std::string &key ) const -{ - return isMember( key.c_str() ); -} - - -# ifdef JSON_USE_CPPTL -bool -Value::isMember( const CppTL::ConstString &key ) const -{ - return isMember( key.c_str() ); -} -#endif - -Value::Members -Value::getMemberNames() const -{ - JSON_ASSERT( type_ == nullValue || type_ == objectValue ); - if ( type_ == nullValue ) - return Value::Members(); - Members members; - members.reserve( value_.map_->size() ); -#ifndef JSON_VALUE_USE_INTERNAL_MAP - ObjectValues::const_iterator it = value_.map_->begin(); - ObjectValues::const_iterator itEnd = value_.map_->end(); - for ( ; it != itEnd; ++it ) - members.push_back( std::string( (*it).first.c_str() ) ); -#else - ValueInternalMap::IteratorState it; - ValueInternalMap::IteratorState itEnd; - value_.map_->makeBeginIterator( it ); - value_.map_->makeEndIterator( itEnd ); - for ( ; !ValueInternalMap::equals( it, itEnd ); ValueInternalMap::increment(it) ) - members.push_back( std::string( ValueInternalMap::key( it ) ) ); -#endif - return members; -} -// -//# ifdef JSON_USE_CPPTL -//EnumMemberNames -//Value::enumMemberNames() const -//{ -// if ( type_ == objectValue ) -// { -// return CppTL::Enum::any( CppTL::Enum::transform( -// CppTL::Enum::keys( *(value_.map_), CppTL::Type() ), -// MemberNamesTransform() ) ); -// } -// return EnumMemberNames(); -//} -// -// -//EnumValues -//Value::enumValues() const -//{ -// if ( type_ == objectValue || type_ == arrayValue ) -// return CppTL::Enum::anyValues( *(value_.map_), -// CppTL::Type() ); -// return EnumValues(); -//} -// -//# endif - - -bool -Value::isNull() const -{ - return type_ == nullValue; -} - - -bool -Value::isBool() const -{ - return type_ == booleanValue; -} - - -bool -Value::isInt() const -{ - return type_ == intValue; -} - - -bool -Value::isUInt() const -{ - return type_ == uintValue; -} - - -bool -Value::isIntegral() const -{ - return type_ == intValue - || type_ == uintValue - || type_ == booleanValue; -} - - -bool -Value::isDouble() const -{ - return type_ == realValue; -} - - -bool -Value::isNumeric() const -{ - return isIntegral() || isDouble(); -} - - -bool -Value::isString() const -{ - return type_ == stringValue; -} - - -bool -Value::isArray() const -{ - return type_ == nullValue || type_ == arrayValue; -} - - -bool -Value::isObject() const -{ - return type_ == nullValue || type_ == objectValue; -} - - -void -Value::setComment( const char *comment, - CommentPlacement placement ) -{ - if ( !comments_ ) - comments_ = new CommentInfo[numberOfCommentPlacement]; - comments_[placement].setComment( comment ); -} - - -void -Value::setComment( const std::string &comment, - CommentPlacement placement ) -{ - setComment( comment.c_str(), placement ); -} - - -bool -Value::hasComment( CommentPlacement placement ) const -{ - return comments_ != 0 && comments_[placement].comment_ != 0; -} - -std::string -Value::getComment( CommentPlacement placement ) const -{ - if ( hasComment(placement) ) - return comments_[placement].comment_; - return ""; -} - - -std::string -Value::toStyledString() const -{ - StyledWriter writer; - return writer.write( *this ); -} - - -Value::const_iterator -Value::begin() const -{ - switch ( type_ ) - { -#ifdef JSON_VALUE_USE_INTERNAL_MAP - case arrayValue: - if ( value_.array_ ) - { - ValueInternalArray::IteratorState it; - value_.array_->makeBeginIterator( it ); - return const_iterator( it ); - } - break; - case objectValue: - if ( value_.map_ ) - { - ValueInternalMap::IteratorState it; - value_.map_->makeBeginIterator( it ); - return const_iterator( it ); - } - break; -#else - case arrayValue: - case objectValue: - if ( value_.map_ ) - return const_iterator( value_.map_->begin() ); - break; -#endif - default: - break; - } - return const_iterator(); -} - -Value::const_iterator -Value::end() const -{ - switch ( type_ ) - { -#ifdef JSON_VALUE_USE_INTERNAL_MAP - case arrayValue: - if ( value_.array_ ) - { - ValueInternalArray::IteratorState it; - value_.array_->makeEndIterator( it ); - return const_iterator( it ); - } - break; - case objectValue: - if ( value_.map_ ) - { - ValueInternalMap::IteratorState it; - value_.map_->makeEndIterator( it ); - return const_iterator( it ); - } - break; -#else - case arrayValue: - case objectValue: - if ( value_.map_ ) - return const_iterator( value_.map_->end() ); - break; -#endif - default: - break; - } - return const_iterator(); -} - - -Value::iterator -Value::begin() -{ - switch ( type_ ) - { -#ifdef JSON_VALUE_USE_INTERNAL_MAP - case arrayValue: - if ( value_.array_ ) - { - ValueInternalArray::IteratorState it; - value_.array_->makeBeginIterator( it ); - return iterator( it ); - } - break; - case objectValue: - if ( value_.map_ ) - { - ValueInternalMap::IteratorState it; - value_.map_->makeBeginIterator( it ); - return iterator( it ); - } - break; -#else - case arrayValue: - case objectValue: - if ( value_.map_ ) - return iterator( value_.map_->begin() ); - break; -#endif - default: - break; - } - return iterator(); -} - -Value::iterator -Value::end() -{ - switch ( type_ ) - { -#ifdef JSON_VALUE_USE_INTERNAL_MAP - case arrayValue: - if ( value_.array_ ) - { - ValueInternalArray::IteratorState it; - value_.array_->makeEndIterator( it ); - return iterator( it ); - } - break; - case objectValue: - if ( value_.map_ ) - { - ValueInternalMap::IteratorState it; - value_.map_->makeEndIterator( it ); - return iterator( it ); - } - break; -#else - case arrayValue: - case objectValue: - if ( value_.map_ ) - return iterator( value_.map_->end() ); - break; -#endif - default: - break; - } - return iterator(); -} - - -// class PathArgument -// ////////////////////////////////////////////////////////////////// - -PathArgument::PathArgument() - : kind_( kindNone ) -{ -} - - -PathArgument::PathArgument( ArrayIndex index ) - : index_( index ) - , kind_( kindIndex ) -{ -} - - -PathArgument::PathArgument( const char *key ) - : key_( key ) - , kind_( kindKey ) -{ -} - - -PathArgument::PathArgument( const std::string &key ) - : key_( key.c_str() ) - , kind_( kindKey ) -{ -} - -// class Path -// ////////////////////////////////////////////////////////////////// - -Path::Path( const std::string &path, - const PathArgument &a1, - const PathArgument &a2, - const PathArgument &a3, - const PathArgument &a4, - const PathArgument &a5 ) -{ - InArgs in; - in.push_back( &a1 ); - in.push_back( &a2 ); - in.push_back( &a3 ); - in.push_back( &a4 ); - in.push_back( &a5 ); - makePath( path, in ); -} - - -void -Path::makePath( const std::string &path, - const InArgs &in ) -{ - const char *current = path.c_str(); - const char *end = current + path.length(); - InArgs::const_iterator itInArg = in.begin(); - while ( current != end ) - { - if ( *current == '[' ) - { - ++current; - if ( *current == '%' ) - addPathInArg( path, in, itInArg, PathArgument::kindIndex ); - else - { - ArrayIndex index = 0; - for ( ; current != end && *current >= '0' && *current <= '9'; ++current ) - index = index * 10 + ArrayIndex(*current - '0'); - args_.push_back( index ); - } - if ( current == end || *current++ != ']' ) - invalidPath( path, int(current - path.c_str()) ); - } - else if ( *current == '%' ) - { - addPathInArg( path, in, itInArg, PathArgument::kindKey ); - ++current; - } - else if ( *current == '.' ) - { - ++current; - } - else - { - const char *beginName = current; - while ( current != end && !strchr( "[.", *current ) ) - ++current; - args_.push_back( std::string( beginName, current ) ); - } - } -} - - -void -Path::addPathInArg( const std::string &path, - const InArgs &in, - InArgs::const_iterator &itInArg, - PathArgument::Kind kind ) -{ - if ( itInArg == in.end() ) - { - // Error: missing argument %d - } - else if ( (*itInArg)->kind_ != kind ) - { - // Error: bad argument type - } - else - { - args_.push_back( **itInArg ); - } -} - - -void -Path::invalidPath( const std::string &path, - int location ) -{ - // Error: invalid path. -} - - -const Value & -Path::resolve( const Value &root ) const -{ - const Value *node = &root; - for ( Args::const_iterator it = args_.begin(); it != args_.end(); ++it ) - { - const PathArgument &arg = *it; - if ( arg.kind_ == PathArgument::kindIndex ) - { - if ( !node->isArray() || node->isValidIndex( arg.index_ ) ) - { - // Error: unable to resolve path (array value expected at position... - } - node = &((*node)[arg.index_]); - } - else if ( arg.kind_ == PathArgument::kindKey ) - { - if ( !node->isObject() ) - { - // Error: unable to resolve path (object value expected at position...) - } - node = &((*node)[arg.key_]); - if ( node == &Value::null ) - { - // Error: unable to resolve path (object has no member named '' at position...) - } - } - } - return *node; -} - - -Value -Path::resolve( const Value &root, - const Value &defaultValue ) const -{ - const Value *node = &root; - for ( Args::const_iterator it = args_.begin(); it != args_.end(); ++it ) - { - const PathArgument &arg = *it; - if ( arg.kind_ == PathArgument::kindIndex ) - { - if ( !node->isArray() || node->isValidIndex( arg.index_ ) ) - return defaultValue; - node = &((*node)[arg.index_]); - } - else if ( arg.kind_ == PathArgument::kindKey ) - { - if ( !node->isObject() ) - return defaultValue; - node = &((*node)[arg.key_]); - if ( node == &Value::null ) - return defaultValue; - } - } - return *node; -} - - -Value & -Path::make( Value &root ) const -{ - Value *node = &root; - for ( Args::const_iterator it = args_.begin(); it != args_.end(); ++it ) - { - const PathArgument &arg = *it; - if ( arg.kind_ == PathArgument::kindIndex ) - { - if ( !node->isArray() ) - { - // Error: node is not an array at position ... - } - node = &((*node)[arg.index_]); - } - else if ( arg.kind_ == PathArgument::kindKey ) - { - if ( !node->isObject() ) - { - // Error: node is not an object at position... - } - node = &((*node)[arg.key_]); - } - } - return *node; -} - - -} // namespace Json diff --git a/ext/jsoncpp/src/json_valueiterator.inl b/ext/jsoncpp/src/json_valueiterator.inl deleted file mode 100644 index 7457ca38..00000000 --- a/ext/jsoncpp/src/json_valueiterator.inl +++ /dev/null @@ -1,299 +0,0 @@ -// Copyright 2007-2010 Baptiste Lepilleur -// Distributed under MIT license, or public domain if desired and -// recognized in your jurisdiction. -// See file LICENSE for detail or copy at http://jsoncpp.sourceforge.net/LICENSE - -// included by json_value.cpp - -namespace Json { - -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// class ValueIteratorBase -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// - -ValueIteratorBase::ValueIteratorBase() -#ifndef JSON_VALUE_USE_INTERNAL_MAP - : current_() - , isNull_( true ) -{ -} -#else - : isArray_( true ) - , isNull_( true ) -{ - iterator_.array_ = ValueInternalArray::IteratorState(); -} -#endif - - -#ifndef JSON_VALUE_USE_INTERNAL_MAP -ValueIteratorBase::ValueIteratorBase( const Value::ObjectValues::iterator ¤t ) - : current_( current ) - , isNull_( false ) -{ -} -#else -ValueIteratorBase::ValueIteratorBase( const ValueInternalArray::IteratorState &state ) - : isArray_( true ) -{ - iterator_.array_ = state; -} - - -ValueIteratorBase::ValueIteratorBase( const ValueInternalMap::IteratorState &state ) - : isArray_( false ) -{ - iterator_.map_ = state; -} -#endif - -Value & -ValueIteratorBase::deref() const -{ -#ifndef JSON_VALUE_USE_INTERNAL_MAP - return current_->second; -#else - if ( isArray_ ) - return ValueInternalArray::dereference( iterator_.array_ ); - return ValueInternalMap::value( iterator_.map_ ); -#endif -} - - -void -ValueIteratorBase::increment() -{ -#ifndef JSON_VALUE_USE_INTERNAL_MAP - ++current_; -#else - if ( isArray_ ) - ValueInternalArray::increment( iterator_.array_ ); - ValueInternalMap::increment( iterator_.map_ ); -#endif -} - - -void -ValueIteratorBase::decrement() -{ -#ifndef JSON_VALUE_USE_INTERNAL_MAP - --current_; -#else - if ( isArray_ ) - ValueInternalArray::decrement( iterator_.array_ ); - ValueInternalMap::decrement( iterator_.map_ ); -#endif -} - - -ValueIteratorBase::difference_type -ValueIteratorBase::computeDistance( const SelfType &other ) const -{ -#ifndef JSON_VALUE_USE_INTERNAL_MAP -# ifdef JSON_USE_CPPTL_SMALLMAP - return current_ - other.current_; -# else - // Iterator for null value are initialized using the default - // constructor, which initialize current_ to the default - // std::map::iterator. As begin() and end() are two instance - // of the default std::map::iterator, they can not be compared. - // To allow this, we handle this comparison specifically. - if ( isNull_ && other.isNull_ ) - { - return 0; - } - - - // Usage of std::distance is not portable (does not compile with Sun Studio 12 RogueWave STL, - // which is the one used by default). - // Using a portable hand-made version for non random iterator instead: - // return difference_type( std::distance( current_, other.current_ ) ); - difference_type myDistance = 0; - for ( Value::ObjectValues::iterator it = current_; it != other.current_; ++it ) - { - ++myDistance; - } - return myDistance; -# endif -#else - if ( isArray_ ) - return ValueInternalArray::distance( iterator_.array_, other.iterator_.array_ ); - return ValueInternalMap::distance( iterator_.map_, other.iterator_.map_ ); -#endif -} - - -bool -ValueIteratorBase::isEqual( const SelfType &other ) const -{ -#ifndef JSON_VALUE_USE_INTERNAL_MAP - if ( isNull_ ) - { - return other.isNull_; - } - return current_ == other.current_; -#else - if ( isArray_ ) - return ValueInternalArray::equals( iterator_.array_, other.iterator_.array_ ); - return ValueInternalMap::equals( iterator_.map_, other.iterator_.map_ ); -#endif -} - - -void -ValueIteratorBase::copy( const SelfType &other ) -{ -#ifndef JSON_VALUE_USE_INTERNAL_MAP - current_ = other.current_; -#else - if ( isArray_ ) - iterator_.array_ = other.iterator_.array_; - iterator_.map_ = other.iterator_.map_; -#endif -} - - -Value -ValueIteratorBase::key() const -{ -#ifndef JSON_VALUE_USE_INTERNAL_MAP - const Value::CZString czstring = (*current_).first; - if ( czstring.c_str() ) - { - if ( czstring.isStaticString() ) - return Value( StaticString( czstring.c_str() ) ); - return Value( czstring.c_str() ); - } - return Value( czstring.index() ); -#else - if ( isArray_ ) - return Value( ValueInternalArray::indexOf( iterator_.array_ ) ); - bool isStatic; - const char *memberName = ValueInternalMap::key( iterator_.map_, isStatic ); - if ( isStatic ) - return Value( StaticString( memberName ) ); - return Value( memberName ); -#endif -} - - -UInt -ValueIteratorBase::index() const -{ -#ifndef JSON_VALUE_USE_INTERNAL_MAP - const Value::CZString czstring = (*current_).first; - if ( !czstring.c_str() ) - return czstring.index(); - return Value::UInt( -1 ); -#else - if ( isArray_ ) - return Value::UInt( ValueInternalArray::indexOf( iterator_.array_ ) ); - return Value::UInt( -1 ); -#endif -} - - -const char * -ValueIteratorBase::memberName() const -{ -#ifndef JSON_VALUE_USE_INTERNAL_MAP - const char *name = (*current_).first.c_str(); - return name ? name : ""; -#else - if ( !isArray_ ) - return ValueInternalMap::key( iterator_.map_ ); - return ""; -#endif -} - - -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// class ValueConstIterator -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// - -ValueConstIterator::ValueConstIterator() -{ -} - - -#ifndef JSON_VALUE_USE_INTERNAL_MAP -ValueConstIterator::ValueConstIterator( const Value::ObjectValues::iterator ¤t ) - : ValueIteratorBase( current ) -{ -} -#else -ValueConstIterator::ValueConstIterator( const ValueInternalArray::IteratorState &state ) - : ValueIteratorBase( state ) -{ -} - -ValueConstIterator::ValueConstIterator( const ValueInternalMap::IteratorState &state ) - : ValueIteratorBase( state ) -{ -} -#endif - -ValueConstIterator & -ValueConstIterator::operator =( const ValueIteratorBase &other ) -{ - copy( other ); - return *this; -} - - -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// class ValueIterator -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// -// ////////////////////////////////////////////////////////////////// - -ValueIterator::ValueIterator() -{ -} - - -#ifndef JSON_VALUE_USE_INTERNAL_MAP -ValueIterator::ValueIterator( const Value::ObjectValues::iterator ¤t ) - : ValueIteratorBase( current ) -{ -} -#else -ValueIterator::ValueIterator( const ValueInternalArray::IteratorState &state ) - : ValueIteratorBase( state ) -{ -} - -ValueIterator::ValueIterator( const ValueInternalMap::IteratorState &state ) - : ValueIteratorBase( state ) -{ -} -#endif - -ValueIterator::ValueIterator( const ValueConstIterator &other ) - : ValueIteratorBase( other ) -{ -} - -ValueIterator::ValueIterator( const ValueIterator &other ) - : ValueIteratorBase( other ) -{ -} - -ValueIterator & -ValueIterator::operator =( const SelfType &other ) -{ - copy( other ); - return *this; -} - -} // namespace Json diff --git a/ext/jsoncpp/src/json_writer.cpp b/ext/jsoncpp/src/json_writer.cpp deleted file mode 100644 index 1bda183f..00000000 --- a/ext/jsoncpp/src/json_writer.cpp +++ /dev/null @@ -1,838 +0,0 @@ -// Copyright 2007-2010 Baptiste Lepilleur -// Distributed under MIT license, or public domain if desired and -// recognized in your jurisdiction. -// See file LICENSE for detail or copy at http://jsoncpp.sourceforge.net/LICENSE - -#if !defined(JSON_IS_AMALGAMATION) -# include -# include "json_tool.h" -#endif // if !defined(JSON_IS_AMALGAMATION) -#include -#include -#include -#include -#include -#include -#include - -#if _MSC_VER >= 1400 // VC++ 8.0 -#pragma warning( disable : 4996 ) // disable warning about strdup being deprecated. -#endif - -namespace Json { - -static bool containsControlCharacter( const char* str ) -{ - while ( *str ) - { - if ( isControlCharacter( *(str++) ) ) - return true; - } - return false; -} - - -std::string valueToString( LargestInt value ) -{ - UIntToStringBuffer buffer; - char *current = buffer + sizeof(buffer); - bool isNegative = value < 0; - if ( isNegative ) - value = -value; - uintToString( LargestUInt(value), current ); - if ( isNegative ) - *--current = '-'; - assert( current >= buffer ); - return current; -} - - -std::string valueToString( LargestUInt value ) -{ - UIntToStringBuffer buffer; - char *current = buffer + sizeof(buffer); - uintToString( value, current ); - assert( current >= buffer ); - return current; -} - -#if defined(JSON_HAS_INT64) - -std::string valueToString( Int value ) -{ - return valueToString( LargestInt(value) ); -} - - -std::string valueToString( UInt value ) -{ - return valueToString( LargestUInt(value) ); -} - -#endif // # if defined(JSON_HAS_INT64) - - -std::string valueToString( double value ) -{ - char buffer[32]; -#if defined(_MSC_VER) && defined(__STDC_SECURE_LIB__) // Use secure version with visual studio 2005 to avoid warning. - sprintf_s(buffer, sizeof(buffer), "%#.16g", value); -#else - sprintf(buffer, "%#.16g", value); -#endif - char* ch = buffer + strlen(buffer) - 1; - if (*ch != '0') return buffer; // nothing to truncate, so save time - while(ch > buffer && *ch == '0'){ - --ch; - } - char* last_nonzero = ch; - while(ch >= buffer){ - switch(*ch){ - case '0': - case '1': - case '2': - case '3': - case '4': - case '5': - case '6': - case '7': - case '8': - case '9': - --ch; - continue; - case '.': - // Truncate zeroes to save bytes in output, but keep one. - *(last_nonzero+2) = '\0'; - return buffer; - default: - return buffer; - } - } - return buffer; -} - - -std::string valueToString( bool value ) -{ - return value ? "true" : "false"; -} - -std::string valueToQuotedString( const char *value ) -{ - // Not sure how to handle unicode... - if (strpbrk(value, "\"\\\b\f\n\r\t") == NULL && !containsControlCharacter( value )) - return std::string("\"") + value + "\""; - // We have to walk value and escape any special characters. - // Appending to std::string is not efficient, but this should be rare. - // (Note: forward slashes are *not* rare, but I am not escaping them.) - std::string::size_type maxsize = strlen(value)*2 + 3; // allescaped+quotes+NULL - std::string result; - result.reserve(maxsize); // to avoid lots of mallocs - result += "\""; - for (const char* c=value; *c != 0; ++c) - { - switch(*c) - { - case '\"': - result += "\\\""; - break; - case '\\': - result += "\\\\"; - break; - case '\b': - result += "\\b"; - break; - case '\f': - result += "\\f"; - break; - case '\n': - result += "\\n"; - break; - case '\r': - result += "\\r"; - break; - case '\t': - result += "\\t"; - break; - //case '/': - // Even though \/ is considered a legal escape in JSON, a bare - // slash is also legal, so I see no reason to escape it. - // (I hope I am not misunderstanding something. - // blep notes: actually escaping \/ may be useful in javascript to avoid (*c); - result += oss.str(); - } - else - { - result += *c; - } - break; - } - } - result += "\""; - return result; -} - -// Class Writer -// ////////////////////////////////////////////////////////////////// -Writer::~Writer() -{ -} - - -// Class FastWriter -// ////////////////////////////////////////////////////////////////// - -FastWriter::FastWriter() - : yamlCompatiblityEnabled_( false ) -{ -} - - -void -FastWriter::enableYAMLCompatibility() -{ - yamlCompatiblityEnabled_ = true; -} - - -std::string -FastWriter::write( const Value &root ) -{ - document_ = ""; - writeValue( root ); - document_ += "\n"; - return document_; -} - - -void -FastWriter::writeValue( const Value &value ) -{ - switch ( value.type() ) - { - case nullValue: - document_ += "null"; - break; - case intValue: - document_ += valueToString( value.asLargestInt() ); - break; - case uintValue: - document_ += valueToString( value.asLargestUInt() ); - break; - case realValue: - document_ += valueToString( value.asDouble() ); - break; - case stringValue: - document_ += valueToQuotedString( value.asCString() ); - break; - case booleanValue: - document_ += valueToString( value.asBool() ); - break; - case arrayValue: - { - document_ += "["; - int size = value.size(); - for ( int index =0; index < size; ++index ) - { - if ( index > 0 ) - document_ += ","; - writeValue( value[index] ); - } - document_ += "]"; - } - break; - case objectValue: - { - Value::Members members( value.getMemberNames() ); - document_ += "{"; - for ( Value::Members::iterator it = members.begin(); - it != members.end(); - ++it ) - { - const std::string &name = *it; - if ( it != members.begin() ) - document_ += ","; - document_ += valueToQuotedString( name.c_str() ); - document_ += yamlCompatiblityEnabled_ ? ": " - : ":"; - writeValue( value[name] ); - } - document_ += "}"; - } - break; - } -} - - -// Class StyledWriter -// ////////////////////////////////////////////////////////////////// - -StyledWriter::StyledWriter() - : rightMargin_( 74 ) - , indentSize_( 3 ) -{ -} - - -std::string -StyledWriter::write( const Value &root ) -{ - document_ = ""; - addChildValues_ = false; - indentString_ = ""; - writeCommentBeforeValue( root ); - writeValue( root ); - writeCommentAfterValueOnSameLine( root ); - document_ += "\n"; - return document_; -} - - -void -StyledWriter::writeValue( const Value &value ) -{ - switch ( value.type() ) - { - case nullValue: - pushValue( "null" ); - break; - case intValue: - pushValue( valueToString( value.asLargestInt() ) ); - break; - case uintValue: - pushValue( valueToString( value.asLargestUInt() ) ); - break; - case realValue: - pushValue( valueToString( value.asDouble() ) ); - break; - case stringValue: - pushValue( valueToQuotedString( value.asCString() ) ); - break; - case booleanValue: - pushValue( valueToString( value.asBool() ) ); - break; - case arrayValue: - writeArrayValue( value); - break; - case objectValue: - { - Value::Members members( value.getMemberNames() ); - if ( members.empty() ) - pushValue( "{}" ); - else - { - writeWithIndent( "{" ); - indent(); - Value::Members::iterator it = members.begin(); - for (;;) - { - const std::string &name = *it; - const Value &childValue = value[name]; - writeCommentBeforeValue( childValue ); - writeWithIndent( valueToQuotedString( name.c_str() ) ); - document_ += " : "; - writeValue( childValue ); - if ( ++it == members.end() ) - { - writeCommentAfterValueOnSameLine( childValue ); - break; - } - document_ += ","; - writeCommentAfterValueOnSameLine( childValue ); - } - unindent(); - writeWithIndent( "}" ); - } - } - break; - } -} - - -void -StyledWriter::writeArrayValue( const Value &value ) -{ - unsigned size = value.size(); - if ( size == 0 ) - pushValue( "[]" ); - else - { - bool isArrayMultiLine = isMultineArray( value ); - if ( isArrayMultiLine ) - { - writeWithIndent( "[" ); - indent(); - bool hasChildValue = !childValues_.empty(); - unsigned index =0; - for (;;) - { - const Value &childValue = value[index]; - writeCommentBeforeValue( childValue ); - if ( hasChildValue ) - writeWithIndent( childValues_[index] ); - else - { - writeIndent(); - writeValue( childValue ); - } - if ( ++index == size ) - { - writeCommentAfterValueOnSameLine( childValue ); - break; - } - document_ += ","; - writeCommentAfterValueOnSameLine( childValue ); - } - unindent(); - writeWithIndent( "]" ); - } - else // output on a single line - { - assert( childValues_.size() == size ); - document_ += "[ "; - for ( unsigned index =0; index < size; ++index ) - { - if ( index > 0 ) - document_ += ", "; - document_ += childValues_[index]; - } - document_ += " ]"; - } - } -} - - -bool -StyledWriter::isMultineArray( const Value &value ) -{ - int size = value.size(); - bool isMultiLine = size*3 >= rightMargin_ ; - childValues_.clear(); - for ( int index =0; index < size && !isMultiLine; ++index ) - { - const Value &childValue = value[index]; - isMultiLine = isMultiLine || - ( (childValue.isArray() || childValue.isObject()) && - childValue.size() > 0 ); - } - if ( !isMultiLine ) // check if line length > max line length - { - childValues_.reserve( size ); - addChildValues_ = true; - int lineLength = 4 + (size-1)*2; // '[ ' + ', '*n + ' ]' - for ( int index =0; index < size && !isMultiLine; ++index ) - { - writeValue( value[index] ); - lineLength += int( childValues_[index].length() ); - isMultiLine = isMultiLine && hasCommentForValue( value[index] ); - } - addChildValues_ = false; - isMultiLine = isMultiLine || lineLength >= rightMargin_; - } - return isMultiLine; -} - - -void -StyledWriter::pushValue( const std::string &value ) -{ - if ( addChildValues_ ) - childValues_.push_back( value ); - else - document_ += value; -} - - -void -StyledWriter::writeIndent() -{ - if ( !document_.empty() ) - { - char last = document_[document_.length()-1]; - if ( last == ' ' ) // already indented - return; - if ( last != '\n' ) // Comments may add new-line - document_ += '\n'; - } - document_ += indentString_; -} - - -void -StyledWriter::writeWithIndent( const std::string &value ) -{ - writeIndent(); - document_ += value; -} - - -void -StyledWriter::indent() -{ - indentString_ += std::string( indentSize_, ' ' ); -} - - -void -StyledWriter::unindent() -{ - assert( int(indentString_.size()) >= indentSize_ ); - indentString_.resize( indentString_.size() - indentSize_ ); -} - - -void -StyledWriter::writeCommentBeforeValue( const Value &root ) -{ - if ( !root.hasComment( commentBefore ) ) - return; - document_ += normalizeEOL( root.getComment( commentBefore ) ); - document_ += "\n"; -} - - -void -StyledWriter::writeCommentAfterValueOnSameLine( const Value &root ) -{ - if ( root.hasComment( commentAfterOnSameLine ) ) - document_ += " " + normalizeEOL( root.getComment( commentAfterOnSameLine ) ); - - if ( root.hasComment( commentAfter ) ) - { - document_ += "\n"; - document_ += normalizeEOL( root.getComment( commentAfter ) ); - document_ += "\n"; - } -} - - -bool -StyledWriter::hasCommentForValue( const Value &value ) -{ - return value.hasComment( commentBefore ) - || value.hasComment( commentAfterOnSameLine ) - || value.hasComment( commentAfter ); -} - - -std::string -StyledWriter::normalizeEOL( const std::string &text ) -{ - std::string normalized; - normalized.reserve( text.length() ); - const char *begin = text.c_str(); - const char *end = begin + text.length(); - const char *current = begin; - while ( current != end ) - { - char c = *current++; - if ( c == '\r' ) // mac or dos EOL - { - if ( *current == '\n' ) // convert dos EOL - ++current; - normalized += '\n'; - } - else // handle unix EOL & other char - normalized += c; - } - return normalized; -} - - -// Class StyledStreamWriter -// ////////////////////////////////////////////////////////////////// - -StyledStreamWriter::StyledStreamWriter( std::string indentation ) - : document_(NULL) - , rightMargin_( 74 ) - , indentation_( indentation ) -{ -} - - -void -StyledStreamWriter::write( std::ostream &out, const Value &root ) -{ - document_ = &out; - addChildValues_ = false; - indentString_ = ""; - writeCommentBeforeValue( root ); - writeValue( root ); - writeCommentAfterValueOnSameLine( root ); - *document_ << "\n"; - document_ = NULL; // Forget the stream, for safety. -} - - -void -StyledStreamWriter::writeValue( const Value &value ) -{ - switch ( value.type() ) - { - case nullValue: - pushValue( "null" ); - break; - case intValue: - pushValue( valueToString( value.asLargestInt() ) ); - break; - case uintValue: - pushValue( valueToString( value.asLargestUInt() ) ); - break; - case realValue: - pushValue( valueToString( value.asDouble() ) ); - break; - case stringValue: - pushValue( valueToQuotedString( value.asCString() ) ); - break; - case booleanValue: - pushValue( valueToString( value.asBool() ) ); - break; - case arrayValue: - writeArrayValue( value); - break; - case objectValue: - { - Value::Members members( value.getMemberNames() ); - if ( members.empty() ) - pushValue( "{}" ); - else - { - writeWithIndent( "{" ); - indent(); - Value::Members::iterator it = members.begin(); - for (;;) - { - const std::string &name = *it; - const Value &childValue = value[name]; - writeCommentBeforeValue( childValue ); - writeWithIndent( valueToQuotedString( name.c_str() ) ); - *document_ << " : "; - writeValue( childValue ); - if ( ++it == members.end() ) - { - writeCommentAfterValueOnSameLine( childValue ); - break; - } - *document_ << ","; - writeCommentAfterValueOnSameLine( childValue ); - } - unindent(); - writeWithIndent( "}" ); - } - } - break; - } -} - - -void -StyledStreamWriter::writeArrayValue( const Value &value ) -{ - unsigned size = value.size(); - if ( size == 0 ) - pushValue( "[]" ); - else - { - bool isArrayMultiLine = isMultineArray( value ); - if ( isArrayMultiLine ) - { - writeWithIndent( "[" ); - indent(); - bool hasChildValue = !childValues_.empty(); - unsigned index =0; - for (;;) - { - const Value &childValue = value[index]; - writeCommentBeforeValue( childValue ); - if ( hasChildValue ) - writeWithIndent( childValues_[index] ); - else - { - writeIndent(); - writeValue( childValue ); - } - if ( ++index == size ) - { - writeCommentAfterValueOnSameLine( childValue ); - break; - } - *document_ << ","; - writeCommentAfterValueOnSameLine( childValue ); - } - unindent(); - writeWithIndent( "]" ); - } - else // output on a single line - { - assert( childValues_.size() == size ); - *document_ << "[ "; - for ( unsigned index =0; index < size; ++index ) - { - if ( index > 0 ) - *document_ << ", "; - *document_ << childValues_[index]; - } - *document_ << " ]"; - } - } -} - - -bool -StyledStreamWriter::isMultineArray( const Value &value ) -{ - int size = value.size(); - bool isMultiLine = size*3 >= rightMargin_ ; - childValues_.clear(); - for ( int index =0; index < size && !isMultiLine; ++index ) - { - const Value &childValue = value[index]; - isMultiLine = isMultiLine || - ( (childValue.isArray() || childValue.isObject()) && - childValue.size() > 0 ); - } - if ( !isMultiLine ) // check if line length > max line length - { - childValues_.reserve( size ); - addChildValues_ = true; - int lineLength = 4 + (size-1)*2; // '[ ' + ', '*n + ' ]' - for ( int index =0; index < size && !isMultiLine; ++index ) - { - writeValue( value[index] ); - lineLength += int( childValues_[index].length() ); - isMultiLine = isMultiLine && hasCommentForValue( value[index] ); - } - addChildValues_ = false; - isMultiLine = isMultiLine || lineLength >= rightMargin_; - } - return isMultiLine; -} - - -void -StyledStreamWriter::pushValue( const std::string &value ) -{ - if ( addChildValues_ ) - childValues_.push_back( value ); - else - *document_ << value; -} - - -void -StyledStreamWriter::writeIndent() -{ - /* - Some comments in this method would have been nice. ;-) - - if ( !document_.empty() ) - { - char last = document_[document_.length()-1]; - if ( last == ' ' ) // already indented - return; - if ( last != '\n' ) // Comments may add new-line - *document_ << '\n'; - } - */ - *document_ << '\n' << indentString_; -} - - -void -StyledStreamWriter::writeWithIndent( const std::string &value ) -{ - writeIndent(); - *document_ << value; -} - - -void -StyledStreamWriter::indent() -{ - indentString_ += indentation_; -} - - -void -StyledStreamWriter::unindent() -{ - assert( indentString_.size() >= indentation_.size() ); - indentString_.resize( indentString_.size() - indentation_.size() ); -} - - -void -StyledStreamWriter::writeCommentBeforeValue( const Value &root ) -{ - if ( !root.hasComment( commentBefore ) ) - return; - *document_ << normalizeEOL( root.getComment( commentBefore ) ); - *document_ << "\n"; -} - - -void -StyledStreamWriter::writeCommentAfterValueOnSameLine( const Value &root ) -{ - if ( root.hasComment( commentAfterOnSameLine ) ) - *document_ << " " + normalizeEOL( root.getComment( commentAfterOnSameLine ) ); - - if ( root.hasComment( commentAfter ) ) - { - *document_ << "\n"; - *document_ << normalizeEOL( root.getComment( commentAfter ) ); - *document_ << "\n"; - } -} - - -bool -StyledStreamWriter::hasCommentForValue( const Value &value ) -{ - return value.hasComment( commentBefore ) - || value.hasComment( commentAfterOnSameLine ) - || value.hasComment( commentAfter ); -} - - -std::string -StyledStreamWriter::normalizeEOL( const std::string &text ) -{ - std::string normalized; - normalized.reserve( text.length() ); - const char *begin = text.c_str(); - const char *end = begin + text.length(); - const char *current = begin; - while ( current != end ) - { - char c = *current++; - if ( c == '\r' ) // mac or dos EOL - { - if ( *current == '\n' ) // convert dos EOL - ++current; - normalized += '\n'; - } - else // handle unix EOL & other char - normalized += c; - } - return normalized; -} - - -std::ostream& operator<<( std::ostream &sout, const Value &root ) -{ - Json::StyledStreamWriter writer; - writer.write(sout, root); - return sout; -} - - -} // namespace Json -- cgit v1.2.3 From e6e825da706f6a59e597aac7ab65b846ea3a7fd4 Mon Sep 17 00:00:00 2001 From: Adam Ierymenko Date: Sat, 27 Jul 2013 13:54:59 -0400 Subject: Get rid of built-in HTTP. If we do HTTP it will probably be via libcurl to support HTTPS and chunking and other complexities. --- AUTHORS.txt | 6 - ext/http-parser/AUTHORS | 41 - ext/http-parser/CONTRIBUTIONS | 4 - ext/http-parser/LICENSE-MIT | 23 - ext/http-parser/README.md | 180 -- ext/http-parser/contrib/parsertrace.c | 156 -- ext/http-parser/contrib/url_parser.c | 44 - ext/http-parser/http_parser.c | 2175 --------------------- ext/http-parser/http_parser.gyp | 111 -- ext/http-parser/http_parser.h | 304 --- ext/http-parser/test.c | 3425 --------------------------------- node/Http.cpp | 323 ---- node/Http.hpp | 129 -- objects.mk | 2 - 14 files changed, 6923 deletions(-) delete mode 100644 ext/http-parser/AUTHORS delete mode 100644 ext/http-parser/CONTRIBUTIONS delete mode 100644 ext/http-parser/LICENSE-MIT delete mode 100644 ext/http-parser/README.md delete mode 100644 ext/http-parser/contrib/parsertrace.c delete mode 100644 ext/http-parser/contrib/url_parser.c delete mode 100644 ext/http-parser/http_parser.c delete mode 100644 ext/http-parser/http_parser.gyp delete mode 100644 ext/http-parser/http_parser.h delete mode 100644 ext/http-parser/test.c delete mode 100644 node/Http.cpp delete mode 100644 node/Http.hpp (limited to 'ext') diff --git a/AUTHORS.txt b/AUTHORS.txt index 7374b7c8..f304f6af 100644 --- a/AUTHORS.txt +++ b/AUTHORS.txt @@ -16,12 +16,6 @@ ZeroTier One includes or links with the following third party software: * LZ4 compression algorithm by Yann Collet (BSD license) http://code.google.com/p/lz4/ - * JsonCPP by Baptiste Lepilleur (public domain) - http://jsoncpp.sourceforge.net - - * http-parser, a simple C http parser library (MIT license) - https://github.com/joyent/http-parser - * OpenSSL libcrypto (BSD-style OpenSSL license) http://www.openssl.org/ diff --git a/ext/http-parser/AUTHORS b/ext/http-parser/AUTHORS deleted file mode 100644 index 92ee45ca..00000000 --- a/ext/http-parser/AUTHORS +++ /dev/null @@ -1,41 +0,0 @@ -# Authors ordered by first contribution. -Ryan Dahl -Jeremy Hinegardner -Sergey Shepelev -Joe Damato -tomika -Phoenix Sol -Cliff Frey -Ewen Cheslack-Postava -Santiago Gala -Tim Becker -Jeff Terrace -Ben Noordhuis -Nathan Rajlich -Mark Nottingham -Aman Gupta -Tim Becker -Sean Cunningham -Peter Griess -Salman Haq -Cliff Frey -Jon Kolb -Fouad Mardini -Paul Querna -Felix Geisendörfer -koichik -Andre Caron -Ivo Raisr -James McLaughlin -David Gwynne -Thomas LE ROUX -Randy Rizun -Andre Louis Caron -Simon Zimmermann -Erik Dubbelboer -Martell Malone -Bertrand Paquet -BogDan Vatra -Peter Faiman -Corey Richardson -Tóth Tamás diff --git a/ext/http-parser/CONTRIBUTIONS b/ext/http-parser/CONTRIBUTIONS deleted file mode 100644 index 11ba31e4..00000000 --- a/ext/http-parser/CONTRIBUTIONS +++ /dev/null @@ -1,4 +0,0 @@ -Contributors must agree to the Contributor License Agreement before patches -can be accepted. - -http://spreadsheets2.google.com/viewform?hl=en&formkey=dDJXOGUwbzlYaWM4cHN1MERwQS1CSnc6MQ diff --git a/ext/http-parser/LICENSE-MIT b/ext/http-parser/LICENSE-MIT deleted file mode 100644 index 58010b38..00000000 --- a/ext/http-parser/LICENSE-MIT +++ /dev/null @@ -1,23 +0,0 @@ -http_parser.c is based on src/http/ngx_http_parse.c from NGINX copyright -Igor Sysoev. - -Additional changes are licensed under the same terms as NGINX and -copyright Joyent, Inc. and other Node contributors. All rights reserved. - -Permission is hereby granted, free of charge, to any person obtaining a copy -of this software and associated documentation files (the "Software"), to -deal in the Software without restriction, including without limitation the -rights to use, copy, modify, merge, publish, distribute, sublicense, and/or -sell copies of the Software, and to permit persons to whom the Software is -furnished to do so, subject to the following conditions: - -The above copyright notice and this permission notice shall be included in -all copies or substantial portions of the Software. - -THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR -IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, -FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE -AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER -LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING -FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS -IN THE SOFTWARE. diff --git a/ext/http-parser/README.md b/ext/http-parser/README.md deleted file mode 100644 index b63418af..00000000 --- a/ext/http-parser/README.md +++ /dev/null @@ -1,180 +0,0 @@ -HTTP Parser -=========== - -[![Build Status](https://travis-ci.org/joyent/http-parser.png?branch=master)](https://travis-ci.org/joyent/http-parser) - -This is a parser for HTTP messages written in C. It parses both requests and -responses. The parser is designed to be used in performance HTTP -applications. It does not make any syscalls nor allocations, it does not -buffer data, it can be interrupted at anytime. Depending on your -architecture, it only requires about 40 bytes of data per message -stream (in a web server that is per connection). - -Features: - - * No dependencies - * Handles persistent streams (keep-alive). - * Decodes chunked encoding. - * Upgrade support - * Defends against buffer overflow attacks. - -The parser extracts the following information from HTTP messages: - - * Header fields and values - * Content-Length - * Request method - * Response status code - * Transfer-Encoding - * HTTP version - * Request URL - * Message body - - -Usage ------ - -One `http_parser` object is used per TCP connection. Initialize the struct -using `http_parser_init()` and set the callbacks. That might look something -like this for a request parser: - - http_parser_settings settings; - settings.on_url = my_url_callback; - settings.on_header_field = my_header_field_callback; - /* ... */ - - http_parser *parser = malloc(sizeof(http_parser)); - http_parser_init(parser, HTTP_REQUEST); - parser->data = my_socket; - -When data is received on the socket execute the parser and check for errors. - - size_t len = 80*1024, nparsed; - char buf[len]; - ssize_t recved; - - recved = recv(fd, buf, len, 0); - - if (recved < 0) { - /* Handle error. */ - } - - /* Start up / continue the parser. - * Note we pass recved==0 to signal that EOF has been recieved. - */ - nparsed = http_parser_execute(parser, &settings, buf, recved); - - if (parser->upgrade) { - /* handle new protocol */ - } else if (nparsed != recved) { - /* Handle error. Usually just close the connection. */ - } - -HTTP needs to know where the end of the stream is. For example, sometimes -servers send responses without Content-Length and expect the client to -consume input (for the body) until EOF. To tell http_parser about EOF, give -`0` as the forth parameter to `http_parser_execute()`. Callbacks and errors -can still be encountered during an EOF, so one must still be prepared -to receive them. - -Scalar valued message information such as `status_code`, `method`, and the -HTTP version are stored in the parser structure. This data is only -temporally stored in `http_parser` and gets reset on each new message. If -this information is needed later, copy it out of the structure during the -`headers_complete` callback. - -The parser decodes the transfer-encoding for both requests and responses -transparently. That is, a chunked encoding is decoded before being sent to -the on_body callback. - - -The Special Problem of Upgrade ------------------------------- - -HTTP supports upgrading the connection to a different protocol. An -increasingly common example of this is the Web Socket protocol which sends -a request like - - GET /demo HTTP/1.1 - Upgrade: WebSocket - Connection: Upgrade - Host: example.com - Origin: http://example.com - WebSocket-Protocol: sample - -followed by non-HTTP data. - -(See http://tools.ietf.org/html/draft-hixie-thewebsocketprotocol-75 for more -information the Web Socket protocol.) - -To support this, the parser will treat this as a normal HTTP message without a -body. Issuing both on_headers_complete and on_message_complete callbacks. However -http_parser_execute() will stop parsing at the end of the headers and return. - -The user is expected to check if `parser->upgrade` has been set to 1 after -`http_parser_execute()` returns. Non-HTTP data begins at the buffer supplied -offset by the return value of `http_parser_execute()`. - - -Callbacks ---------- - -During the `http_parser_execute()` call, the callbacks set in -`http_parser_settings` will be executed. The parser maintains state and -never looks behind, so buffering the data is not necessary. If you need to -save certain data for later usage, you can do that from the callbacks. - -There are two types of callbacks: - -* notification `typedef int (*http_cb) (http_parser*);` - Callbacks: on_message_begin, on_headers_complete, on_message_complete. -* data `typedef int (*http_data_cb) (http_parser*, const char *at, size_t length);` - Callbacks: (requests only) on_uri, - (common) on_header_field, on_header_value, on_body; - -Callbacks must return 0 on success. Returning a non-zero value indicates -error to the parser, making it exit immediately. - -In case you parse HTTP message in chunks (i.e. `read()` request line -from socket, parse, read half headers, parse, etc) your data callbacks -may be called more than once. Http-parser guarantees that data pointer is only -valid for the lifetime of callback. You can also `read()` into a heap allocated -buffer to avoid copying memory around if this fits your application. - -Reading headers may be a tricky task if you read/parse headers partially. -Basically, you need to remember whether last header callback was field or value -and apply following logic: - - (on_header_field and on_header_value shortened to on_h_*) - ------------------------ ------------ -------------------------------------------- - | State (prev. callback) | Callback | Description/action | - ------------------------ ------------ -------------------------------------------- - | nothing (first call) | on_h_field | Allocate new buffer and copy callback data | - | | | into it | - ------------------------ ------------ -------------------------------------------- - | value | on_h_field | New header started. | - | | | Copy current name,value buffers to headers | - | | | list and allocate new buffer for new name | - ------------------------ ------------ -------------------------------------------- - | field | on_h_field | Previous name continues. Reallocate name | - | | | buffer and append callback data to it | - ------------------------ ------------ -------------------------------------------- - | field | on_h_value | Value for current header started. Allocate | - | | | new buffer and copy callback data to it | - ------------------------ ------------ -------------------------------------------- - | value | on_h_value | Value continues. Reallocate value buffer | - | | | and append callback data to it | - ------------------------ ------------ -------------------------------------------- - - -Parsing URLs ------------- - -A simplistic zero-copy URL parser is provided as `http_parser_parse_url()`. -Users of this library may wish to use it to parse URLs constructed from -consecutive `on_url` callbacks. - -See examples of reading in headers: - -* [partial example](http://gist.github.com/155877) in C -* [from http-parser tests](http://github.com/joyent/http-parser/blob/37a0ff8/test.c#L403) in C -* [from Node library](http://github.com/joyent/node/blob/842eaf4/src/http.js#L284) in Javascript diff --git a/ext/http-parser/contrib/parsertrace.c b/ext/http-parser/contrib/parsertrace.c deleted file mode 100644 index c9bc71ec..00000000 --- a/ext/http-parser/contrib/parsertrace.c +++ /dev/null @@ -1,156 +0,0 @@ -/* Based on src/http/ngx_http_parse.c from NGINX copyright Igor Sysoev - * - * Additional changes are licensed under the same terms as NGINX and - * copyright Joyent, Inc. and other Node contributors. All rights reserved. - * - * Permission is hereby granted, free of charge, to any person obtaining a copy - * of this software and associated documentation files (the "Software"), to - * deal in the Software without restriction, including without limitation the - * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or - * sell copies of the Software, and to permit persons to whom the Software is - * furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE - * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER - * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING - * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS - * IN THE SOFTWARE. - */ - -/* Dump what the parser finds to stdout as it happen */ - -#include "http_parser.h" -#include -#include -#include - -int on_message_begin(http_parser* _) { - (void)_; - printf("\n***MESSAGE BEGIN***\n\n"); - return 0; -} - -int on_headers_complete(http_parser* _) { - (void)_; - printf("\n***HEADERS COMPLETE***\n\n"); - return 0; -} - -int on_message_complete(http_parser* _) { - (void)_; - printf("\n***MESSAGE COMPLETE***\n\n"); - return 0; -} - -int on_url(http_parser* _, const char* at, size_t length) { - (void)_; - printf("Url: %.*s\n", (int)length, at); - return 0; -} - -int on_header_field(http_parser* _, const char* at, size_t length) { - (void)_; - printf("Header field: %.*s\n", (int)length, at); - return 0; -} - -int on_header_value(http_parser* _, const char* at, size_t length) { - (void)_; - printf("Header value: %.*s\n", (int)length, at); - return 0; -} - -int on_body(http_parser* _, const char* at, size_t length) { - (void)_; - printf("Body: %.*s\n", (int)length, at); - return 0; -} - -void usage(const char* name) { - fprintf(stderr, - "Usage: %s $type $filename\n" - " type: -x, where x is one of {r,b,q}\n" - " parses file as a Response, reQuest, or Both\n", - name); - exit(EXIT_FAILURE); -} - -int main(int argc, char* argv[]) { - enum http_parser_type file_type; - - if (argc != 3) { - usage(argv[0]); - } - - char* type = argv[1]; - if (type[0] != '-') { - usage(argv[0]); - } - - switch (type[1]) { - /* in the case of "-", type[1] will be NUL */ - case 'r': - file_type = HTTP_RESPONSE; - break; - case 'q': - file_type = HTTP_REQUEST; - break; - case 'b': - file_type = HTTP_BOTH; - break; - default: - usage(argv[0]); - } - - char* filename = argv[2]; - FILE* file = fopen(filename, "r"); - if (file == NULL) { - perror("fopen"); - return EXIT_FAILURE; - } - - fseek(file, 0, SEEK_END); - long file_length = ftell(file); - if (file_length == -1) { - perror("ftell"); - return EXIT_FAILURE; - } - fseek(file, 0, SEEK_SET); - - char* data = malloc(file_length); - if (fread(data, 1, file_length, file) != (size_t)file_length) { - fprintf(stderr, "couldn't read entire file\n"); - free(data); - return EXIT_FAILURE; - } - - http_parser_settings settings; - memset(&settings, 0, sizeof(settings)); - settings.on_message_begin = on_message_begin; - settings.on_url = on_url; - settings.on_header_field = on_header_field; - settings.on_header_value = on_header_value; - settings.on_headers_complete = on_headers_complete; - settings.on_body = on_body; - settings.on_message_complete = on_message_complete; - - http_parser parser; - http_parser_init(&parser, file_type); - size_t nparsed = http_parser_execute(&parser, &settings, data, file_length); - free(data); - - if (nparsed != (size_t)file_length) { - fprintf(stderr, - "Error: %s (%s)\n", - http_errno_description(HTTP_PARSER_ERRNO(&parser)), - http_errno_name(HTTP_PARSER_ERRNO(&parser))); - return EXIT_FAILURE; - } - - return EXIT_SUCCESS; -} diff --git a/ext/http-parser/contrib/url_parser.c b/ext/http-parser/contrib/url_parser.c deleted file mode 100644 index b1f9c979..00000000 --- a/ext/http-parser/contrib/url_parser.c +++ /dev/null @@ -1,44 +0,0 @@ -#include "http_parser.h" -#include -#include - -void -dump_url (const char *url, const struct http_parser_url *u) -{ - unsigned int i; - - printf("\tfield_set: 0x%x, port: %u\n", u->field_set, u->port); - for (i = 0; i < UF_MAX; i++) { - if ((u->field_set & (1 << i)) == 0) { - printf("\tfield_data[%u]: unset\n", i); - continue; - } - - printf("\tfield_data[%u]: off: %u len: %u part: \"%.*s\n", - i, - u->field_data[i].off, - u->field_data[i].len, - u->field_data[i].len, - url + u->field_data[i].off); - } -} - -int main(int argc, char ** argv) { - if (argc != 3) { - printf("Syntax : %s connect|get url\n", argv[0]); - return 1; - } - struct http_parser_url u; - int len = strlen(argv[2]); - int connect = strcmp("connect", argv[1]) == 0 ? 1 : 0; - printf("Parsing %s, connect %d\n", argv[2], connect); - - int result = http_parser_parse_url(argv[2], len, connect, &u); - if (result != 0) { - printf("Parse error : %d\n", result); - return result; - } - printf("Parse ok, result : \n"); - dump_url(argv[2], &u); - return 0; -} \ No newline at end of file diff --git a/ext/http-parser/http_parser.c b/ext/http-parser/http_parser.c deleted file mode 100644 index ed3a9232..00000000 --- a/ext/http-parser/http_parser.c +++ /dev/null @@ -1,2175 +0,0 @@ -/* Based on src/http/ngx_http_parse.c from NGINX copyright Igor Sysoev - * - * Additional changes are licensed under the same terms as NGINX and - * copyright Joyent, Inc. and other Node contributors. All rights reserved. - * - * Permission is hereby granted, free of charge, to any person obtaining a copy - * of this software and associated documentation files (the "Software"), to - * deal in the Software without restriction, including without limitation the - * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or - * sell copies of the Software, and to permit persons to whom the Software is - * furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE - * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER - * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING - * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS - * IN THE SOFTWARE. - */ -#include "http_parser.h" -#include -#include -#include -#include -#include -#include - -#ifndef ULLONG_MAX -# define ULLONG_MAX ((uint64_t) -1) /* 2^64-1 */ -#endif - -#ifndef MIN -# define MIN(a,b) ((a) < (b) ? (a) : (b)) -#endif - -#ifndef ARRAY_SIZE -# define ARRAY_SIZE(a) (sizeof(a) / sizeof((a)[0])) -#endif - -#ifndef BIT_AT -# define BIT_AT(a, i) \ - (!!((unsigned int) (a)[(unsigned int) (i) >> 3] & \ - (1 << ((unsigned int) (i) & 7)))) -#endif - -#ifndef ELEM_AT -# define ELEM_AT(a, i, v) ((unsigned int) (i) < ARRAY_SIZE(a) ? (a)[(i)] : (v)) -#endif - -#define SET_ERRNO(e) \ -do { \ - parser->http_errno = (e); \ -} while(0) - - -/* Run the notify callback FOR, returning ER if it fails */ -#define CALLBACK_NOTIFY_(FOR, ER) \ -do { \ - assert(HTTP_PARSER_ERRNO(parser) == HPE_OK); \ - \ - if (settings->on_##FOR) { \ - if (0 != settings->on_##FOR(parser)) { \ - SET_ERRNO(HPE_CB_##FOR); \ - } \ - \ - /* We either errored above or got paused; get out */ \ - if (HTTP_PARSER_ERRNO(parser) != HPE_OK) { \ - return (ER); \ - } \ - } \ -} while (0) - -/* Run the notify callback FOR and consume the current byte */ -#define CALLBACK_NOTIFY(FOR) CALLBACK_NOTIFY_(FOR, p - data + 1) - -/* Run the notify callback FOR and don't consume the current byte */ -#define CALLBACK_NOTIFY_NOADVANCE(FOR) CALLBACK_NOTIFY_(FOR, p - data) - -/* Run data callback FOR with LEN bytes, returning ER if it fails */ -#define CALLBACK_DATA_(FOR, LEN, ER) \ -do { \ - assert(HTTP_PARSER_ERRNO(parser) == HPE_OK); \ - \ - if (FOR##_mark) { \ - if (settings->on_##FOR) { \ - if (0 != settings->on_##FOR(parser, FOR##_mark, (LEN))) { \ - SET_ERRNO(HPE_CB_##FOR); \ - } \ - \ - /* We either errored above or got paused; get out */ \ - if (HTTP_PARSER_ERRNO(parser) != HPE_OK) { \ - return (ER); \ - } \ - } \ - FOR##_mark = NULL; \ - } \ -} while (0) - -/* Run the data callback FOR and consume the current byte */ -#define CALLBACK_DATA(FOR) \ - CALLBACK_DATA_(FOR, p - FOR##_mark, p - data + 1) - -/* Run the data callback FOR and don't consume the current byte */ -#define CALLBACK_DATA_NOADVANCE(FOR) \ - CALLBACK_DATA_(FOR, p - FOR##_mark, p - data) - -/* Set the mark FOR; non-destructive if mark is already set */ -#define MARK(FOR) \ -do { \ - if (!FOR##_mark) { \ - FOR##_mark = p; \ - } \ -} while (0) - - -#define PROXY_CONNECTION "proxy-connection" -#define CONNECTION "connection" -#define CONTENT_LENGTH "content-length" -#define TRANSFER_ENCODING "transfer-encoding" -#define UPGRADE "upgrade" -#define CHUNKED "chunked" -#define KEEP_ALIVE "keep-alive" -#define CLOSE "close" - - -static const char *method_strings[] = - { -#define XX(num, name, string) #string, - HTTP_METHOD_MAP(XX) -#undef XX - }; - - -/* Tokens as defined by rfc 2616. Also lowercases them. - * token = 1* - * separators = "(" | ")" | "<" | ">" | "@" - * | "," | ";" | ":" | "\" | <"> - * | "/" | "[" | "]" | "?" | "=" - * | "{" | "}" | SP | HT - */ -static const char tokens[256] = { -/* 0 nul 1 soh 2 stx 3 etx 4 eot 5 enq 6 ack 7 bel */ - 0, 0, 0, 0, 0, 0, 0, 0, -/* 8 bs 9 ht 10 nl 11 vt 12 np 13 cr 14 so 15 si */ - 0, 0, 0, 0, 0, 0, 0, 0, -/* 16 dle 17 dc1 18 dc2 19 dc3 20 dc4 21 nak 22 syn 23 etb */ - 0, 0, 0, 0, 0, 0, 0, 0, -/* 24 can 25 em 26 sub 27 esc 28 fs 29 gs 30 rs 31 us */ - 0, 0, 0, 0, 0, 0, 0, 0, -/* 32 sp 33 ! 34 " 35 # 36 $ 37 % 38 & 39 ' */ - 0, '!', 0, '#', '$', '%', '&', '\'', -/* 40 ( 41 ) 42 * 43 + 44 , 45 - 46 . 47 / */ - 0, 0, '*', '+', 0, '-', '.', 0, -/* 48 0 49 1 50 2 51 3 52 4 53 5 54 6 55 7 */ - '0', '1', '2', '3', '4', '5', '6', '7', -/* 56 8 57 9 58 : 59 ; 60 < 61 = 62 > 63 ? */ - '8', '9', 0, 0, 0, 0, 0, 0, -/* 64 @ 65 A 66 B 67 C 68 D 69 E 70 F 71 G */ - 0, 'a', 'b', 'c', 'd', 'e', 'f', 'g', -/* 72 H 73 I 74 J 75 K 76 L 77 M 78 N 79 O */ - 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', -/* 80 P 81 Q 82 R 83 S 84 T 85 U 86 V 87 W */ - 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', -/* 88 X 89 Y 90 Z 91 [ 92 \ 93 ] 94 ^ 95 _ */ - 'x', 'y', 'z', 0, 0, 0, '^', '_', -/* 96 ` 97 a 98 b 99 c 100 d 101 e 102 f 103 g */ - '`', 'a', 'b', 'c', 'd', 'e', 'f', 'g', -/* 104 h 105 i 106 j 107 k 108 l 109 m 110 n 111 o */ - 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', -/* 112 p 113 q 114 r 115 s 116 t 117 u 118 v 119 w */ - 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', -/* 120 x 121 y 122 z 123 { 124 | 125 } 126 ~ 127 del */ - 'x', 'y', 'z', 0, '|', 0, '~', 0 }; - - -static const int8_t unhex[256] = - {-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 - ,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 - ,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 - , 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,-1,-1,-1,-1,-1,-1 - ,-1,10,11,12,13,14,15,-1,-1,-1,-1,-1,-1,-1,-1,-1 - ,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 - ,-1,10,11,12,13,14,15,-1,-1,-1,-1,-1,-1,-1,-1,-1 - ,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 - }; - - -#if HTTP_PARSER_STRICT -# define T(v) 0 -#else -# define T(v) v -#endif - - -static const uint8_t normal_url_char[32] = { -/* 0 nul 1 soh 2 stx 3 etx 4 eot 5 enq 6 ack 7 bel */ - 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0, -/* 8 bs 9 ht 10 nl 11 vt 12 np 13 cr 14 so 15 si */ - 0 | T(2) | 0 | 0 | T(16) | 0 | 0 | 0, -/* 16 dle 17 dc1 18 dc2 19 dc3 20 dc4 21 nak 22 syn 23 etb */ - 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0, -/* 24 can 25 em 26 sub 27 esc 28 fs 29 gs 30 rs 31 us */ - 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0, -/* 32 sp 33 ! 34 " 35 # 36 $ 37 % 38 & 39 ' */ - 0 | 2 | 4 | 0 | 16 | 32 | 64 | 128, -/* 40 ( 41 ) 42 * 43 + 44 , 45 - 46 . 47 / */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, -/* 48 0 49 1 50 2 51 3 52 4 53 5 54 6 55 7 */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, -/* 56 8 57 9 58 : 59 ; 60 < 61 = 62 > 63 ? */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 0, -/* 64 @ 65 A 66 B 67 C 68 D 69 E 70 F 71 G */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, -/* 72 H 73 I 74 J 75 K 76 L 77 M 78 N 79 O */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, -/* 80 P 81 Q 82 R 83 S 84 T 85 U 86 V 87 W */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, -/* 88 X 89 Y 90 Z 91 [ 92 \ 93 ] 94 ^ 95 _ */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, -/* 96 ` 97 a 98 b 99 c 100 d 101 e 102 f 103 g */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, -/* 104 h 105 i 106 j 107 k 108 l 109 m 110 n 111 o */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, -/* 112 p 113 q 114 r 115 s 116 t 117 u 118 v 119 w */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, -/* 120 x 121 y 122 z 123 { 124 | 125 } 126 ~ 127 del */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 0, }; - -#undef T - -enum state - { s_dead = 1 /* important that this is > 0 */ - - , s_start_req_or_res - , s_res_or_resp_H - , s_start_res - , s_res_H - , s_res_HT - , s_res_HTT - , s_res_HTTP - , s_res_first_http_major - , s_res_http_major - , s_res_first_http_minor - , s_res_http_minor - , s_res_first_status_code - , s_res_status_code - , s_res_status - , s_res_line_almost_done - - , s_start_req - - , s_req_method - , s_req_spaces_before_url - , s_req_schema - , s_req_schema_slash - , s_req_schema_slash_slash - , s_req_server_start - , s_req_server - , s_req_server_with_at - , s_req_path - , s_req_query_string_start - , s_req_query_string - , s_req_fragment_start - , s_req_fragment - , s_req_http_start - , s_req_http_H - , s_req_http_HT - , s_req_http_HTT - , s_req_http_HTTP - , s_req_first_http_major - , s_req_http_major - , s_req_first_http_minor - , s_req_http_minor - , s_req_line_almost_done - - , s_header_field_start - , s_header_field - , s_header_value_start - , s_header_value - , s_header_value_lws - - , s_header_almost_done - - , s_chunk_size_start - , s_chunk_size - , s_chunk_parameters - , s_chunk_size_almost_done - - , s_headers_almost_done - , s_headers_done - - /* Important: 's_headers_done' must be the last 'header' state. All - * states beyond this must be 'body' states. It is used for overflow - * checking. See the PARSING_HEADER() macro. - */ - - , s_chunk_data - , s_chunk_data_almost_done - , s_chunk_data_done - - , s_body_identity - , s_body_identity_eof - - , s_message_done - }; - - -#define PARSING_HEADER(state) (state <= s_headers_done) - - -enum header_states - { h_general = 0 - , h_C - , h_CO - , h_CON - - , h_matching_connection - , h_matching_proxy_connection - , h_matching_content_length - , h_matching_transfer_encoding - , h_matching_upgrade - - , h_connection - , h_content_length - , h_transfer_encoding - , h_upgrade - - , h_matching_transfer_encoding_chunked - , h_matching_connection_keep_alive - , h_matching_connection_close - - , h_transfer_encoding_chunked - , h_connection_keep_alive - , h_connection_close - }; - -enum http_host_state - { - s_http_host_dead = 1 - , s_http_userinfo_start - , s_http_userinfo - , s_http_host_start - , s_http_host_v6_start - , s_http_host - , s_http_host_v6 - , s_http_host_v6_end - , s_http_host_port_start - , s_http_host_port -}; - -/* Macros for character classes; depends on strict-mode */ -#define CR '\r' -#define LF '\n' -#define LOWER(c) (unsigned char)(c | 0x20) -#define IS_ALPHA(c) (LOWER(c) >= 'a' && LOWER(c) <= 'z') -#define IS_NUM(c) ((c) >= '0' && (c) <= '9') -#define IS_ALPHANUM(c) (IS_ALPHA(c) || IS_NUM(c)) -#define IS_HEX(c) (IS_NUM(c) || (LOWER(c) >= 'a' && LOWER(c) <= 'f')) -#define IS_MARK(c) ((c) == '-' || (c) == '_' || (c) == '.' || \ - (c) == '!' || (c) == '~' || (c) == '*' || (c) == '\'' || (c) == '(' || \ - (c) == ')') -#define IS_USERINFO_CHAR(c) (IS_ALPHANUM(c) || IS_MARK(c) || (c) == '%' || \ - (c) == ';' || (c) == ':' || (c) == '&' || (c) == '=' || (c) == '+' || \ - (c) == '$' || (c) == ',') - -#if HTTP_PARSER_STRICT -#define TOKEN(c) (tokens[(unsigned char)c]) -#define IS_URL_CHAR(c) (BIT_AT(normal_url_char, (unsigned char)c)) -#define IS_HOST_CHAR(c) (IS_ALPHANUM(c) || (c) == '.' || (c) == '-') -#else -#define TOKEN(c) ((c == ' ') ? ' ' : tokens[(unsigned char)c]) -#define IS_URL_CHAR(c) \ - (BIT_AT(normal_url_char, (unsigned char)c) || ((c) & 0x80)) -#define IS_HOST_CHAR(c) \ - (IS_ALPHANUM(c) || (c) == '.' || (c) == '-' || (c) == '_') -#endif - - -#define start_state (parser->type == HTTP_REQUEST ? s_start_req : s_start_res) - - -#if HTTP_PARSER_STRICT -# define STRICT_CHECK(cond) \ -do { \ - if (cond) { \ - SET_ERRNO(HPE_STRICT); \ - goto error; \ - } \ -} while (0) -# define NEW_MESSAGE() (http_should_keep_alive(parser) ? start_state : s_dead) -#else -# define STRICT_CHECK(cond) -# define NEW_MESSAGE() start_state -#endif - - -/* Map errno values to strings for human-readable output */ -#define HTTP_STRERROR_GEN(n, s) { "HPE_" #n, s }, -static struct { - const char *name; - const char *description; -} http_strerror_tab[] = { - HTTP_ERRNO_MAP(HTTP_STRERROR_GEN) -}; -#undef HTTP_STRERROR_GEN - -int http_message_needs_eof(const http_parser *parser); - -/* Our URL parser. - * - * This is designed to be shared by http_parser_execute() for URL validation, - * hence it has a state transition + byte-for-byte interface. In addition, it - * is meant to be embedded in http_parser_parse_url(), which does the dirty - * work of turning state transitions URL components for its API. - * - * This function should only be invoked with non-space characters. It is - * assumed that the caller cares about (and can detect) the transition between - * URL and non-URL states by looking for these. - */ -static enum state -parse_url_char(enum state s, const char ch) -{ - if (ch == ' ' || ch == '\r' || ch == '\n') { - return s_dead; - } - -#if HTTP_PARSER_STRICT - if (ch == '\t' || ch == '\f') { - return s_dead; - } -#endif - - switch (s) { - case s_req_spaces_before_url: - /* Proxied requests are followed by scheme of an absolute URI (alpha). - * All methods except CONNECT are followed by '/' or '*'. - */ - - if (ch == '/' || ch == '*') { - return s_req_path; - } - - if (IS_ALPHA(ch)) { - return s_req_schema; - } - - break; - - case s_req_schema: - if (IS_ALPHA(ch)) { - return s; - } - - if (ch == ':') { - return s_req_schema_slash; - } - - break; - - case s_req_schema_slash: - if (ch == '/') { - return s_req_schema_slash_slash; - } - - break; - - case s_req_schema_slash_slash: - if (ch == '/') { - return s_req_server_start; - } - - break; - - case s_req_server_with_at: - if (ch == '@') { - return s_dead; - } - - /* FALLTHROUGH */ - case s_req_server_start: - case s_req_server: - if (ch == '/') { - return s_req_path; - } - - if (ch == '?') { - return s_req_query_string_start; - } - - if (ch == '@') { - return s_req_server_with_at; - } - - if (IS_USERINFO_CHAR(ch) || ch == '[' || ch == ']') { - return s_req_server; - } - - break; - - case s_req_path: - if (IS_URL_CHAR(ch)) { - return s; - } - - switch (ch) { - case '?': - return s_req_query_string_start; - - case '#': - return s_req_fragment_start; - } - - break; - - case s_req_query_string_start: - case s_req_query_string: - if (IS_URL_CHAR(ch)) { - return s_req_query_string; - } - - switch (ch) { - case '?': - /* allow extra '?' in query string */ - return s_req_query_string; - - case '#': - return s_req_fragment_start; - } - - break; - - case s_req_fragment_start: - if (IS_URL_CHAR(ch)) { - return s_req_fragment; - } - - switch (ch) { - case '?': - return s_req_fragment; - - case '#': - return s; - } - - break; - - case s_req_fragment: - if (IS_URL_CHAR(ch)) { - return s; - } - - switch (ch) { - case '?': - case '#': - return s; - } - - break; - - default: - break; - } - - /* We should never fall out of the switch above unless there's an error */ - return s_dead; -} - -size_t http_parser_execute (http_parser *parser, - const http_parser_settings *settings, - const char *data, - size_t len) -{ - char c, ch; - int8_t unhex_val; - const char *p = data; - const char *header_field_mark = 0; - const char *header_value_mark = 0; - const char *url_mark = 0; - const char *body_mark = 0; - - /* We're in an error state. Don't bother doing anything. */ - if (HTTP_PARSER_ERRNO(parser) != HPE_OK) { - return 0; - } - - if (len == 0) { - switch (parser->state) { - case s_body_identity_eof: - /* Use of CALLBACK_NOTIFY() here would erroneously return 1 byte read if - * we got paused. - */ - CALLBACK_NOTIFY_NOADVANCE(message_complete); - return 0; - - case s_dead: - case s_start_req_or_res: - case s_start_res: - case s_start_req: - return 0; - - default: - SET_ERRNO(HPE_INVALID_EOF_STATE); - return 1; - } - } - - - if (parser->state == s_header_field) - header_field_mark = data; - if (parser->state == s_header_value) - header_value_mark = data; - switch (parser->state) { - case s_req_path: - case s_req_schema: - case s_req_schema_slash: - case s_req_schema_slash_slash: - case s_req_server_start: - case s_req_server: - case s_req_server_with_at: - case s_req_query_string_start: - case s_req_query_string: - case s_req_fragment_start: - case s_req_fragment: - url_mark = data; - break; - } - - for (p=data; p != data + len; p++) { - ch = *p; - - if (PARSING_HEADER(parser->state)) { - ++parser->nread; - /* Buffer overflow attack */ - if (parser->nread > HTTP_MAX_HEADER_SIZE) { - SET_ERRNO(HPE_HEADER_OVERFLOW); - goto error; - } - } - - reexecute_byte: - switch (parser->state) { - - case s_dead: - /* this state is used after a 'Connection: close' message - * the parser will error out if it reads another message - */ - if (ch == CR || ch == LF) - break; - - SET_ERRNO(HPE_CLOSED_CONNECTION); - goto error; - - case s_start_req_or_res: - { - if (ch == CR || ch == LF) - break; - parser->flags = 0; - parser->content_length = ULLONG_MAX; - - if (ch == 'H') { - parser->state = s_res_or_resp_H; - - CALLBACK_NOTIFY(message_begin); - } else { - parser->type = HTTP_REQUEST; - parser->state = s_start_req; - goto reexecute_byte; - } - - break; - } - - case s_res_or_resp_H: - if (ch == 'T') { - parser->type = HTTP_RESPONSE; - parser->state = s_res_HT; - } else { - if (ch != 'E') { - SET_ERRNO(HPE_INVALID_CONSTANT); - goto error; - } - - parser->type = HTTP_REQUEST; - parser->method = HTTP_HEAD; - parser->index = 2; - parser->state = s_req_method; - } - break; - - case s_start_res: - { - parser->flags = 0; - parser->content_length = ULLONG_MAX; - - switch (ch) { - case 'H': - parser->state = s_res_H; - break; - - case CR: - case LF: - break; - - default: - SET_ERRNO(HPE_INVALID_CONSTANT); - goto error; - } - - CALLBACK_NOTIFY(message_begin); - break; - } - - case s_res_H: - STRICT_CHECK(ch != 'T'); - parser->state = s_res_HT; - break; - - case s_res_HT: - STRICT_CHECK(ch != 'T'); - parser->state = s_res_HTT; - break; - - case s_res_HTT: - STRICT_CHECK(ch != 'P'); - parser->state = s_res_HTTP; - break; - - case s_res_HTTP: - STRICT_CHECK(ch != '/'); - parser->state = s_res_first_http_major; - break; - - case s_res_first_http_major: - if (ch < '0' || ch > '9') { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - parser->http_major = ch - '0'; - parser->state = s_res_http_major; - break; - - /* major HTTP version or dot */ - case s_res_http_major: - { - if (ch == '.') { - parser->state = s_res_first_http_minor; - break; - } - - if (!IS_NUM(ch)) { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - parser->http_major *= 10; - parser->http_major += ch - '0'; - - if (parser->http_major > 999) { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - break; - } - - /* first digit of minor HTTP version */ - case s_res_first_http_minor: - if (!IS_NUM(ch)) { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - parser->http_minor = ch - '0'; - parser->state = s_res_http_minor; - break; - - /* minor HTTP version or end of request line */ - case s_res_http_minor: - { - if (ch == ' ') { - parser->state = s_res_first_status_code; - break; - } - - if (!IS_NUM(ch)) { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - parser->http_minor *= 10; - parser->http_minor += ch - '0'; - - if (parser->http_minor > 999) { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - break; - } - - case s_res_first_status_code: - { - if (!IS_NUM(ch)) { - if (ch == ' ') { - break; - } - - SET_ERRNO(HPE_INVALID_STATUS); - goto error; - } - parser->status_code = ch - '0'; - parser->state = s_res_status_code; - break; - } - - case s_res_status_code: - { - if (!IS_NUM(ch)) { - switch (ch) { - case ' ': - parser->state = s_res_status; - break; - case CR: - parser->state = s_res_line_almost_done; - break; - case LF: - parser->state = s_header_field_start; - break; - default: - SET_ERRNO(HPE_INVALID_STATUS); - goto error; - } - break; - } - - parser->status_code *= 10; - parser->status_code += ch - '0'; - - if (parser->status_code > 999) { - SET_ERRNO(HPE_INVALID_STATUS); - goto error; - } - - break; - } - - case s_res_status: - /* the human readable status. e.g. "NOT FOUND" - * we are not humans so just ignore this */ - if (ch == CR) { - parser->state = s_res_line_almost_done; - break; - } - - if (ch == LF) { - parser->state = s_header_field_start; - break; - } - break; - - case s_res_line_almost_done: - STRICT_CHECK(ch != LF); - parser->state = s_header_field_start; - CALLBACK_NOTIFY(status_complete); - break; - - case s_start_req: - { - if (ch == CR || ch == LF) - break; - parser->flags = 0; - parser->content_length = ULLONG_MAX; - - if (!IS_ALPHA(ch)) { - SET_ERRNO(HPE_INVALID_METHOD); - goto error; - } - - parser->method = (enum http_method) 0; - parser->index = 1; - switch (ch) { - case 'C': parser->method = HTTP_CONNECT; /* or COPY, CHECKOUT */ break; - case 'D': parser->method = HTTP_DELETE; break; - case 'G': parser->method = HTTP_GET; break; - case 'H': parser->method = HTTP_HEAD; break; - case 'L': parser->method = HTTP_LOCK; break; - case 'M': parser->method = HTTP_MKCOL; /* or MOVE, MKACTIVITY, MERGE, M-SEARCH */ break; - case 'N': parser->method = HTTP_NOTIFY; break; - case 'O': parser->method = HTTP_OPTIONS; break; - case 'P': parser->method = HTTP_POST; - /* or PROPFIND|PROPPATCH|PUT|PATCH|PURGE */ - break; - case 'R': parser->method = HTTP_REPORT; break; - case 'S': parser->method = HTTP_SUBSCRIBE; /* or SEARCH */ break; - case 'T': parser->method = HTTP_TRACE; break; - case 'U': parser->method = HTTP_UNLOCK; /* or UNSUBSCRIBE */ break; - default: - SET_ERRNO(HPE_INVALID_METHOD); - goto error; - } - parser->state = s_req_method; - - CALLBACK_NOTIFY(message_begin); - - break; - } - - case s_req_method: - { - const char *matcher; - if (ch == '\0') { - SET_ERRNO(HPE_INVALID_METHOD); - goto error; - } - - matcher = method_strings[parser->method]; - if (ch == ' ' && matcher[parser->index] == '\0') { - parser->state = s_req_spaces_before_url; - } else if (ch == matcher[parser->index]) { - ; /* nada */ - } else if (parser->method == HTTP_CONNECT) { - if (parser->index == 1 && ch == 'H') { - parser->method = HTTP_CHECKOUT; - } else if (parser->index == 2 && ch == 'P') { - parser->method = HTTP_COPY; - } else { - goto error; - } - } else if (parser->method == HTTP_MKCOL) { - if (parser->index == 1 && ch == 'O') { - parser->method = HTTP_MOVE; - } else if (parser->index == 1 && ch == 'E') { - parser->method = HTTP_MERGE; - } else if (parser->index == 1 && ch == '-') { - parser->method = HTTP_MSEARCH; - } else if (parser->index == 2 && ch == 'A') { - parser->method = HTTP_MKACTIVITY; - } else { - goto error; - } - } else if (parser->method == HTTP_SUBSCRIBE) { - if (parser->index == 1 && ch == 'E') { - parser->method = HTTP_SEARCH; - } else { - goto error; - } - } else if (parser->index == 1 && parser->method == HTTP_POST) { - if (ch == 'R') { - parser->method = HTTP_PROPFIND; /* or HTTP_PROPPATCH */ - } else if (ch == 'U') { - parser->method = HTTP_PUT; /* or HTTP_PURGE */ - } else if (ch == 'A') { - parser->method = HTTP_PATCH; - } else { - goto error; - } - } else if (parser->index == 2) { - if (parser->method == HTTP_PUT) { - if (ch == 'R') parser->method = HTTP_PURGE; - } else if (parser->method == HTTP_UNLOCK) { - if (ch == 'S') parser->method = HTTP_UNSUBSCRIBE; - } - } else if (parser->index == 4 && parser->method == HTTP_PROPFIND && ch == 'P') { - parser->method = HTTP_PROPPATCH; - } else { - SET_ERRNO(HPE_INVALID_METHOD); - goto error; - } - - ++parser->index; - break; - } - - case s_req_spaces_before_url: - { - if (ch == ' ') break; - - MARK(url); - if (parser->method == HTTP_CONNECT) { - parser->state = s_req_server_start; - } - - parser->state = parse_url_char((enum state)parser->state, ch); - if (parser->state == s_dead) { - SET_ERRNO(HPE_INVALID_URL); - goto error; - } - - break; - } - - case s_req_schema: - case s_req_schema_slash: - case s_req_schema_slash_slash: - case s_req_server_start: - { - switch (ch) { - /* No whitespace allowed here */ - case ' ': - case CR: - case LF: - SET_ERRNO(HPE_INVALID_URL); - goto error; - default: - parser->state = parse_url_char((enum state)parser->state, ch); - if (parser->state == s_dead) { - SET_ERRNO(HPE_INVALID_URL); - goto error; - } - } - - break; - } - - case s_req_server: - case s_req_server_with_at: - case s_req_path: - case s_req_query_string_start: - case s_req_query_string: - case s_req_fragment_start: - case s_req_fragment: - { - switch (ch) { - case ' ': - parser->state = s_req_http_start; - CALLBACK_DATA(url); - break; - case CR: - case LF: - parser->http_major = 0; - parser->http_minor = 9; - parser->state = (ch == CR) ? - s_req_line_almost_done : - s_header_field_start; - CALLBACK_DATA(url); - break; - default: - parser->state = parse_url_char((enum state)parser->state, ch); - if (parser->state == s_dead) { - SET_ERRNO(HPE_INVALID_URL); - goto error; - } - } - break; - } - - case s_req_http_start: - switch (ch) { - case 'H': - parser->state = s_req_http_H; - break; - case ' ': - break; - default: - SET_ERRNO(HPE_INVALID_CONSTANT); - goto error; - } - break; - - case s_req_http_H: - STRICT_CHECK(ch != 'T'); - parser->state = s_req_http_HT; - break; - - case s_req_http_HT: - STRICT_CHECK(ch != 'T'); - parser->state = s_req_http_HTT; - break; - - case s_req_http_HTT: - STRICT_CHECK(ch != 'P'); - parser->state = s_req_http_HTTP; - break; - - case s_req_http_HTTP: - STRICT_CHECK(ch != '/'); - parser->state = s_req_first_http_major; - break; - - /* first digit of major HTTP version */ - case s_req_first_http_major: - if (ch < '1' || ch > '9') { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - parser->http_major = ch - '0'; - parser->state = s_req_http_major; - break; - - /* major HTTP version or dot */ - case s_req_http_major: - { - if (ch == '.') { - parser->state = s_req_first_http_minor; - break; - } - - if (!IS_NUM(ch)) { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - parser->http_major *= 10; - parser->http_major += ch - '0'; - - if (parser->http_major > 999) { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - break; - } - - /* first digit of minor HTTP version */ - case s_req_first_http_minor: - if (!IS_NUM(ch)) { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - parser->http_minor = ch - '0'; - parser->state = s_req_http_minor; - break; - - /* minor HTTP version or end of request line */ - case s_req_http_minor: - { - if (ch == CR) { - parser->state = s_req_line_almost_done; - break; - } - - if (ch == LF) { - parser->state = s_header_field_start; - break; - } - - /* XXX allow spaces after digit? */ - - if (!IS_NUM(ch)) { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - parser->http_minor *= 10; - parser->http_minor += ch - '0'; - - if (parser->http_minor > 999) { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - break; - } - - /* end of request line */ - case s_req_line_almost_done: - { - if (ch != LF) { - SET_ERRNO(HPE_LF_EXPECTED); - goto error; - } - - parser->state = s_header_field_start; - break; - } - - case s_header_field_start: - { - if (ch == CR) { - parser->state = s_headers_almost_done; - break; - } - - if (ch == LF) { - /* they might be just sending \n instead of \r\n so this would be - * the second \n to denote the end of headers*/ - parser->state = s_headers_almost_done; - goto reexecute_byte; - } - - c = TOKEN(ch); - - if (!c) { - SET_ERRNO(HPE_INVALID_HEADER_TOKEN); - goto error; - } - - MARK(header_field); - - parser->index = 0; - parser->state = s_header_field; - - switch (c) { - case 'c': - parser->header_state = h_C; - break; - - case 'p': - parser->header_state = h_matching_proxy_connection; - break; - - case 't': - parser->header_state = h_matching_transfer_encoding; - break; - - case 'u': - parser->header_state = h_matching_upgrade; - break; - - default: - parser->header_state = h_general; - break; - } - break; - } - - case s_header_field: - { - c = TOKEN(ch); - - if (c) { - switch (parser->header_state) { - case h_general: - break; - - case h_C: - parser->index++; - parser->header_state = (c == 'o' ? h_CO : h_general); - break; - - case h_CO: - parser->index++; - parser->header_state = (c == 'n' ? h_CON : h_general); - break; - - case h_CON: - parser->index++; - switch (c) { - case 'n': - parser->header_state = h_matching_connection; - break; - case 't': - parser->header_state = h_matching_content_length; - break; - default: - parser->header_state = h_general; - break; - } - break; - - /* connection */ - - case h_matching_connection: - parser->index++; - if (parser->index > sizeof(CONNECTION)-1 - || c != CONNECTION[parser->index]) { - parser->header_state = h_general; - } else if (parser->index == sizeof(CONNECTION)-2) { - parser->header_state = h_connection; - } - break; - - /* proxy-connection */ - - case h_matching_proxy_connection: - parser->index++; - if (parser->index > sizeof(PROXY_CONNECTION)-1 - || c != PROXY_CONNECTION[parser->index]) { - parser->header_state = h_general; - } else if (parser->index == sizeof(PROXY_CONNECTION)-2) { - parser->header_state = h_connection; - } - break; - - /* content-length */ - - case h_matching_content_length: - parser->index++; - if (parser->index > sizeof(CONTENT_LENGTH)-1 - || c != CONTENT_LENGTH[parser->index]) { - parser->header_state = h_general; - } else if (parser->index == sizeof(CONTENT_LENGTH)-2) { - parser->header_state = h_content_length; - } - break; - - /* transfer-encoding */ - - case h_matching_transfer_encoding: - parser->index++; - if (parser->index > sizeof(TRANSFER_ENCODING)-1 - || c != TRANSFER_ENCODING[parser->index]) { - parser->header_state = h_general; - } else if (parser->index == sizeof(TRANSFER_ENCODING)-2) { - parser->header_state = h_transfer_encoding; - } - break; - - /* upgrade */ - - case h_matching_upgrade: - parser->index++; - if (parser->index > sizeof(UPGRADE)-1 - || c != UPGRADE[parser->index]) { - parser->header_state = h_general; - } else if (parser->index == sizeof(UPGRADE)-2) { - parser->header_state = h_upgrade; - } - break; - - case h_connection: - case h_content_length: - case h_transfer_encoding: - case h_upgrade: - if (ch != ' ') parser->header_state = h_general; - break; - - default: - assert(0 && "Unknown header_state"); - break; - } - break; - } - - if (ch == ':') { - parser->state = s_header_value_start; - CALLBACK_DATA(header_field); - break; - } - - if (ch == CR) { - parser->state = s_header_almost_done; - CALLBACK_DATA(header_field); - break; - } - - if (ch == LF) { - parser->state = s_header_field_start; - CALLBACK_DATA(header_field); - break; - } - - SET_ERRNO(HPE_INVALID_HEADER_TOKEN); - goto error; - } - - case s_header_value_start: - { - if (ch == ' ' || ch == '\t') break; - - MARK(header_value); - - parser->state = s_header_value; - parser->index = 0; - - if (ch == CR) { - parser->header_state = h_general; - parser->state = s_header_almost_done; - CALLBACK_DATA(header_value); - break; - } - - if (ch == LF) { - parser->state = s_header_field_start; - CALLBACK_DATA(header_value); - break; - } - - c = LOWER(ch); - - switch (parser->header_state) { - case h_upgrade: - parser->flags |= F_UPGRADE; - parser->header_state = h_general; - break; - - case h_transfer_encoding: - /* looking for 'Transfer-Encoding: chunked' */ - if ('c' == c) { - parser->header_state = h_matching_transfer_encoding_chunked; - } else { - parser->header_state = h_general; - } - break; - - case h_content_length: - if (!IS_NUM(ch)) { - SET_ERRNO(HPE_INVALID_CONTENT_LENGTH); - goto error; - } - - parser->content_length = ch - '0'; - break; - - case h_connection: - /* looking for 'Connection: keep-alive' */ - if (c == 'k') { - parser->header_state = h_matching_connection_keep_alive; - /* looking for 'Connection: close' */ - } else if (c == 'c') { - parser->header_state = h_matching_connection_close; - } else { - parser->header_state = h_general; - } - break; - - default: - parser->header_state = h_general; - break; - } - break; - } - - case s_header_value: - { - - if (ch == CR) { - parser->state = s_header_almost_done; - CALLBACK_DATA(header_value); - break; - } - - if (ch == LF) { - parser->state = s_header_almost_done; - CALLBACK_DATA_NOADVANCE(header_value); - goto reexecute_byte; - } - - c = LOWER(ch); - - switch (parser->header_state) { - case h_general: - break; - - case h_connection: - case h_transfer_encoding: - assert(0 && "Shouldn't get here."); - break; - - case h_content_length: - { - uint64_t t; - - if (ch == ' ') break; - - if (!IS_NUM(ch)) { - SET_ERRNO(HPE_INVALID_CONTENT_LENGTH); - goto error; - } - - t = parser->content_length; - t *= 10; - t += ch - '0'; - - /* Overflow? */ - if (t < parser->content_length || t == ULLONG_MAX) { - SET_ERRNO(HPE_INVALID_CONTENT_LENGTH); - goto error; - } - - parser->content_length = t; - break; - } - - /* Transfer-Encoding: chunked */ - case h_matching_transfer_encoding_chunked: - parser->index++; - if (parser->index > sizeof(CHUNKED)-1 - || c != CHUNKED[parser->index]) { - parser->header_state = h_general; - } else if (parser->index == sizeof(CHUNKED)-2) { - parser->header_state = h_transfer_encoding_chunked; - } - break; - - /* looking for 'Connection: keep-alive' */ - case h_matching_connection_keep_alive: - parser->index++; - if (parser->index > sizeof(KEEP_ALIVE)-1 - || c != KEEP_ALIVE[parser->index]) { - parser->header_state = h_general; - } else if (parser->index == sizeof(KEEP_ALIVE)-2) { - parser->header_state = h_connection_keep_alive; - } - break; - - /* looking for 'Connection: close' */ - case h_matching_connection_close: - parser->index++; - if (parser->index > sizeof(CLOSE)-1 || c != CLOSE[parser->index]) { - parser->header_state = h_general; - } else if (parser->index == sizeof(CLOSE)-2) { - parser->header_state = h_connection_close; - } - break; - - case h_transfer_encoding_chunked: - case h_connection_keep_alive: - case h_connection_close: - if (ch != ' ') parser->header_state = h_general; - break; - - default: - parser->state = s_header_value; - parser->header_state = h_general; - break; - } - break; - } - - case s_header_almost_done: - { - STRICT_CHECK(ch != LF); - - parser->state = s_header_value_lws; - - switch (parser->header_state) { - case h_connection_keep_alive: - parser->flags |= F_CONNECTION_KEEP_ALIVE; - break; - case h_connection_close: - parser->flags |= F_CONNECTION_CLOSE; - break; - case h_transfer_encoding_chunked: - parser->flags |= F_CHUNKED; - break; - default: - break; - } - - break; - } - - case s_header_value_lws: - { - if (ch == ' ' || ch == '\t') - parser->state = s_header_value_start; - else - { - parser->state = s_header_field_start; - goto reexecute_byte; - } - break; - } - - case s_headers_almost_done: - { - STRICT_CHECK(ch != LF); - - if (parser->flags & F_TRAILING) { - /* End of a chunked request */ - parser->state = NEW_MESSAGE(); - CALLBACK_NOTIFY(message_complete); - break; - } - - parser->state = s_headers_done; - - /* Set this here so that on_headers_complete() callbacks can see it */ - parser->upgrade = - (parser->flags & F_UPGRADE || parser->method == HTTP_CONNECT); - - /* Here we call the headers_complete callback. This is somewhat - * different than other callbacks because if the user returns 1, we - * will interpret that as saying that this message has no body. This - * is needed for the annoying case of recieving a response to a HEAD - * request. - * - * We'd like to use CALLBACK_NOTIFY_NOADVANCE() here but we cannot, so - * we have to simulate it by handling a change in errno below. - */ - if (settings->on_headers_complete) { - switch (settings->on_headers_complete(parser)) { - case 0: - break; - - case 1: - parser->flags |= F_SKIPBODY; - break; - - default: - SET_ERRNO(HPE_CB_headers_complete); - return p - data; /* Error */ - } - } - - if (HTTP_PARSER_ERRNO(parser) != HPE_OK) { - return p - data; - } - - goto reexecute_byte; - } - - case s_headers_done: - { - STRICT_CHECK(ch != LF); - - parser->nread = 0; - - /* Exit, the rest of the connect is in a different protocol. */ - if (parser->upgrade) { - parser->state = NEW_MESSAGE(); - CALLBACK_NOTIFY(message_complete); - return (p - data) + 1; - } - - if (parser->flags & F_SKIPBODY) { - parser->state = NEW_MESSAGE(); - CALLBACK_NOTIFY(message_complete); - } else if (parser->flags & F_CHUNKED) { - /* chunked encoding - ignore Content-Length header */ - parser->state = s_chunk_size_start; - } else { - if (parser->content_length == 0) { - /* Content-Length header given but zero: Content-Length: 0\r\n */ - parser->state = NEW_MESSAGE(); - CALLBACK_NOTIFY(message_complete); - } else if (parser->content_length != ULLONG_MAX) { - /* Content-Length header given and non-zero */ - parser->state = s_body_identity; - } else { - if (parser->type == HTTP_REQUEST || - !http_message_needs_eof(parser)) { - /* Assume content-length 0 - read the next */ - parser->state = NEW_MESSAGE(); - CALLBACK_NOTIFY(message_complete); - } else { - /* Read body until EOF */ - parser->state = s_body_identity_eof; - } - } - } - - break; - } - - case s_body_identity: - { - uint64_t to_read = MIN(parser->content_length, - (uint64_t) ((data + len) - p)); - - assert(parser->content_length != 0 - && parser->content_length != ULLONG_MAX); - - /* The difference between advancing content_length and p is because - * the latter will automaticaly advance on the next loop iteration. - * Further, if content_length ends up at 0, we want to see the last - * byte again for our message complete callback. - */ - MARK(body); - parser->content_length -= to_read; - p += to_read - 1; - - if (parser->content_length == 0) { - parser->state = s_message_done; - - /* Mimic CALLBACK_DATA_NOADVANCE() but with one extra byte. - * - * The alternative to doing this is to wait for the next byte to - * trigger the data callback, just as in every other case. The - * problem with this is that this makes it difficult for the test - * harness to distinguish between complete-on-EOF and - * complete-on-length. It's not clear that this distinction is - * important for applications, but let's keep it for now. - */ - CALLBACK_DATA_(body, p - body_mark + 1, p - data); - goto reexecute_byte; - } - - break; - } - - /* read until EOF */ - case s_body_identity_eof: - MARK(body); - p = data + len - 1; - - break; - - case s_message_done: - parser->state = NEW_MESSAGE(); - CALLBACK_NOTIFY(message_complete); - break; - - case s_chunk_size_start: - { - assert(parser->nread == 1); - assert(parser->flags & F_CHUNKED); - - unhex_val = unhex[(unsigned char)ch]; - if (unhex_val == -1) { - SET_ERRNO(HPE_INVALID_CHUNK_SIZE); - goto error; - } - - parser->content_length = unhex_val; - parser->state = s_chunk_size; - break; - } - - case s_chunk_size: - { - uint64_t t; - - assert(parser->flags & F_CHUNKED); - - if (ch == CR) { - parser->state = s_chunk_size_almost_done; - break; - } - - unhex_val = unhex[(unsigned char)ch]; - - if (unhex_val == -1) { - if (ch == ';' || ch == ' ') { - parser->state = s_chunk_parameters; - break; - } - - SET_ERRNO(HPE_INVALID_CHUNK_SIZE); - goto error; - } - - t = parser->content_length; - t *= 16; - t += unhex_val; - - /* Overflow? */ - if (t < parser->content_length || t == ULLONG_MAX) { - SET_ERRNO(HPE_INVALID_CONTENT_LENGTH); - goto error; - } - - parser->content_length = t; - break; - } - - case s_chunk_parameters: - { - assert(parser->flags & F_CHUNKED); - /* just ignore this shit. TODO check for overflow */ - if (ch == CR) { - parser->state = s_chunk_size_almost_done; - break; - } - break; - } - - case s_chunk_size_almost_done: - { - assert(parser->flags & F_CHUNKED); - STRICT_CHECK(ch != LF); - - parser->nread = 0; - - if (parser->content_length == 0) { - parser->flags |= F_TRAILING; - parser->state = s_header_field_start; - } else { - parser->state = s_chunk_data; - } - break; - } - - case s_chunk_data: - { - uint64_t to_read = MIN(parser->content_length, - (uint64_t) ((data + len) - p)); - - assert(parser->flags & F_CHUNKED); - assert(parser->content_length != 0 - && parser->content_length != ULLONG_MAX); - - /* See the explanation in s_body_identity for why the content - * length and data pointers are managed this way. - */ - MARK(body); - parser->content_length -= to_read; - p += to_read - 1; - - if (parser->content_length == 0) { - parser->state = s_chunk_data_almost_done; - } - - break; - } - - case s_chunk_data_almost_done: - assert(parser->flags & F_CHUNKED); - assert(parser->content_length == 0); - STRICT_CHECK(ch != CR); - parser->state = s_chunk_data_done; - CALLBACK_DATA(body); - break; - - case s_chunk_data_done: - assert(parser->flags & F_CHUNKED); - STRICT_CHECK(ch != LF); - parser->nread = 0; - parser->state = s_chunk_size_start; - break; - - default: - assert(0 && "unhandled state"); - SET_ERRNO(HPE_INVALID_INTERNAL_STATE); - goto error; - } - } - - /* Run callbacks for any marks that we have leftover after we ran our of - * bytes. There should be at most one of these set, so it's OK to invoke - * them in series (unset marks will not result in callbacks). - * - * We use the NOADVANCE() variety of callbacks here because 'p' has already - * overflowed 'data' and this allows us to correct for the off-by-one that - * we'd otherwise have (since CALLBACK_DATA() is meant to be run with a 'p' - * value that's in-bounds). - */ - - assert(((header_field_mark ? 1 : 0) + - (header_value_mark ? 1 : 0) + - (url_mark ? 1 : 0) + - (body_mark ? 1 : 0)) <= 1); - - CALLBACK_DATA_NOADVANCE(header_field); - CALLBACK_DATA_NOADVANCE(header_value); - CALLBACK_DATA_NOADVANCE(url); - CALLBACK_DATA_NOADVANCE(body); - - return len; - -error: - if (HTTP_PARSER_ERRNO(parser) == HPE_OK) { - SET_ERRNO(HPE_UNKNOWN); - } - - return (p - data); -} - - -/* Does the parser need to see an EOF to find the end of the message? */ -int -http_message_needs_eof (const http_parser *parser) -{ - if (parser->type == HTTP_REQUEST) { - return 0; - } - - /* See RFC 2616 section 4.4 */ - if (parser->status_code / 100 == 1 || /* 1xx e.g. Continue */ - parser->status_code == 204 || /* No Content */ - parser->status_code == 304 || /* Not Modified */ - parser->flags & F_SKIPBODY) { /* response to a HEAD request */ - return 0; - } - - if ((parser->flags & F_CHUNKED) || parser->content_length != ULLONG_MAX) { - return 0; - } - - return 1; -} - - -int -http_should_keep_alive (const http_parser *parser) -{ - if (parser->http_major > 0 && parser->http_minor > 0) { - /* HTTP/1.1 */ - if (parser->flags & F_CONNECTION_CLOSE) { - return 0; - } - } else { - /* HTTP/1.0 or earlier */ - if (!(parser->flags & F_CONNECTION_KEEP_ALIVE)) { - return 0; - } - } - - return !http_message_needs_eof(parser); -} - - -const char * -http_method_str (enum http_method m) -{ - return ELEM_AT(method_strings, m, ""); -} - - -void -http_parser_init (http_parser *parser, enum http_parser_type t) -{ - void *data = parser->data; /* preserve application data */ - memset(parser, 0, sizeof(*parser)); - parser->data = data; - parser->type = t; - parser->state = (t == HTTP_REQUEST ? s_start_req : (t == HTTP_RESPONSE ? s_start_res : s_start_req_or_res)); - parser->http_errno = HPE_OK; -} - -const char * -http_errno_name(enum http_errno err) { - assert(err < (sizeof(http_strerror_tab)/sizeof(http_strerror_tab[0]))); - return http_strerror_tab[err].name; -} - -const char * -http_errno_description(enum http_errno err) { - assert(err < (sizeof(http_strerror_tab)/sizeof(http_strerror_tab[0]))); - return http_strerror_tab[err].description; -} - -static enum http_host_state -http_parse_host_char(enum http_host_state s, const char ch) { - switch(s) { - case s_http_userinfo: - case s_http_userinfo_start: - if (ch == '@') { - return s_http_host_start; - } - - if (IS_USERINFO_CHAR(ch)) { - return s_http_userinfo; - } - break; - - case s_http_host_start: - if (ch == '[') { - return s_http_host_v6_start; - } - - if (IS_HOST_CHAR(ch)) { - return s_http_host; - } - - break; - - case s_http_host: - if (IS_HOST_CHAR(ch)) { - return s_http_host; - } - - /* FALLTHROUGH */ - case s_http_host_v6_end: - if (ch == ':') { - return s_http_host_port_start; - } - - break; - - case s_http_host_v6: - if (ch == ']') { - return s_http_host_v6_end; - } - - /* FALLTHROUGH */ - case s_http_host_v6_start: - if (IS_HEX(ch) || ch == ':' || ch == '.') { - return s_http_host_v6; - } - - break; - - case s_http_host_port: - case s_http_host_port_start: - if (IS_NUM(ch)) { - return s_http_host_port; - } - - break; - - default: - break; - } - return s_http_host_dead; -} - -static int -http_parse_host(const char * buf, struct http_parser_url *u, int found_at) { - enum http_host_state s; - - const char *p; - size_t buflen = u->field_data[UF_HOST].off + u->field_data[UF_HOST].len; - - u->field_data[UF_HOST].len = 0; - - s = found_at ? s_http_userinfo_start : s_http_host_start; - - for (p = buf + u->field_data[UF_HOST].off; p < buf + buflen; p++) { - enum http_host_state new_s = http_parse_host_char(s, *p); - - if (new_s == s_http_host_dead) { - return 1; - } - - switch(new_s) { - case s_http_host: - if (s != s_http_host) { - u->field_data[UF_HOST].off = p - buf; - } - u->field_data[UF_HOST].len++; - break; - - case s_http_host_v6: - if (s != s_http_host_v6) { - u->field_data[UF_HOST].off = p - buf; - } - u->field_data[UF_HOST].len++; - break; - - case s_http_host_port: - if (s != s_http_host_port) { - u->field_data[UF_PORT].off = p - buf; - u->field_data[UF_PORT].len = 0; - u->field_set |= (1 << UF_PORT); - } - u->field_data[UF_PORT].len++; - break; - - case s_http_userinfo: - if (s != s_http_userinfo) { - u->field_data[UF_USERINFO].off = p - buf ; - u->field_data[UF_USERINFO].len = 0; - u->field_set |= (1 << UF_USERINFO); - } - u->field_data[UF_USERINFO].len++; - break; - - default: - break; - } - s = new_s; - } - - /* Make sure we don't end somewhere unexpected */ - switch (s) { - case s_http_host_start: - case s_http_host_v6_start: - case s_http_host_v6: - case s_http_host_port_start: - case s_http_userinfo: - case s_http_userinfo_start: - return 1; - default: - break; - } - - return 0; -} - -int -http_parser_parse_url(const char *buf, size_t buflen, int is_connect, - struct http_parser_url *u) -{ - enum state s; - const char *p; - enum http_parser_url_fields uf, old_uf; - int found_at = 0; - - u->port = u->field_set = 0; - s = is_connect ? s_req_server_start : s_req_spaces_before_url; - uf = old_uf = UF_MAX; - - for (p = buf; p < buf + buflen; p++) { - s = parse_url_char(s, *p); - - /* Figure out the next field that we're operating on */ - switch (s) { - case s_dead: - return 1; - - /* Skip delimeters */ - case s_req_schema_slash: - case s_req_schema_slash_slash: - case s_req_server_start: - case s_req_query_string_start: - case s_req_fragment_start: - continue; - - case s_req_schema: - uf = UF_SCHEMA; - break; - - case s_req_server_with_at: - found_at = 1; - - /* FALLTROUGH */ - case s_req_server: - uf = UF_HOST; - break; - - case s_req_path: - uf = UF_PATH; - break; - - case s_req_query_string: - uf = UF_QUERY; - break; - - case s_req_fragment: - uf = UF_FRAGMENT; - break; - - default: - assert(!"Unexpected state"); - return 1; - } - - /* Nothing's changed; soldier on */ - if (uf == old_uf) { - u->field_data[uf].len++; - continue; - } - - u->field_data[uf].off = p - buf; - u->field_data[uf].len = 1; - - u->field_set |= (1 << uf); - old_uf = uf; - } - - /* host must be present if there is a schema */ - /* parsing http:///toto will fail */ - if ((u->field_set & ((1 << UF_SCHEMA) | (1 << UF_HOST))) != 0) { - if (http_parse_host(buf, u, found_at) != 0) { - return 1; - } - } - - /* CONNECT requests can only contain "hostname:port" */ - if (is_connect && u->field_set != ((1 << UF_HOST)|(1 << UF_PORT))) { - return 1; - } - - if (u->field_set & (1 << UF_PORT)) { - /* Don't bother with endp; we've already validated the string */ - unsigned long v = strtoul(buf + u->field_data[UF_PORT].off, NULL, 10); - - /* Ports have a max value of 2^16 */ - if (v > 0xffff) { - return 1; - } - - u->port = (uint16_t) v; - } - - return 0; -} - -void -http_parser_pause(http_parser *parser, int paused) { - /* Users should only be pausing/unpausing a parser that is not in an error - * state. In non-debug builds, there's not much that we can do about this - * other than ignore it. - */ - if (HTTP_PARSER_ERRNO(parser) == HPE_OK || - HTTP_PARSER_ERRNO(parser) == HPE_PAUSED) { - SET_ERRNO((paused) ? HPE_PAUSED : HPE_OK); - } else { - assert(0 && "Attempting to pause parser in error state"); - } -} - -int -http_body_is_final(const struct http_parser *parser) { - return parser->state == s_message_done; -} diff --git a/ext/http-parser/http_parser.gyp b/ext/http-parser/http_parser.gyp deleted file mode 100644 index ef34ecae..00000000 --- a/ext/http-parser/http_parser.gyp +++ /dev/null @@ -1,111 +0,0 @@ -# This file is used with the GYP meta build system. -# http://code.google.com/p/gyp/ -# To build try this: -# svn co http://gyp.googlecode.com/svn/trunk gyp -# ./gyp/gyp -f make --depth=`pwd` http_parser.gyp -# ./out/Debug/test -{ - 'target_defaults': { - 'default_configuration': 'Debug', - 'configurations': { - # TODO: hoist these out and put them somewhere common, because - # RuntimeLibrary MUST MATCH across the entire project - 'Debug': { - 'defines': [ 'DEBUG', '_DEBUG' ], - 'cflags': [ '-Wall', '-Wextra', '-O0', '-g', '-ftrapv' ], - 'msvs_settings': { - 'VCCLCompilerTool': { - 'RuntimeLibrary': 1, # static debug - }, - }, - }, - 'Release': { - 'defines': [ 'NDEBUG' ], - 'cflags': [ '-Wall', '-Wextra', '-O3' ], - 'msvs_settings': { - 'VCCLCompilerTool': { - 'RuntimeLibrary': 0, # static release - }, - }, - } - }, - 'msvs_settings': { - 'VCCLCompilerTool': { - }, - 'VCLibrarianTool': { - }, - 'VCLinkerTool': { - 'GenerateDebugInformation': 'true', - }, - }, - 'conditions': [ - ['OS == "win"', { - 'defines': [ - 'WIN32' - ], - }] - ], - }, - - 'targets': [ - { - 'target_name': 'http_parser', - 'type': 'static_library', - 'include_dirs': [ '.' ], - 'direct_dependent_settings': { - 'defines': [ 'HTTP_PARSER_STRICT=0' ], - 'include_dirs': [ '.' ], - }, - 'defines': [ 'HTTP_PARSER_STRICT=0' ], - 'sources': [ './http_parser.c', ], - 'conditions': [ - ['OS=="win"', { - 'msvs_settings': { - 'VCCLCompilerTool': { - # Compile as C++. http_parser.c is actually C99, but C++ is - # close enough in this case. - 'CompileAs': 2, - }, - }, - }] - ], - }, - - { - 'target_name': 'http_parser_strict', - 'type': 'static_library', - 'include_dirs': [ '.' ], - 'direct_dependent_settings': { - 'defines': [ 'HTTP_PARSER_STRICT=1' ], - 'include_dirs': [ '.' ], - }, - 'defines': [ 'HTTP_PARSER_STRICT=1' ], - 'sources': [ './http_parser.c', ], - 'conditions': [ - ['OS=="win"', { - 'msvs_settings': { - 'VCCLCompilerTool': { - # Compile as C++. http_parser.c is actually C99, but C++ is - # close enough in this case. - 'CompileAs': 2, - }, - }, - }] - ], - }, - - { - 'target_name': 'test-nonstrict', - 'type': 'executable', - 'dependencies': [ 'http_parser' ], - 'sources': [ 'test.c' ] - }, - - { - 'target_name': 'test-strict', - 'type': 'executable', - 'dependencies': [ 'http_parser_strict' ], - 'sources': [ 'test.c' ] - } - ] -} diff --git a/ext/http-parser/http_parser.h b/ext/http-parser/http_parser.h deleted file mode 100644 index a992c742..00000000 --- a/ext/http-parser/http_parser.h +++ /dev/null @@ -1,304 +0,0 @@ -/* Copyright Joyent, Inc. and other Node contributors. All rights reserved. - * - * Permission is hereby granted, free of charge, to any person obtaining a copy - * of this software and associated documentation files (the "Software"), to - * deal in the Software without restriction, including without limitation the - * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or - * sell copies of the Software, and to permit persons to whom the Software is - * furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE - * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER - * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING - * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS - * IN THE SOFTWARE. - */ -#ifndef http_parser_h -#define http_parser_h -#ifdef __cplusplus -extern "C" { -#endif - -#define HTTP_PARSER_VERSION_MAJOR 2 -#define HTTP_PARSER_VERSION_MINOR 0 - -#include -#if defined(_WIN32) && !defined(__MINGW32__) && (!defined(_MSC_VER) || _MSC_VER<1600) -#include -#include -typedef __int8 int8_t; -typedef unsigned __int8 uint8_t; -typedef __int16 int16_t; -typedef unsigned __int16 uint16_t; -typedef __int32 int32_t; -typedef unsigned __int32 uint32_t; -typedef __int64 int64_t; -typedef unsigned __int64 uint64_t; -#else -#include -#endif - -/* Compile with -DHTTP_PARSER_STRICT=0 to make less checks, but run - * faster - */ -#ifndef HTTP_PARSER_STRICT -# define HTTP_PARSER_STRICT 1 -#endif - -/* Maximium header size allowed */ -#define HTTP_MAX_HEADER_SIZE (80*1024) - - -typedef struct http_parser http_parser; -typedef struct http_parser_settings http_parser_settings; - - -/* Callbacks should return non-zero to indicate an error. The parser will - * then halt execution. - * - * The one exception is on_headers_complete. In a HTTP_RESPONSE parser - * returning '1' from on_headers_complete will tell the parser that it - * should not expect a body. This is used when receiving a response to a - * HEAD request which may contain 'Content-Length' or 'Transfer-Encoding: - * chunked' headers that indicate the presence of a body. - * - * http_data_cb does not return data chunks. It will be call arbitrarally - * many times for each string. E.G. you might get 10 callbacks for "on_url" - * each providing just a few characters more data. - */ -typedef int (*http_data_cb) (http_parser*, const char *at, size_t length); -typedef int (*http_cb) (http_parser*); - - -/* Request Methods */ -#define HTTP_METHOD_MAP(XX) \ - XX(0, DELETE, DELETE) \ - XX(1, GET, GET) \ - XX(2, HEAD, HEAD) \ - XX(3, POST, POST) \ - XX(4, PUT, PUT) \ - /* pathological */ \ - XX(5, CONNECT, CONNECT) \ - XX(6, OPTIONS, OPTIONS) \ - XX(7, TRACE, TRACE) \ - /* webdav */ \ - XX(8, COPY, COPY) \ - XX(9, LOCK, LOCK) \ - XX(10, MKCOL, MKCOL) \ - XX(11, MOVE, MOVE) \ - XX(12, PROPFIND, PROPFIND) \ - XX(13, PROPPATCH, PROPPATCH) \ - XX(14, SEARCH, SEARCH) \ - XX(15, UNLOCK, UNLOCK) \ - /* subversion */ \ - XX(16, REPORT, REPORT) \ - XX(17, MKACTIVITY, MKACTIVITY) \ - XX(18, CHECKOUT, CHECKOUT) \ - XX(19, MERGE, MERGE) \ - /* upnp */ \ - XX(20, MSEARCH, M-SEARCH) \ - XX(21, NOTIFY, NOTIFY) \ - XX(22, SUBSCRIBE, SUBSCRIBE) \ - XX(23, UNSUBSCRIBE, UNSUBSCRIBE) \ - /* RFC-5789 */ \ - XX(24, PATCH, PATCH) \ - XX(25, PURGE, PURGE) \ - -enum http_method - { -#define XX(num, name, string) HTTP_##name = num, - HTTP_METHOD_MAP(XX) -#undef XX - }; - - -enum http_parser_type { HTTP_REQUEST, HTTP_RESPONSE, HTTP_BOTH }; - - -/* Flag values for http_parser.flags field */ -enum flags - { F_CHUNKED = 1 << 0 - , F_CONNECTION_KEEP_ALIVE = 1 << 1 - , F_CONNECTION_CLOSE = 1 << 2 - , F_TRAILING = 1 << 3 - , F_UPGRADE = 1 << 4 - , F_SKIPBODY = 1 << 5 - }; - - -/* Map for errno-related constants - * - * The provided argument should be a macro that takes 2 arguments. - */ -#define HTTP_ERRNO_MAP(XX) \ - /* No error */ \ - XX(OK, "success") \ - \ - /* Callback-related errors */ \ - XX(CB_message_begin, "the on_message_begin callback failed") \ - XX(CB_status_complete, "the on_status_complete callback failed") \ - XX(CB_url, "the on_url callback failed") \ - XX(CB_header_field, "the on_header_field callback failed") \ - XX(CB_header_value, "the on_header_value callback failed") \ - XX(CB_headers_complete, "the on_headers_complete callback failed") \ - XX(CB_body, "the on_body callback failed") \ - XX(CB_message_complete, "the on_message_complete callback failed") \ - \ - /* Parsing-related errors */ \ - XX(INVALID_EOF_STATE, "stream ended at an unexpected time") \ - XX(HEADER_OVERFLOW, \ - "too many header bytes seen; overflow detected") \ - XX(CLOSED_CONNECTION, \ - "data received after completed connection: close message") \ - XX(INVALID_VERSION, "invalid HTTP version") \ - XX(INVALID_STATUS, "invalid HTTP status code") \ - XX(INVALID_METHOD, "invalid HTTP method") \ - XX(INVALID_URL, "invalid URL") \ - XX(INVALID_HOST, "invalid host") \ - XX(INVALID_PORT, "invalid port") \ - XX(INVALID_PATH, "invalid path") \ - XX(INVALID_QUERY_STRING, "invalid query string") \ - XX(INVALID_FRAGMENT, "invalid fragment") \ - XX(LF_EXPECTED, "LF character expected") \ - XX(INVALID_HEADER_TOKEN, "invalid character in header") \ - XX(INVALID_CONTENT_LENGTH, \ - "invalid character in content-length header") \ - XX(INVALID_CHUNK_SIZE, \ - "invalid character in chunk size header") \ - XX(INVALID_CONSTANT, "invalid constant string") \ - XX(INVALID_INTERNAL_STATE, "encountered unexpected internal state")\ - XX(STRICT, "strict mode assertion failed") \ - XX(PAUSED, "parser is paused") \ - XX(UNKNOWN, "an unknown error occurred") - - -/* Define HPE_* values for each errno value above */ -#define HTTP_ERRNO_GEN(n, s) HPE_##n, -enum http_errno { - HTTP_ERRNO_MAP(HTTP_ERRNO_GEN) -}; -#undef HTTP_ERRNO_GEN - - -/* Get an http_errno value from an http_parser */ -#define HTTP_PARSER_ERRNO(p) ((enum http_errno) (p)->http_errno) - - -struct http_parser { - /** PRIVATE **/ - unsigned char type : 2; /* enum http_parser_type */ - unsigned char flags : 6; /* F_* values from 'flags' enum; semi-public */ - unsigned char state; /* enum state from http_parser.c */ - unsigned char header_state; /* enum header_state from http_parser.c */ - unsigned char index; /* index into current matcher */ - - uint32_t nread; /* # bytes read in various scenarios */ - uint64_t content_length; /* # bytes in body (0 if no Content-Length header) */ - - /** READ-ONLY **/ - unsigned short http_major; - unsigned short http_minor; - unsigned short status_code; /* responses only */ - unsigned char method; /* requests only */ - unsigned char http_errno : 7; - - /* 1 = Upgrade header was present and the parser has exited because of that. - * 0 = No upgrade header present. - * Should be checked when http_parser_execute() returns in addition to - * error checking. - */ - unsigned char upgrade : 1; - - /** PUBLIC **/ - void *data; /* A pointer to get hook to the "connection" or "socket" object */ -}; - - -struct http_parser_settings { - http_cb on_message_begin; - http_data_cb on_url; - http_cb on_status_complete; - http_data_cb on_header_field; - http_data_cb on_header_value; - http_cb on_headers_complete; - http_data_cb on_body; - http_cb on_message_complete; -}; - - -enum http_parser_url_fields - { UF_SCHEMA = 0 - , UF_HOST = 1 - , UF_PORT = 2 - , UF_PATH = 3 - , UF_QUERY = 4 - , UF_FRAGMENT = 5 - , UF_USERINFO = 6 - , UF_MAX = 7 - }; - - -/* Result structure for http_parser_parse_url(). - * - * Callers should index into field_data[] with UF_* values iff field_set - * has the relevant (1 << UF_*) bit set. As a courtesy to clients (and - * because we probably have padding left over), we convert any port to - * a uint16_t. - */ -struct http_parser_url { - uint16_t field_set; /* Bitmask of (1 << UF_*) values */ - uint16_t port; /* Converted UF_PORT string */ - - struct { - uint16_t off; /* Offset into buffer in which field starts */ - uint16_t len; /* Length of run in buffer */ - } field_data[UF_MAX]; -}; - - -void http_parser_init(http_parser *parser, enum http_parser_type type); - - -size_t http_parser_execute(http_parser *parser, - const http_parser_settings *settings, - const char *data, - size_t len); - - -/* If http_should_keep_alive() in the on_headers_complete or - * on_message_complete callback returns 0, then this should be - * the last message on the connection. - * If you are the server, respond with the "Connection: close" header. - * If you are the client, close the connection. - */ -int http_should_keep_alive(const http_parser *parser); - -/* Returns a string version of the HTTP method. */ -const char *http_method_str(enum http_method m); - -/* Return a string name of the given error */ -const char *http_errno_name(enum http_errno err); - -/* Return a string description of the given error */ -const char *http_errno_description(enum http_errno err); - -/* Parse a URL; return nonzero on failure */ -int http_parser_parse_url(const char *buf, size_t buflen, - int is_connect, - struct http_parser_url *u); - -/* Pause or un-pause the parser; a nonzero value pauses */ -void http_parser_pause(http_parser *parser, int paused); - -/* Checks if this is the final chunk of the body. */ -int http_body_is_final(const http_parser *parser); - -#ifdef __cplusplus -} -#endif -#endif diff --git a/ext/http-parser/test.c b/ext/http-parser/test.c deleted file mode 100644 index 83723b7f..00000000 --- a/ext/http-parser/test.c +++ /dev/null @@ -1,3425 +0,0 @@ -/* Copyright Joyent, Inc. and other Node contributors. All rights reserved. - * - * Permission is hereby granted, free of charge, to any person obtaining a copy - * of this software and associated documentation files (the "Software"), to - * deal in the Software without restriction, including without limitation the - * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or - * sell copies of the Software, and to permit persons to whom the Software is - * furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE - * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER - * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING - * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS - * IN THE SOFTWARE. - */ -#include "http_parser.h" -#include -#include -#include -#include /* rand */ -#include -#include - -#undef TRUE -#define TRUE 1 -#undef FALSE -#define FALSE 0 - -#define MAX_HEADERS 13 -#define MAX_ELEMENT_SIZE 2048 - -#define MIN(a,b) ((a) < (b) ? (a) : (b)) - -static http_parser *parser; - -struct message { - const char *name; // for debugging purposes - const char *raw; - enum http_parser_type type; - enum http_method method; - int status_code; - char request_path[MAX_ELEMENT_SIZE]; - char request_url[MAX_ELEMENT_SIZE]; - char fragment[MAX_ELEMENT_SIZE]; - char query_string[MAX_ELEMENT_SIZE]; - char body[MAX_ELEMENT_SIZE]; - size_t body_size; - const char *host; - const char *userinfo; - uint16_t port; - int num_headers; - enum { NONE=0, FIELD, VALUE } last_header_element; - char headers [MAX_HEADERS][2][MAX_ELEMENT_SIZE]; - int should_keep_alive; - - const char *upgrade; // upgraded body - - unsigned short http_major; - unsigned short http_minor; - - int message_begin_cb_called; - int headers_complete_cb_called; - int message_complete_cb_called; - int message_complete_on_eof; - int body_is_final; -}; - -static int currently_parsing_eof; - -static struct message messages[5]; -static int num_messages; -static http_parser_settings *current_pause_parser; - -/* * R E Q U E S T S * */ -const struct message requests[] = -#define CURL_GET 0 -{ {.name= "curl get" - ,.type= HTTP_REQUEST - ,.raw= "GET /test HTTP/1.1\r\n" - "User-Agent: curl/7.18.0 (i486-pc-linux-gnu) libcurl/7.18.0 OpenSSL/0.9.8g zlib/1.2.3.3 libidn/1.1\r\n" - "Host: 0.0.0.0=5000\r\n" - "Accept: */*\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/test" - ,.request_url= "/test" - ,.num_headers= 3 - ,.headers= - { { "User-Agent", "curl/7.18.0 (i486-pc-linux-gnu) libcurl/7.18.0 OpenSSL/0.9.8g zlib/1.2.3.3 libidn/1.1" } - , { "Host", "0.0.0.0=5000" } - , { "Accept", "*/*" } - } - ,.body= "" - } - -#define FIREFOX_GET 1 -, {.name= "firefox get" - ,.type= HTTP_REQUEST - ,.raw= "GET /favicon.ico HTTP/1.1\r\n" - "Host: 0.0.0.0=5000\r\n" - "User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9) Gecko/2008061015 Firefox/3.0\r\n" - "Accept: text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8\r\n" - "Accept-Language: en-us,en;q=0.5\r\n" - "Accept-Encoding: gzip,deflate\r\n" - "Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7\r\n" - "Keep-Alive: 300\r\n" - "Connection: keep-alive\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/favicon.ico" - ,.request_url= "/favicon.ico" - ,.num_headers= 8 - ,.headers= - { { "Host", "0.0.0.0=5000" } - , { "User-Agent", "Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9) Gecko/2008061015 Firefox/3.0" } - , { "Accept", "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8" } - , { "Accept-Language", "en-us,en;q=0.5" } - , { "Accept-Encoding", "gzip,deflate" } - , { "Accept-Charset", "ISO-8859-1,utf-8;q=0.7,*;q=0.7" } - , { "Keep-Alive", "300" } - , { "Connection", "keep-alive" } - } - ,.body= "" - } - -#define DUMBFUCK 2 -, {.name= "dumbfuck" - ,.type= HTTP_REQUEST - ,.raw= "GET /dumbfuck HTTP/1.1\r\n" - "aaaaaaaaaaaaa:++++++++++\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/dumbfuck" - ,.request_url= "/dumbfuck" - ,.num_headers= 1 - ,.headers= - { { "aaaaaaaaaaaaa", "++++++++++" } - } - ,.body= "" - } - -#define FRAGMENT_IN_URI 3 -, {.name= "fragment in url" - ,.type= HTTP_REQUEST - ,.raw= "GET /forums/1/topics/2375?page=1#posts-17408 HTTP/1.1\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "page=1" - ,.fragment= "posts-17408" - ,.request_path= "/forums/1/topics/2375" - /* XXX request url does include fragment? */ - ,.request_url= "/forums/1/topics/2375?page=1#posts-17408" - ,.num_headers= 0 - ,.body= "" - } - -#define GET_NO_HEADERS_NO_BODY 4 -, {.name= "get no headers no body" - ,.type= HTTP_REQUEST - ,.raw= "GET /get_no_headers_no_body/world HTTP/1.1\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE /* would need Connection: close */ - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/get_no_headers_no_body/world" - ,.request_url= "/get_no_headers_no_body/world" - ,.num_headers= 0 - ,.body= "" - } - -#define GET_ONE_HEADER_NO_BODY 5 -, {.name= "get one header no body" - ,.type= HTTP_REQUEST - ,.raw= "GET /get_one_header_no_body HTTP/1.1\r\n" - "Accept: */*\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE /* would need Connection: close */ - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/get_one_header_no_body" - ,.request_url= "/get_one_header_no_body" - ,.num_headers= 1 - ,.headers= - { { "Accept" , "*/*" } - } - ,.body= "" - } - -#define GET_FUNKY_CONTENT_LENGTH 6 -, {.name= "get funky content length body hello" - ,.type= HTTP_REQUEST - ,.raw= "GET /get_funky_content_length_body_hello HTTP/1.0\r\n" - "conTENT-Length: 5\r\n" - "\r\n" - "HELLO" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 0 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/get_funky_content_length_body_hello" - ,.request_url= "/get_funky_content_length_body_hello" - ,.num_headers= 1 - ,.headers= - { { "conTENT-Length" , "5" } - } - ,.body= "HELLO" - } - -#define POST_IDENTITY_BODY_WORLD 7 -, {.name= "post identity body world" - ,.type= HTTP_REQUEST - ,.raw= "POST /post_identity_body_world?q=search#hey HTTP/1.1\r\n" - "Accept: */*\r\n" - "Transfer-Encoding: identity\r\n" - "Content-Length: 5\r\n" - "\r\n" - "World" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_POST - ,.query_string= "q=search" - ,.fragment= "hey" - ,.request_path= "/post_identity_body_world" - ,.request_url= "/post_identity_body_world?q=search#hey" - ,.num_headers= 3 - ,.headers= - { { "Accept", "*/*" } - , { "Transfer-Encoding", "identity" } - , { "Content-Length", "5" } - } - ,.body= "World" - } - -#define POST_CHUNKED_ALL_YOUR_BASE 8 -, {.name= "post - chunked body: all your base are belong to us" - ,.type= HTTP_REQUEST - ,.raw= "POST /post_chunked_all_your_base HTTP/1.1\r\n" - "Transfer-Encoding: chunked\r\n" - "\r\n" - "1e\r\nall your base are belong to us\r\n" - "0\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_POST - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/post_chunked_all_your_base" - ,.request_url= "/post_chunked_all_your_base" - ,.num_headers= 1 - ,.headers= - { { "Transfer-Encoding" , "chunked" } - } - ,.body= "all your base are belong to us" - } - -#define TWO_CHUNKS_MULT_ZERO_END 9 -, {.name= "two chunks ; triple zero ending" - ,.type= HTTP_REQUEST - ,.raw= "POST /two_chunks_mult_zero_end HTTP/1.1\r\n" - "Transfer-Encoding: chunked\r\n" - "\r\n" - "5\r\nhello\r\n" - "6\r\n world\r\n" - "000\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_POST - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/two_chunks_mult_zero_end" - ,.request_url= "/two_chunks_mult_zero_end" - ,.num_headers= 1 - ,.headers= - { { "Transfer-Encoding", "chunked" } - } - ,.body= "hello world" - } - -#define CHUNKED_W_TRAILING_HEADERS 10 -, {.name= "chunked with trailing headers. blech." - ,.type= HTTP_REQUEST - ,.raw= "POST /chunked_w_trailing_headers HTTP/1.1\r\n" - "Transfer-Encoding: chunked\r\n" - "\r\n" - "5\r\nhello\r\n" - "6\r\n world\r\n" - "0\r\n" - "Vary: *\r\n" - "Content-Type: text/plain\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_POST - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/chunked_w_trailing_headers" - ,.request_url= "/chunked_w_trailing_headers" - ,.num_headers= 3 - ,.headers= - { { "Transfer-Encoding", "chunked" } - , { "Vary", "*" } - , { "Content-Type", "text/plain" } - } - ,.body= "hello world" - } - -#define CHUNKED_W_BULLSHIT_AFTER_LENGTH 11 -, {.name= "with bullshit after the length" - ,.type= HTTP_REQUEST - ,.raw= "POST /chunked_w_bullshit_after_length HTTP/1.1\r\n" - "Transfer-Encoding: chunked\r\n" - "\r\n" - "5; ihatew3;whatthefuck=aretheseparametersfor\r\nhello\r\n" - "6; blahblah; blah\r\n world\r\n" - "0\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_POST - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/chunked_w_bullshit_after_length" - ,.request_url= "/chunked_w_bullshit_after_length" - ,.num_headers= 1 - ,.headers= - { { "Transfer-Encoding", "chunked" } - } - ,.body= "hello world" - } - -#define WITH_QUOTES 12 -, {.name= "with quotes" - ,.type= HTTP_REQUEST - ,.raw= "GET /with_\"stupid\"_quotes?foo=\"bar\" HTTP/1.1\r\n\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "foo=\"bar\"" - ,.fragment= "" - ,.request_path= "/with_\"stupid\"_quotes" - ,.request_url= "/with_\"stupid\"_quotes?foo=\"bar\"" - ,.num_headers= 0 - ,.headers= { } - ,.body= "" - } - -#define APACHEBENCH_GET 13 -/* The server receiving this request SHOULD NOT wait for EOF - * to know that content-length == 0. - * How to represent this in a unit test? message_complete_on_eof - * Compare with NO_CONTENT_LENGTH_RESPONSE. - */ -, {.name = "apachebench get" - ,.type= HTTP_REQUEST - ,.raw= "GET /test HTTP/1.0\r\n" - "Host: 0.0.0.0:5000\r\n" - "User-Agent: ApacheBench/2.3\r\n" - "Accept: */*\r\n\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 0 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/test" - ,.request_url= "/test" - ,.num_headers= 3 - ,.headers= { { "Host", "0.0.0.0:5000" } - , { "User-Agent", "ApacheBench/2.3" } - , { "Accept", "*/*" } - } - ,.body= "" - } - -#define QUERY_URL_WITH_QUESTION_MARK_GET 14 -/* Some clients include '?' characters in query strings. - */ -, {.name = "query url with question mark" - ,.type= HTTP_REQUEST - ,.raw= "GET /test.cgi?foo=bar?baz HTTP/1.1\r\n\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "foo=bar?baz" - ,.fragment= "" - ,.request_path= "/test.cgi" - ,.request_url= "/test.cgi?foo=bar?baz" - ,.num_headers= 0 - ,.headers= {} - ,.body= "" - } - -#define PREFIX_NEWLINE_GET 15 -/* Some clients, especially after a POST in a keep-alive connection, - * will send an extra CRLF before the next request - */ -, {.name = "newline prefix get" - ,.type= HTTP_REQUEST - ,.raw= "\r\nGET /test HTTP/1.1\r\n\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/test" - ,.request_url= "/test" - ,.num_headers= 0 - ,.headers= { } - ,.body= "" - } - -#define UPGRADE_REQUEST 16 -, {.name = "upgrade request" - ,.type= HTTP_REQUEST - ,.raw= "GET /demo HTTP/1.1\r\n" - "Host: example.com\r\n" - "Connection: Upgrade\r\n" - "Sec-WebSocket-Key2: 12998 5 Y3 1 .P00\r\n" - "Sec-WebSocket-Protocol: sample\r\n" - "Upgrade: WebSocket\r\n" - "Sec-WebSocket-Key1: 4 @1 46546xW%0l 1 5\r\n" - "Origin: http://example.com\r\n" - "\r\n" - "Hot diggity dogg" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/demo" - ,.request_url= "/demo" - ,.num_headers= 7 - ,.upgrade="Hot diggity dogg" - ,.headers= { { "Host", "example.com" } - , { "Connection", "Upgrade" } - , { "Sec-WebSocket-Key2", "12998 5 Y3 1 .P00" } - , { "Sec-WebSocket-Protocol", "sample" } - , { "Upgrade", "WebSocket" } - , { "Sec-WebSocket-Key1", "4 @1 46546xW%0l 1 5" } - , { "Origin", "http://example.com" } - } - ,.body= "" - } - -#define CONNECT_REQUEST 17 -, {.name = "connect request" - ,.type= HTTP_REQUEST - ,.raw= "CONNECT 0-home0.netscape.com:443 HTTP/1.0\r\n" - "User-agent: Mozilla/1.1N\r\n" - "Proxy-authorization: basic aGVsbG86d29ybGQ=\r\n" - "\r\n" - "some data\r\n" - "and yet even more data" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 0 - ,.method= HTTP_CONNECT - ,.query_string= "" - ,.fragment= "" - ,.request_path= "" - ,.request_url= "0-home0.netscape.com:443" - ,.num_headers= 2 - ,.upgrade="some data\r\nand yet even more data" - ,.headers= { { "User-agent", "Mozilla/1.1N" } - , { "Proxy-authorization", "basic aGVsbG86d29ybGQ=" } - } - ,.body= "" - } - -#define REPORT_REQ 18 -, {.name= "report request" - ,.type= HTTP_REQUEST - ,.raw= "REPORT /test HTTP/1.1\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_REPORT - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/test" - ,.request_url= "/test" - ,.num_headers= 0 - ,.headers= {} - ,.body= "" - } - -#define NO_HTTP_VERSION 19 -, {.name= "request with no http version" - ,.type= HTTP_REQUEST - ,.raw= "GET /\r\n" - "\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 0 - ,.http_minor= 9 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/" - ,.request_url= "/" - ,.num_headers= 0 - ,.headers= {} - ,.body= "" - } - -#define MSEARCH_REQ 20 -, {.name= "m-search request" - ,.type= HTTP_REQUEST - ,.raw= "M-SEARCH * HTTP/1.1\r\n" - "HOST: 239.255.255.250:1900\r\n" - "MAN: \"ssdp:discover\"\r\n" - "ST: \"ssdp:all\"\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_MSEARCH - ,.query_string= "" - ,.fragment= "" - ,.request_path= "*" - ,.request_url= "*" - ,.num_headers= 3 - ,.headers= { { "HOST", "239.255.255.250:1900" } - , { "MAN", "\"ssdp:discover\"" } - , { "ST", "\"ssdp:all\"" } - } - ,.body= "" - } - -#define LINE_FOLDING_IN_HEADER 21 -, {.name= "line folding in header value" - ,.type= HTTP_REQUEST - ,.raw= "GET / HTTP/1.1\r\n" - "Line1: abc\r\n" - "\tdef\r\n" - " ghi\r\n" - "\t\tjkl\r\n" - " mno \r\n" - "\t \tqrs\r\n" - "Line2: \t line2\t\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/" - ,.request_url= "/" - ,.num_headers= 2 - ,.headers= { { "Line1", "abcdefghijklmno qrs" } - , { "Line2", "line2\t" } - } - ,.body= "" - } - - -#define QUERY_TERMINATED_HOST 22 -, {.name= "host terminated by a query string" - ,.type= HTTP_REQUEST - ,.raw= "GET http://hypnotoad.org?hail=all HTTP/1.1\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "hail=all" - ,.fragment= "" - ,.request_path= "" - ,.request_url= "http://hypnotoad.org?hail=all" - ,.host= "hypnotoad.org" - ,.num_headers= 0 - ,.headers= { } - ,.body= "" - } - -#define QUERY_TERMINATED_HOSTPORT 23 -, {.name= "host:port terminated by a query string" - ,.type= HTTP_REQUEST - ,.raw= "GET http://hypnotoad.org:1234?hail=all HTTP/1.1\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "hail=all" - ,.fragment= "" - ,.request_path= "" - ,.request_url= "http://hypnotoad.org:1234?hail=all" - ,.host= "hypnotoad.org" - ,.port= 1234 - ,.num_headers= 0 - ,.headers= { } - ,.body= "" - } - -#define SPACE_TERMINATED_HOSTPORT 24 -, {.name= "host:port terminated by a space" - ,.type= HTTP_REQUEST - ,.raw= "GET http://hypnotoad.org:1234 HTTP/1.1\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "" - ,.request_url= "http://hypnotoad.org:1234" - ,.host= "hypnotoad.org" - ,.port= 1234 - ,.num_headers= 0 - ,.headers= { } - ,.body= "" - } - -#define PATCH_REQ 25 -, {.name = "PATCH request" - ,.type= HTTP_REQUEST - ,.raw= "PATCH /file.txt HTTP/1.1\r\n" - "Host: www.example.com\r\n" - "Content-Type: application/example\r\n" - "If-Match: \"e0023aa4e\"\r\n" - "Content-Length: 10\r\n" - "\r\n" - "cccccccccc" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_PATCH - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/file.txt" - ,.request_url= "/file.txt" - ,.num_headers= 4 - ,.headers= { { "Host", "www.example.com" } - , { "Content-Type", "application/example" } - , { "If-Match", "\"e0023aa4e\"" } - , { "Content-Length", "10" } - } - ,.body= "cccccccccc" - } - -#define CONNECT_CAPS_REQUEST 26 -, {.name = "connect caps request" - ,.type= HTTP_REQUEST - ,.raw= "CONNECT HOME0.NETSCAPE.COM:443 HTTP/1.0\r\n" - "User-agent: Mozilla/1.1N\r\n" - "Proxy-authorization: basic aGVsbG86d29ybGQ=\r\n" - "\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 0 - ,.method= HTTP_CONNECT - ,.query_string= "" - ,.fragment= "" - ,.request_path= "" - ,.request_url= "HOME0.NETSCAPE.COM:443" - ,.num_headers= 2 - ,.upgrade="" - ,.headers= { { "User-agent", "Mozilla/1.1N" } - , { "Proxy-authorization", "basic aGVsbG86d29ybGQ=" } - } - ,.body= "" - } - -#if !HTTP_PARSER_STRICT -#define UTF8_PATH_REQ 27 -, {.name= "utf-8 path request" - ,.type= HTTP_REQUEST - ,.raw= "GET /δ¶/δt/pope?q=1#narf HTTP/1.1\r\n" - "Host: github.com\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "q=1" - ,.fragment= "narf" - ,.request_path= "/δ¶/δt/pope" - ,.request_url= "/δ¶/δt/pope?q=1#narf" - ,.num_headers= 1 - ,.headers= { {"Host", "github.com" } - } - ,.body= "" - } - -#define HOSTNAME_UNDERSCORE 28 -, {.name = "hostname underscore" - ,.type= HTTP_REQUEST - ,.raw= "CONNECT home_0.netscape.com:443 HTTP/1.0\r\n" - "User-agent: Mozilla/1.1N\r\n" - "Proxy-authorization: basic aGVsbG86d29ybGQ=\r\n" - "\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 0 - ,.method= HTTP_CONNECT - ,.query_string= "" - ,.fragment= "" - ,.request_path= "" - ,.request_url= "home_0.netscape.com:443" - ,.num_headers= 2 - ,.upgrade="" - ,.headers= { { "User-agent", "Mozilla/1.1N" } - , { "Proxy-authorization", "basic aGVsbG86d29ybGQ=" } - } - ,.body= "" - } -#endif /* !HTTP_PARSER_STRICT */ - -/* see https://github.com/ry/http-parser/issues/47 */ -#define EAT_TRAILING_CRLF_NO_CONNECTION_CLOSE 29 -, {.name = "eat CRLF between requests, no \"Connection: close\" header" - ,.raw= "POST / HTTP/1.1\r\n" - "Host: www.example.com\r\n" - "Content-Type: application/x-www-form-urlencoded\r\n" - "Content-Length: 4\r\n" - "\r\n" - "q=42\r\n" /* note the trailing CRLF */ - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_POST - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/" - ,.request_url= "/" - ,.num_headers= 3 - ,.upgrade= 0 - ,.headers= { { "Host", "www.example.com" } - , { "Content-Type", "application/x-www-form-urlencoded" } - , { "Content-Length", "4" } - } - ,.body= "q=42" - } - -/* see https://github.com/ry/http-parser/issues/47 */ -#define EAT_TRAILING_CRLF_WITH_CONNECTION_CLOSE 30 -, {.name = "eat CRLF between requests even if \"Connection: close\" is set" - ,.raw= "POST / HTTP/1.1\r\n" - "Host: www.example.com\r\n" - "Content-Type: application/x-www-form-urlencoded\r\n" - "Content-Length: 4\r\n" - "Connection: close\r\n" - "\r\n" - "q=42\r\n" /* note the trailing CRLF */ - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE /* input buffer isn't empty when on_message_complete is called */ - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_POST - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/" - ,.request_url= "/" - ,.num_headers= 4 - ,.upgrade= 0 - ,.headers= { { "Host", "www.example.com" } - , { "Content-Type", "application/x-www-form-urlencoded" } - , { "Content-Length", "4" } - , { "Connection", "close" } - } - ,.body= "q=42" - } - -#define PURGE_REQ 31 -, {.name = "PURGE request" - ,.type= HTTP_REQUEST - ,.raw= "PURGE /file.txt HTTP/1.1\r\n" - "Host: www.example.com\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_PURGE - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/file.txt" - ,.request_url= "/file.txt" - ,.num_headers= 1 - ,.headers= { { "Host", "www.example.com" } } - ,.body= "" - } - -#define SEARCH_REQ 32 -, {.name = "SEARCH request" - ,.type= HTTP_REQUEST - ,.raw= "SEARCH / HTTP/1.1\r\n" - "Host: www.example.com\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_SEARCH - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/" - ,.request_url= "/" - ,.num_headers= 1 - ,.headers= { { "Host", "www.example.com" } } - ,.body= "" - } - -#define PROXY_WITH_BASIC_AUTH 33 -, {.name= "host:port and basic_auth" - ,.type= HTTP_REQUEST - ,.raw= "GET http://a%12:b!&*$@hypnotoad.org:1234/toto HTTP/1.1\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.fragment= "" - ,.request_path= "/toto" - ,.request_url= "http://a%12:b!&*$@hypnotoad.org:1234/toto" - ,.host= "hypnotoad.org" - ,.userinfo= "a%12:b!&*$" - ,.port= 1234 - ,.num_headers= 0 - ,.headers= { } - ,.body= "" - } - - -, {.name= NULL } /* sentinel */ -}; - -/* * R E S P O N S E S * */ -const struct message responses[] = -#define GOOGLE_301 0 -{ {.name= "google 301" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 301 Moved Permanently\r\n" - "Location: http://www.google.com/\r\n" - "Content-Type: text/html; charset=UTF-8\r\n" - "Date: Sun, 26 Apr 2009 11:11:49 GMT\r\n" - "Expires: Tue, 26 May 2009 11:11:49 GMT\r\n" - "X-$PrototypeBI-Version: 1.6.0.3\r\n" /* $ char in header field */ - "Cache-Control: public, max-age=2592000\r\n" - "Server: gws\r\n" - "Content-Length: 219 \r\n" - "\r\n" - "\n" - "301 Moved\n" - "

301 Moved

\n" - "The document has moved\n" - "here.\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 301 - ,.num_headers= 8 - ,.headers= - { { "Location", "http://www.google.com/" } - , { "Content-Type", "text/html; charset=UTF-8" } - , { "Date", "Sun, 26 Apr 2009 11:11:49 GMT" } - , { "Expires", "Tue, 26 May 2009 11:11:49 GMT" } - , { "X-$PrototypeBI-Version", "1.6.0.3" } - , { "Cache-Control", "public, max-age=2592000" } - , { "Server", "gws" } - , { "Content-Length", "219 " } - } - ,.body= "\n" - "301 Moved\n" - "

301 Moved

\n" - "The document has moved\n" - "here.\r\n" - "\r\n" - } - -#define NO_CONTENT_LENGTH_RESPONSE 1 -/* The client should wait for the server's EOF. That is, when content-length - * is not specified, and "Connection: close", the end of body is specified - * by the EOF. - * Compare with APACHEBENCH_GET - */ -, {.name= "no content-length response" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 200 OK\r\n" - "Date: Tue, 04 Aug 2009 07:59:32 GMT\r\n" - "Server: Apache\r\n" - "X-Powered-By: Servlet/2.5 JSP/2.1\r\n" - "Content-Type: text/xml; charset=utf-8\r\n" - "Connection: close\r\n" - "\r\n" - "\n" - "\n" - " \n" - " \n" - " SOAP-ENV:Client\n" - " Client Error\n" - " \n" - " \n" - "" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= TRUE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 200 - ,.num_headers= 5 - ,.headers= - { { "Date", "Tue, 04 Aug 2009 07:59:32 GMT" } - , { "Server", "Apache" } - , { "X-Powered-By", "Servlet/2.5 JSP/2.1" } - , { "Content-Type", "text/xml; charset=utf-8" } - , { "Connection", "close" } - } - ,.body= "\n" - "\n" - " \n" - " \n" - " SOAP-ENV:Client\n" - " Client Error\n" - " \n" - " \n" - "" - } - -#define NO_HEADERS_NO_BODY_404 2 -, {.name= "404 no headers no body" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 404 Not Found\r\n\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= TRUE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 404 - ,.num_headers= 0 - ,.headers= {} - ,.body_size= 0 - ,.body= "" - } - -#define NO_REASON_PHRASE 3 -, {.name= "301 no response phrase" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 301\r\n\r\n" - ,.should_keep_alive = FALSE - ,.message_complete_on_eof= TRUE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 301 - ,.num_headers= 0 - ,.headers= {} - ,.body= "" - } - -#define TRAILING_SPACE_ON_CHUNKED_BODY 4 -, {.name="200 trailing space on chunked body" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 200 OK\r\n" - "Content-Type: text/plain\r\n" - "Transfer-Encoding: chunked\r\n" - "\r\n" - "25 \r\n" - "This is the data in the first chunk\r\n" - "\r\n" - "1C\r\n" - "and this is the second one\r\n" - "\r\n" - "0 \r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 200 - ,.num_headers= 2 - ,.headers= - { {"Content-Type", "text/plain" } - , {"Transfer-Encoding", "chunked" } - } - ,.body_size = 37+28 - ,.body = - "This is the data in the first chunk\r\n" - "and this is the second one\r\n" - - } - -#define NO_CARRIAGE_RET 5 -, {.name="no carriage ret" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 200 OK\n" - "Content-Type: text/html; charset=utf-8\n" - "Connection: close\n" - "\n" - "these headers are from http://news.ycombinator.com/" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= TRUE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 200 - ,.num_headers= 2 - ,.headers= - { {"Content-Type", "text/html; charset=utf-8" } - , {"Connection", "close" } - } - ,.body= "these headers are from http://news.ycombinator.com/" - } - -#define PROXY_CONNECTION 6 -, {.name="proxy connection" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 200 OK\r\n" - "Content-Type: text/html; charset=UTF-8\r\n" - "Content-Length: 11\r\n" - "Proxy-Connection: close\r\n" - "Date: Thu, 31 Dec 2009 20:55:48 +0000\r\n" - "\r\n" - "hello world" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 200 - ,.num_headers= 4 - ,.headers= - { {"Content-Type", "text/html; charset=UTF-8" } - , {"Content-Length", "11" } - , {"Proxy-Connection", "close" } - , {"Date", "Thu, 31 Dec 2009 20:55:48 +0000"} - } - ,.body= "hello world" - } - -#define UNDERSTORE_HEADER_KEY 7 - // shown by - // curl -o /dev/null -v "http://ad.doubleclick.net/pfadx/DARTSHELLCONFIGXML;dcmt=text/xml;" -, {.name="underscore header key" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 200 OK\r\n" - "Server: DCLK-AdSvr\r\n" - "Content-Type: text/xml\r\n" - "Content-Length: 0\r\n" - "DCLK_imp: v7;x;114750856;0-0;0;17820020;0/0;21603567/21621457/1;;~okv=;dcmt=text/xml;;~cs=o\r\n\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 200 - ,.num_headers= 4 - ,.headers= - { {"Server", "DCLK-AdSvr" } - , {"Content-Type", "text/xml" } - , {"Content-Length", "0" } - , {"DCLK_imp", "v7;x;114750856;0-0;0;17820020;0/0;21603567/21621457/1;;~okv=;dcmt=text/xml;;~cs=o" } - } - ,.body= "" - } - -#define BONJOUR_MADAME_FR 8 -/* The client should not merge two headers fields when the first one doesn't - * have a value. - */ -, {.name= "bonjourmadame.fr" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.0 301 Moved Permanently\r\n" - "Date: Thu, 03 Jun 2010 09:56:32 GMT\r\n" - "Server: Apache/2.2.3 (Red Hat)\r\n" - "Cache-Control: public\r\n" - "Pragma: \r\n" - "Location: http://www.bonjourmadame.fr/\r\n" - "Vary: Accept-Encoding\r\n" - "Content-Length: 0\r\n" - "Content-Type: text/html; charset=UTF-8\r\n" - "Connection: keep-alive\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 0 - ,.status_code= 301 - ,.num_headers= 9 - ,.headers= - { { "Date", "Thu, 03 Jun 2010 09:56:32 GMT" } - , { "Server", "Apache/2.2.3 (Red Hat)" } - , { "Cache-Control", "public" } - , { "Pragma", "" } - , { "Location", "http://www.bonjourmadame.fr/" } - , { "Vary", "Accept-Encoding" } - , { "Content-Length", "0" } - , { "Content-Type", "text/html; charset=UTF-8" } - , { "Connection", "keep-alive" } - } - ,.body= "" - } - -#define RES_FIELD_UNDERSCORE 9 -/* Should handle spaces in header fields */ -, {.name= "field underscore" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 200 OK\r\n" - "Date: Tue, 28 Sep 2010 01:14:13 GMT\r\n" - "Server: Apache\r\n" - "Cache-Control: no-cache, must-revalidate\r\n" - "Expires: Mon, 26 Jul 1997 05:00:00 GMT\r\n" - ".et-Cookie: PlaxoCS=1274804622353690521; path=/; domain=.plaxo.com\r\n" - "Vary: Accept-Encoding\r\n" - "_eep-Alive: timeout=45\r\n" /* semantic value ignored */ - "_onnection: Keep-Alive\r\n" /* semantic value ignored */ - "Transfer-Encoding: chunked\r\n" - "Content-Type: text/html\r\n" - "Connection: close\r\n" - "\r\n" - "0\r\n\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 200 - ,.num_headers= 11 - ,.headers= - { { "Date", "Tue, 28 Sep 2010 01:14:13 GMT" } - , { "Server", "Apache" } - , { "Cache-Control", "no-cache, must-revalidate" } - , { "Expires", "Mon, 26 Jul 1997 05:00:00 GMT" } - , { ".et-Cookie", "PlaxoCS=1274804622353690521; path=/; domain=.plaxo.com" } - , { "Vary", "Accept-Encoding" } - , { "_eep-Alive", "timeout=45" } - , { "_onnection", "Keep-Alive" } - , { "Transfer-Encoding", "chunked" } - , { "Content-Type", "text/html" } - , { "Connection", "close" } - } - ,.body= "" - } - -#define NON_ASCII_IN_STATUS_LINE 10 -/* Should handle non-ASCII in status line */ -, {.name= "non-ASCII in status line" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 500 Oriëntatieprobleem\r\n" - "Date: Fri, 5 Nov 2010 23:07:12 GMT+2\r\n" - "Content-Length: 0\r\n" - "Connection: close\r\n" - "\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 500 - ,.num_headers= 3 - ,.headers= - { { "Date", "Fri, 5 Nov 2010 23:07:12 GMT+2" } - , { "Content-Length", "0" } - , { "Connection", "close" } - } - ,.body= "" - } - -#define HTTP_VERSION_0_9 11 -/* Should handle HTTP/0.9 */ -, {.name= "http version 0.9" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/0.9 200 OK\r\n" - "\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= TRUE - ,.http_major= 0 - ,.http_minor= 9 - ,.status_code= 200 - ,.num_headers= 0 - ,.headers= - {} - ,.body= "" - } - -#define NO_CONTENT_LENGTH_NO_TRANSFER_ENCODING_RESPONSE 12 -/* The client should wait for the server's EOF. That is, when neither - * content-length nor transfer-encoding is specified, the end of body - * is specified by the EOF. - */ -, {.name= "neither content-length nor transfer-encoding response" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 200 OK\r\n" - "Content-Type: text/plain\r\n" - "\r\n" - "hello world" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= TRUE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 200 - ,.num_headers= 1 - ,.headers= - { { "Content-Type", "text/plain" } - } - ,.body= "hello world" - } - -#define NO_BODY_HTTP10_KA_200 13 -, {.name= "HTTP/1.0 with keep-alive and EOF-terminated 200 status" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.0 200 OK\r\n" - "Connection: keep-alive\r\n" - "\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= TRUE - ,.http_major= 1 - ,.http_minor= 0 - ,.status_code= 200 - ,.num_headers= 1 - ,.headers= - { { "Connection", "keep-alive" } - } - ,.body_size= 0 - ,.body= "" - } - -#define NO_BODY_HTTP10_KA_204 14 -, {.name= "HTTP/1.0 with keep-alive and a 204 status" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.0 204 No content\r\n" - "Connection: keep-alive\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 0 - ,.status_code= 204 - ,.num_headers= 1 - ,.headers= - { { "Connection", "keep-alive" } - } - ,.body_size= 0 - ,.body= "" - } - -#define NO_BODY_HTTP11_KA_200 15 -, {.name= "HTTP/1.1 with an EOF-terminated 200 status" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 200 OK\r\n" - "\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= TRUE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 200 - ,.num_headers= 0 - ,.headers={} - ,.body_size= 0 - ,.body= "" - } - -#define NO_BODY_HTTP11_KA_204 16 -, {.name= "HTTP/1.1 with a 204 status" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 204 No content\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 204 - ,.num_headers= 0 - ,.headers={} - ,.body_size= 0 - ,.body= "" - } - -#define NO_BODY_HTTP11_NOKA_204 17 -, {.name= "HTTP/1.1 with a 204 status and keep-alive disabled" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 204 No content\r\n" - "Connection: close\r\n" - "\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 204 - ,.num_headers= 1 - ,.headers= - { { "Connection", "close" } - } - ,.body_size= 0 - ,.body= "" - } - -#define NO_BODY_HTTP11_KA_CHUNKED_200 18 -, {.name= "HTTP/1.1 with chunked endocing and a 200 response" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 200 OK\r\n" - "Transfer-Encoding: chunked\r\n" - "\r\n" - "0\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 200 - ,.num_headers= 1 - ,.headers= - { { "Transfer-Encoding", "chunked" } - } - ,.body_size= 0 - ,.body= "" - } - -#if !HTTP_PARSER_STRICT -#define SPACE_IN_FIELD_RES 19 -/* Should handle spaces in header fields */ -, {.name= "field space" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 200 OK\r\n" - "Server: Microsoft-IIS/6.0\r\n" - "X-Powered-By: ASP.NET\r\n" - "en-US Content-Type: text/xml\r\n" /* this is the problem */ - "Content-Type: text/xml\r\n" - "Content-Length: 16\r\n" - "Date: Fri, 23 Jul 2010 18:45:38 GMT\r\n" - "Connection: keep-alive\r\n" - "\r\n" - "hello" /* fake body */ - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 200 - ,.num_headers= 7 - ,.headers= - { { "Server", "Microsoft-IIS/6.0" } - , { "X-Powered-By", "ASP.NET" } - , { "en-US Content-Type", "text/xml" } - , { "Content-Type", "text/xml" } - , { "Content-Length", "16" } - , { "Date", "Fri, 23 Jul 2010 18:45:38 GMT" } - , { "Connection", "keep-alive" } - } - ,.body= "hello" - } -#endif /* !HTTP_PARSER_STRICT */ - -, {.name= NULL } /* sentinel */ -}; - -/* strnlen() is a POSIX.2008 addition. Can't rely on it being available so - * define it ourselves. - */ -size_t -strnlen(const char *s, size_t maxlen) -{ - const char *p; - - p = memchr(s, '\0', maxlen); - if (p == NULL) - return maxlen; - - return p - s; -} - -size_t -strlncat(char *dst, size_t len, const char *src, size_t n) -{ - size_t slen; - size_t dlen; - size_t rlen; - size_t ncpy; - - slen = strnlen(src, n); - dlen = strnlen(dst, len); - - if (dlen < len) { - rlen = len - dlen; - ncpy = slen < rlen ? slen : (rlen - 1); - memcpy(dst + dlen, src, ncpy); - dst[dlen + ncpy] = '\0'; - } - - assert(len > slen + dlen); - return slen + dlen; -} - -size_t -strlcat(char *dst, const char *src, size_t len) -{ - return strlncat(dst, len, src, (size_t) -1); -} - -size_t -strlncpy(char *dst, size_t len, const char *src, size_t n) -{ - size_t slen; - size_t ncpy; - - slen = strnlen(src, n); - - if (len > 0) { - ncpy = slen < len ? slen : (len - 1); - memcpy(dst, src, ncpy); - dst[ncpy] = '\0'; - } - - assert(len > slen); - return slen; -} - -size_t -strlcpy(char *dst, const char *src, size_t len) -{ - return strlncpy(dst, len, src, (size_t) -1); -} - -int -request_url_cb (http_parser *p, const char *buf, size_t len) -{ - assert(p == parser); - strlncat(messages[num_messages].request_url, - sizeof(messages[num_messages].request_url), - buf, - len); - return 0; -} - -int -status_complete_cb (http_parser *p) { - assert(p == parser); - p->data++; - return 0; -} - -int -header_field_cb (http_parser *p, const char *buf, size_t len) -{ - assert(p == parser); - struct message *m = &messages[num_messages]; - - if (m->last_header_element != FIELD) - m->num_headers++; - - strlncat(m->headers[m->num_headers-1][0], - sizeof(m->headers[m->num_headers-1][0]), - buf, - len); - - m->last_header_element = FIELD; - - return 0; -} - -int -header_value_cb (http_parser *p, const char *buf, size_t len) -{ - assert(p == parser); - struct message *m = &messages[num_messages]; - - strlncat(m->headers[m->num_headers-1][1], - sizeof(m->headers[m->num_headers-1][1]), - buf, - len); - - m->last_header_element = VALUE; - - return 0; -} - -void -check_body_is_final (const http_parser *p) -{ - if (messages[num_messages].body_is_final) { - fprintf(stderr, "\n\n *** Error http_body_is_final() should return 1 " - "on last on_body callback call " - "but it doesn't! ***\n\n"); - assert(0); - abort(); - } - messages[num_messages].body_is_final = http_body_is_final(p); -} - -int -body_cb (http_parser *p, const char *buf, size_t len) -{ - assert(p == parser); - strlncat(messages[num_messages].body, - sizeof(messages[num_messages].body), - buf, - len); - messages[num_messages].body_size += len; - check_body_is_final(p); - // printf("body_cb: '%s'\n", requests[num_messages].body); - return 0; -} - -int -count_body_cb (http_parser *p, const char *buf, size_t len) -{ - assert(p == parser); - assert(buf); - messages[num_messages].body_size += len; - check_body_is_final(p); - return 0; -} - -int -message_begin_cb (http_parser *p) -{ - assert(p == parser); - messages[num_messages].message_begin_cb_called = TRUE; - return 0; -} - -int -headers_complete_cb (http_parser *p) -{ - assert(p == parser); - messages[num_messages].method = parser->method; - messages[num_messages].status_code = parser->status_code; - messages[num_messages].http_major = parser->http_major; - messages[num_messages].http_minor = parser->http_minor; - messages[num_messages].headers_complete_cb_called = TRUE; - messages[num_messages].should_keep_alive = http_should_keep_alive(parser); - return 0; -} - -int -message_complete_cb (http_parser *p) -{ - assert(p == parser); - if (messages[num_messages].should_keep_alive != http_should_keep_alive(parser)) - { - fprintf(stderr, "\n\n *** Error http_should_keep_alive() should have same " - "value in both on_message_complete and on_headers_complete " - "but it doesn't! ***\n\n"); - assert(0); - abort(); - } - - if (messages[num_messages].body_size && - http_body_is_final(p) && - !messages[num_messages].body_is_final) - { - fprintf(stderr, "\n\n *** Error http_body_is_final() should return 1 " - "on last on_body callback call " - "but it doesn't! ***\n\n"); - assert(0); - abort(); - } - - messages[num_messages].message_complete_cb_called = TRUE; - - messages[num_messages].message_complete_on_eof = currently_parsing_eof; - - num_messages++; - return 0; -} - -/* These dontcall_* callbacks exist so that we can verify that when we're - * paused, no additional callbacks are invoked */ -int -dontcall_message_begin_cb (http_parser *p) -{ - if (p) { } // gcc - fprintf(stderr, "\n\n*** on_message_begin() called on paused parser ***\n\n"); - abort(); -} - -int -dontcall_header_field_cb (http_parser *p, const char *buf, size_t len) -{ - if (p || buf || len) { } // gcc - fprintf(stderr, "\n\n*** on_header_field() called on paused parser ***\n\n"); - abort(); -} - -int -dontcall_header_value_cb (http_parser *p, const char *buf, size_t len) -{ - if (p || buf || len) { } // gcc - fprintf(stderr, "\n\n*** on_header_value() called on paused parser ***\n\n"); - abort(); -} - -int -dontcall_request_url_cb (http_parser *p, const char *buf, size_t len) -{ - if (p || buf || len) { } // gcc - fprintf(stderr, "\n\n*** on_request_url() called on paused parser ***\n\n"); - abort(); -} - -int -dontcall_body_cb (http_parser *p, const char *buf, size_t len) -{ - if (p || buf || len) { } // gcc - fprintf(stderr, "\n\n*** on_body_cb() called on paused parser ***\n\n"); - abort(); -} - -int -dontcall_headers_complete_cb (http_parser *p) -{ - if (p) { } // gcc - fprintf(stderr, "\n\n*** on_headers_complete() called on paused " - "parser ***\n\n"); - abort(); -} - -int -dontcall_message_complete_cb (http_parser *p) -{ - if (p) { } // gcc - fprintf(stderr, "\n\n*** on_message_complete() called on paused " - "parser ***\n\n"); - abort(); -} - -static http_parser_settings settings_dontcall = - {.on_message_begin = dontcall_message_begin_cb - ,.on_header_field = dontcall_header_field_cb - ,.on_header_value = dontcall_header_value_cb - ,.on_url = dontcall_request_url_cb - ,.on_body = dontcall_body_cb - ,.on_headers_complete = dontcall_headers_complete_cb - ,.on_message_complete = dontcall_message_complete_cb - }; - -/* These pause_* callbacks always pause the parser and just invoke the regular - * callback that tracks content. Before returning, we overwrite the parser - * settings to point to the _dontcall variety so that we can verify that - * the pause actually did, you know, pause. */ -int -pause_message_begin_cb (http_parser *p) -{ - http_parser_pause(p, 1); - *current_pause_parser = settings_dontcall; - return message_begin_cb(p); -} - -int -pause_header_field_cb (http_parser *p, const char *buf, size_t len) -{ - http_parser_pause(p, 1); - *current_pause_parser = settings_dontcall; - return header_field_cb(p, buf, len); -} - -int -pause_header_value_cb (http_parser *p, const char *buf, size_t len) -{ - http_parser_pause(p, 1); - *current_pause_parser = settings_dontcall; - return header_value_cb(p, buf, len); -} - -int -pause_request_url_cb (http_parser *p, const char *buf, size_t len) -{ - http_parser_pause(p, 1); - *current_pause_parser = settings_dontcall; - return request_url_cb(p, buf, len); -} - -int -pause_body_cb (http_parser *p, const char *buf, size_t len) -{ - http_parser_pause(p, 1); - *current_pause_parser = settings_dontcall; - return body_cb(p, buf, len); -} - -int -pause_headers_complete_cb (http_parser *p) -{ - http_parser_pause(p, 1); - *current_pause_parser = settings_dontcall; - return headers_complete_cb(p); -} - -int -pause_message_complete_cb (http_parser *p) -{ - http_parser_pause(p, 1); - *current_pause_parser = settings_dontcall; - return message_complete_cb(p); -} - -static http_parser_settings settings_pause = - {.on_message_begin = pause_message_begin_cb - ,.on_header_field = pause_header_field_cb - ,.on_header_value = pause_header_value_cb - ,.on_url = pause_request_url_cb - ,.on_body = pause_body_cb - ,.on_headers_complete = pause_headers_complete_cb - ,.on_message_complete = pause_message_complete_cb - }; - -static http_parser_settings settings = - {.on_message_begin = message_begin_cb - ,.on_header_field = header_field_cb - ,.on_header_value = header_value_cb - ,.on_url = request_url_cb - ,.on_body = body_cb - ,.on_headers_complete = headers_complete_cb - ,.on_message_complete = message_complete_cb - }; - -static http_parser_settings settings_count_body = - {.on_message_begin = message_begin_cb - ,.on_header_field = header_field_cb - ,.on_header_value = header_value_cb - ,.on_url = request_url_cb - ,.on_body = count_body_cb - ,.on_headers_complete = headers_complete_cb - ,.on_message_complete = message_complete_cb - }; - -static http_parser_settings settings_null = - {.on_message_begin = 0 - ,.on_header_field = 0 - ,.on_header_value = 0 - ,.on_url = 0 - ,.on_body = 0 - ,.on_headers_complete = 0 - ,.on_message_complete = 0 - }; - -void -parser_init (enum http_parser_type type) -{ - num_messages = 0; - - assert(parser == NULL); - - parser = malloc(sizeof(http_parser)); - - http_parser_init(parser, type); - - memset(&messages, 0, sizeof messages); - -} - -void -parser_free () -{ - assert(parser); - free(parser); - parser = NULL; -} - -size_t parse (const char *buf, size_t len) -{ - size_t nparsed; - currently_parsing_eof = (len == 0); - nparsed = http_parser_execute(parser, &settings, buf, len); - return nparsed; -} - -size_t parse_count_body (const char *buf, size_t len) -{ - size_t nparsed; - currently_parsing_eof = (len == 0); - nparsed = http_parser_execute(parser, &settings_count_body, buf, len); - return nparsed; -} - -size_t parse_pause (const char *buf, size_t len) -{ - size_t nparsed; - http_parser_settings s = settings_pause; - - currently_parsing_eof = (len == 0); - current_pause_parser = &s; - nparsed = http_parser_execute(parser, current_pause_parser, buf, len); - return nparsed; -} - -static inline int -check_str_eq (const struct message *m, - const char *prop, - const char *expected, - const char *found) { - if ((expected == NULL) != (found == NULL)) { - printf("\n*** Error: %s in '%s' ***\n\n", prop, m->name); - printf("expected %s\n", (expected == NULL) ? "NULL" : expected); - printf(" found %s\n", (found == NULL) ? "NULL" : found); - return 0; - } - if (expected != NULL && 0 != strcmp(expected, found)) { - printf("\n*** Error: %s in '%s' ***\n\n", prop, m->name); - printf("expected '%s'\n", expected); - printf(" found '%s'\n", found); - return 0; - } - return 1; -} - -static inline int -check_num_eq (const struct message *m, - const char *prop, - int expected, - int found) { - if (expected != found) { - printf("\n*** Error: %s in '%s' ***\n\n", prop, m->name); - printf("expected %d\n", expected); - printf(" found %d\n", found); - return 0; - } - return 1; -} - -#define MESSAGE_CHECK_STR_EQ(expected, found, prop) \ - if (!check_str_eq(expected, #prop, expected->prop, found->prop)) return 0 - -#define MESSAGE_CHECK_NUM_EQ(expected, found, prop) \ - if (!check_num_eq(expected, #prop, expected->prop, found->prop)) return 0 - -#define MESSAGE_CHECK_URL_EQ(u, expected, found, prop, fn) \ -do { \ - char ubuf[256]; \ - \ - if ((u)->field_set & (1 << (fn))) { \ - memcpy(ubuf, (found)->request_url + (u)->field_data[(fn)].off, \ - (u)->field_data[(fn)].len); \ - ubuf[(u)->field_data[(fn)].len] = '\0'; \ - } else { \ - ubuf[0] = '\0'; \ - } \ - \ - check_str_eq(expected, #prop, expected->prop, ubuf); \ -} while(0) - -int -message_eq (int index, const struct message *expected) -{ - int i; - struct message *m = &messages[index]; - - MESSAGE_CHECK_NUM_EQ(expected, m, http_major); - MESSAGE_CHECK_NUM_EQ(expected, m, http_minor); - - if (expected->type == HTTP_REQUEST) { - MESSAGE_CHECK_NUM_EQ(expected, m, method); - } else { - MESSAGE_CHECK_NUM_EQ(expected, m, status_code); - } - - MESSAGE_CHECK_NUM_EQ(expected, m, should_keep_alive); - MESSAGE_CHECK_NUM_EQ(expected, m, message_complete_on_eof); - - assert(m->message_begin_cb_called); - assert(m->headers_complete_cb_called); - assert(m->message_complete_cb_called); - - - MESSAGE_CHECK_STR_EQ(expected, m, request_url); - - /* Check URL components; we can't do this w/ CONNECT since it doesn't - * send us a well-formed URL. - */ - if (*m->request_url && m->method != HTTP_CONNECT) { - struct http_parser_url u; - - if (http_parser_parse_url(m->request_url, strlen(m->request_url), 0, &u)) { - fprintf(stderr, "\n\n*** failed to parse URL %s ***\n\n", - m->request_url); - abort(); - } - - if (expected->host) { - MESSAGE_CHECK_URL_EQ(&u, expected, m, host, UF_HOST); - } - - if (expected->userinfo) { - MESSAGE_CHECK_URL_EQ(&u, expected, m, userinfo, UF_USERINFO); - } - - m->port = (u.field_set & (1 << UF_PORT)) ? - u.port : 0; - - MESSAGE_CHECK_URL_EQ(&u, expected, m, query_string, UF_QUERY); - MESSAGE_CHECK_URL_EQ(&u, expected, m, fragment, UF_FRAGMENT); - MESSAGE_CHECK_URL_EQ(&u, expected, m, request_path, UF_PATH); - MESSAGE_CHECK_NUM_EQ(expected, m, port); - } - - if (expected->body_size) { - MESSAGE_CHECK_NUM_EQ(expected, m, body_size); - } else { - MESSAGE_CHECK_STR_EQ(expected, m, body); - } - - MESSAGE_CHECK_NUM_EQ(expected, m, num_headers); - - int r; - for (i = 0; i < m->num_headers; i++) { - r = check_str_eq(expected, "header field", expected->headers[i][0], m->headers[i][0]); - if (!r) return 0; - r = check_str_eq(expected, "header value", expected->headers[i][1], m->headers[i][1]); - if (!r) return 0; - } - - MESSAGE_CHECK_STR_EQ(expected, m, upgrade); - - return 1; -} - -/* Given a sequence of varargs messages, return the number of them that the - * parser should successfully parse, taking into account that upgraded - * messages prevent all subsequent messages from being parsed. - */ -size_t -count_parsed_messages(const size_t nmsgs, ...) { - size_t i; - va_list ap; - - va_start(ap, nmsgs); - - for (i = 0; i < nmsgs; i++) { - struct message *m = va_arg(ap, struct message *); - - if (m->upgrade) { - va_end(ap); - return i + 1; - } - } - - va_end(ap); - return nmsgs; -} - -/* Given a sequence of bytes and the number of these that we were able to - * parse, verify that upgrade bodies are correct. - */ -void -upgrade_message_fix(char *body, const size_t nread, const size_t nmsgs, ...) { - va_list ap; - size_t i; - size_t off = 0; - - va_start(ap, nmsgs); - - for (i = 0; i < nmsgs; i++) { - struct message *m = va_arg(ap, struct message *); - - off += strlen(m->raw); - - if (m->upgrade) { - off -= strlen(m->upgrade); - - /* Check the portion of the response after its specified upgrade */ - if (!check_str_eq(m, "upgrade", body + off, body + nread)) { - abort(); - } - - /* Fix up the response so that message_eq() will verify the beginning - * of the upgrade */ - *(body + nread + strlen(m->upgrade)) = '\0'; - messages[num_messages -1 ].upgrade = body + nread; - - va_end(ap); - return; - } - } - - va_end(ap); - printf("\n\n*** Error: expected a message with upgrade ***\n"); - - abort(); -} - -static void -print_error (const char *raw, size_t error_location) -{ - fprintf(stderr, "\n*** %s ***\n\n", - http_errno_description(HTTP_PARSER_ERRNO(parser))); - - int this_line = 0, char_len = 0; - size_t i, j, len = strlen(raw), error_location_line = 0; - for (i = 0; i < len; i++) { - if (i == error_location) this_line = 1; - switch (raw[i]) { - case '\r': - char_len = 2; - fprintf(stderr, "\\r"); - break; - - case '\n': - char_len = 2; - fprintf(stderr, "\\n\n"); - - if (this_line) goto print; - - error_location_line = 0; - continue; - - default: - char_len = 1; - fputc(raw[i], stderr); - break; - } - if (!this_line) error_location_line += char_len; - } - - fprintf(stderr, "[eof]\n"); - - print: - for (j = 0; j < error_location_line; j++) { - fputc(' ', stderr); - } - fprintf(stderr, "^\n\nerror location: %u\n", (unsigned int)error_location); -} - -void -test_preserve_data (void) -{ - char my_data[] = "application-specific data"; - http_parser parser; - parser.data = my_data; - http_parser_init(&parser, HTTP_REQUEST); - if (parser.data != my_data) { - printf("\n*** parser.data not preserved accross http_parser_init ***\n\n"); - abort(); - } -} - -struct url_test { - const char *name; - const char *url; - int is_connect; - struct http_parser_url u; - int rv; -}; - -const struct url_test url_tests[] = -{ {.name="proxy request" - ,.url="http://hostname/" - ,.is_connect=0 - ,.u= - {.field_set=(1 << UF_SCHEMA) | (1 << UF_HOST) | (1 << UF_PATH) - ,.port=0 - ,.field_data= - {{ 0, 4 } /* UF_SCHEMA */ - ,{ 7, 8 } /* UF_HOST */ - ,{ 0, 0 } /* UF_PORT */ - ,{ 15, 1 } /* UF_PATH */ - ,{ 0, 0 } /* UF_QUERY */ - ,{ 0, 0 } /* UF_FRAGMENT */ - ,{ 0, 0 } /* UF_USERINFO */ - } - } - ,.rv=0 - } - -, {.name="proxy request with port" - ,.url="http://hostname:444/" - ,.is_connect=0 - ,.u= - {.field_set=(1 << UF_SCHEMA) | (1 << UF_HOST) | (1 << UF_PORT) | (1 << UF_PATH) - ,.port=444 - ,.field_data= - {{ 0, 4 } /* UF_SCHEMA */ - ,{ 7, 8 } /* UF_HOST */ - ,{ 16, 3 } /* UF_PORT */ - ,{ 19, 1 } /* UF_PATH */ - ,{ 0, 0 } /* UF_QUERY */ - ,{ 0, 0 } /* UF_FRAGMENT */ - ,{ 0, 0 } /* UF_USERINFO */ - } - } - ,.rv=0 - } - -, {.name="CONNECT request" - ,.url="hostname:443" - ,.is_connect=1 - ,.u= - {.field_set=(1 << UF_HOST) | (1 << UF_PORT) - ,.port=443 - ,.field_data= - {{ 0, 0 } /* UF_SCHEMA */ - ,{ 0, 8 } /* UF_HOST */ - ,{ 9, 3 } /* UF_PORT */ - ,{ 0, 0 } /* UF_PATH */ - ,{ 0, 0 } /* UF_QUERY */ - ,{ 0, 0 } /* UF_FRAGMENT */ - ,{ 0, 0 } /* UF_USERINFO */ - } - } - ,.rv=0 - } - -, {.name="CONNECT request but not connect" - ,.url="hostname:443" - ,.is_connect=0 - ,.rv=1 - } - -, {.name="proxy ipv6 request" - ,.url="http://[1:2::3:4]/" - ,.is_connect=0 - ,.u= - {.field_set=(1 << UF_SCHEMA) | (1 << UF_HOST) | (1 << UF_PATH) - ,.port=0 - ,.field_data= - {{ 0, 4 } /* UF_SCHEMA */ - ,{ 8, 8 } /* UF_HOST */ - ,{ 0, 0 } /* UF_PORT */ - ,{ 17, 1 } /* UF_PATH */ - ,{ 0, 0 } /* UF_QUERY */ - ,{ 0, 0 } /* UF_FRAGMENT */ - ,{ 0, 0 } /* UF_USERINFO */ - } - } - ,.rv=0 - } - -, {.name="proxy ipv6 request with port" - ,.url="http://[1:2::3:4]:67/" - ,.is_connect=0 - ,.u= - {.field_set=(1 << UF_SCHEMA) | (1 << UF_HOST) | (1 << UF_PORT) | (1 << UF_PATH) - ,.port=67 - ,.field_data= - {{ 0, 4 } /* UF_SCHEMA */ - ,{ 8, 8 } /* UF_HOST */ - ,{ 18, 2 } /* UF_PORT */ - ,{ 20, 1 } /* UF_PATH */ - ,{ 0, 0 } /* UF_QUERY */ - ,{ 0, 0 } /* UF_FRAGMENT */ - ,{ 0, 0 } /* UF_USERINFO */ - } - } - ,.rv=0 - } - -, {.name="CONNECT ipv6 address" - ,.url="[1:2::3:4]:443" - ,.is_connect=1 - ,.u= - {.field_set=(1 << UF_HOST) | (1 << UF_PORT) - ,.port=443 - ,.field_data= - {{ 0, 0 } /* UF_SCHEMA */ - ,{ 1, 8 } /* UF_HOST */ - ,{ 11, 3 } /* UF_PORT */ - ,{ 0, 0 } /* UF_PATH */ - ,{ 0, 0 } /* UF_QUERY */ - ,{ 0, 0 } /* UF_FRAGMENT */ - ,{ 0, 0 } /* UF_USERINFO */ - } - } - ,.rv=0 - } - -, {.name="ipv4 in ipv6 address" - ,.url="http://[2001:0000:0000:0000:0000:0000:1.9.1.1]/" - ,.is_connect=0 - ,.u= - {.field_set=(1 << UF_SCHEMA) | (1 << UF_HOST) | (1 << UF_PATH) - ,.port=0 - ,.field_data= - {{ 0, 4 } /* UF_SCHEMA */ - ,{ 8, 37 } /* UF_HOST */ - ,{ 0, 0 } /* UF_PORT */ - ,{ 46, 1 } /* UF_PATH */ - ,{ 0, 0 } /* UF_QUERY */ - ,{ 0, 0 } /* UF_FRAGMENT */ - ,{ 0, 0 } /* UF_USERINFO */ - } - } - ,.rv=0 - } - -, {.name="extra ? in query string" - ,.url="http://a.tbcdn.cn/p/fp/2010c/??fp-header-min.css,fp-base-min.css," - "fp-channel-min.css,fp-product-min.css,fp-mall-min.css,fp-category-min.css," - "fp-sub-min.css,fp-gdp4p-min.css,fp-css3-min.css,fp-misc-min.css?t=20101022.css" - ,.is_connect=0 - ,.u= - {.field_set=(1<field_set, u->port); - for (i = 0; i < UF_MAX; i++) { - if ((u->field_set & (1 << i)) == 0) { - printf("\tfield_data[%u]: unset\n", i); - continue; - } - - printf("\tfield_data[%u]: off: %u len: %u part: \"%.*s\n\"", - i, - u->field_data[i].off, - u->field_data[i].len, - u->field_data[i].len, - url + u->field_data[i].off); - } -} - -void -test_parse_url (void) -{ - struct http_parser_url u; - const struct url_test *test; - unsigned int i; - int rv; - - for (i = 0; i < (sizeof(url_tests) / sizeof(url_tests[0])); i++) { - test = &url_tests[i]; - memset(&u, 0, sizeof(u)); - - rv = http_parser_parse_url(test->url, - strlen(test->url), - test->is_connect, - &u); - - if (test->rv == 0) { - if (rv != 0) { - printf("\n*** http_parser_parse_url(\"%s\") \"%s\" test failed, " - "unexpected rv %d ***\n\n", test->url, test->name, rv); - abort(); - } - - if (memcmp(&u, &test->u, sizeof(u)) != 0) { - printf("\n*** http_parser_parse_url(\"%s\") \"%s\" failed ***\n", - test->url, test->name); - - printf("target http_parser_url:\n"); - dump_url(test->url, &test->u); - printf("result http_parser_url:\n"); - dump_url(test->url, &u); - - abort(); - } - } else { - /* test->rv != 0 */ - if (rv == 0) { - printf("\n*** http_parser_parse_url(\"%s\") \"%s\" test failed, " - "unexpected rv %d ***\n\n", test->url, test->name, rv); - abort(); - } - } - } -} - -void -test_method_str (void) -{ - assert(0 == strcmp("GET", http_method_str(HTTP_GET))); - assert(0 == strcmp("", http_method_str(1337))); -} - -void -test_message (const struct message *message) -{ - size_t raw_len = strlen(message->raw); - size_t msg1len; - for (msg1len = 0; msg1len < raw_len; msg1len++) { - parser_init(message->type); - - size_t read; - const char *msg1 = message->raw; - const char *msg2 = msg1 + msg1len; - size_t msg2len = raw_len - msg1len; - - if (msg1len) { - read = parse(msg1, msg1len); - - if (message->upgrade && parser->upgrade) { - messages[num_messages - 1].upgrade = msg1 + read; - goto test; - } - - if (read != msg1len) { - print_error(msg1, read); - abort(); - } - } - - - read = parse(msg2, msg2len); - - if (message->upgrade && parser->upgrade) { - messages[num_messages - 1].upgrade = msg2 + read; - goto test; - } - - if (read != msg2len) { - print_error(msg2, read); - abort(); - } - - read = parse(NULL, 0); - - if (read != 0) { - print_error(message->raw, read); - abort(); - } - - test: - - if (num_messages != 1) { - printf("\n*** num_messages != 1 after testing '%s' ***\n\n", message->name); - abort(); - } - - if(!message_eq(0, message)) abort(); - - parser_free(); - } -} - -void -test_message_count_body (const struct message *message) -{ - parser_init(message->type); - - size_t read; - size_t l = strlen(message->raw); - size_t i, toread; - size_t chunk = 4024; - - for (i = 0; i < l; i+= chunk) { - toread = MIN(l-i, chunk); - read = parse_count_body(message->raw + i, toread); - if (read != toread) { - print_error(message->raw, read); - abort(); - } - } - - - read = parse_count_body(NULL, 0); - if (read != 0) { - print_error(message->raw, read); - abort(); - } - - if (num_messages != 1) { - printf("\n*** num_messages != 1 after testing '%s' ***\n\n", message->name); - abort(); - } - - if(!message_eq(0, message)) abort(); - - parser_free(); -} - -void -test_simple (const char *buf, enum http_errno err_expected) -{ - parser_init(HTTP_REQUEST); - - size_t parsed; - int pass; - enum http_errno err; - - parsed = parse(buf, strlen(buf)); - pass = (parsed == strlen(buf)); - err = HTTP_PARSER_ERRNO(parser); - parsed = parse(NULL, 0); - pass &= (parsed == 0); - - parser_free(); - - /* In strict mode, allow us to pass with an unexpected HPE_STRICT as - * long as the caller isn't expecting success. - */ -#if HTTP_PARSER_STRICT - if (err_expected != err && err_expected != HPE_OK && err != HPE_STRICT) { -#else - if (err_expected != err) { -#endif - fprintf(stderr, "\n*** test_simple expected %s, but saw %s ***\n\n%s\n", - http_errno_name(err_expected), http_errno_name(err), buf); - abort(); - } -} - -void -test_header_overflow_error (int req) -{ - http_parser parser; - http_parser_init(&parser, req ? HTTP_REQUEST : HTTP_RESPONSE); - size_t parsed; - const char *buf; - buf = req ? "GET / HTTP/1.1\r\n" : "HTTP/1.0 200 OK\r\n"; - parsed = http_parser_execute(&parser, &settings_null, buf, strlen(buf)); - assert(parsed == strlen(buf)); - - buf = "header-key: header-value\r\n"; - size_t buflen = strlen(buf); - - int i; - for (i = 0; i < 10000; i++) { - parsed = http_parser_execute(&parser, &settings_null, buf, buflen); - if (parsed != buflen) { - //fprintf(stderr, "error found on iter %d\n", i); - assert(HTTP_PARSER_ERRNO(&parser) == HPE_HEADER_OVERFLOW); - return; - } - } - - fprintf(stderr, "\n*** Error expected but none in header overflow test ***\n"); - abort(); -} - -static void -test_content_length_overflow (const char *buf, size_t buflen, int expect_ok) -{ - http_parser parser; - http_parser_init(&parser, HTTP_RESPONSE); - http_parser_execute(&parser, &settings_null, buf, buflen); - - if (expect_ok) - assert(HTTP_PARSER_ERRNO(&parser) == HPE_OK); - else - assert(HTTP_PARSER_ERRNO(&parser) == HPE_INVALID_CONTENT_LENGTH); -} - -void -test_header_content_length_overflow_error (void) -{ -#define X(size) \ - "HTTP/1.1 200 OK\r\n" \ - "Content-Length: " #size "\r\n" \ - "\r\n" - const char a[] = X(18446744073709551614); /* 2^64-2 */ - const char b[] = X(18446744073709551615); /* 2^64-1 */ - const char c[] = X(18446744073709551616); /* 2^64 */ -#undef X - test_content_length_overflow(a, sizeof(a) - 1, 1); /* expect ok */ - test_content_length_overflow(b, sizeof(b) - 1, 0); /* expect failure */ - test_content_length_overflow(c, sizeof(c) - 1, 0); /* expect failure */ -} - -void -test_chunk_content_length_overflow_error (void) -{ -#define X(size) \ - "HTTP/1.1 200 OK\r\n" \ - "Transfer-Encoding: chunked\r\n" \ - "\r\n" \ - #size "\r\n" \ - "..." - const char a[] = X(FFFFFFFFFFFFFFFE); /* 2^64-2 */ - const char b[] = X(FFFFFFFFFFFFFFFF); /* 2^64-1 */ - const char c[] = X(10000000000000000); /* 2^64 */ -#undef X - test_content_length_overflow(a, sizeof(a) - 1, 1); /* expect ok */ - test_content_length_overflow(b, sizeof(b) - 1, 0); /* expect failure */ - test_content_length_overflow(c, sizeof(c) - 1, 0); /* expect failure */ -} - -void -test_no_overflow_long_body (int req, size_t length) -{ - http_parser parser; - http_parser_init(&parser, req ? HTTP_REQUEST : HTTP_RESPONSE); - size_t parsed; - size_t i; - char buf1[3000]; - size_t buf1len = sprintf(buf1, "%s\r\nConnection: Keep-Alive\r\nContent-Length: %lu\r\n\r\n", - req ? "POST / HTTP/1.0" : "HTTP/1.0 200 OK", (unsigned long)length); - parsed = http_parser_execute(&parser, &settings_null, buf1, buf1len); - if (parsed != buf1len) - goto err; - - for (i = 0; i < length; i++) { - char foo = 'a'; - parsed = http_parser_execute(&parser, &settings_null, &foo, 1); - if (parsed != 1) - goto err; - } - - parsed = http_parser_execute(&parser, &settings_null, buf1, buf1len); - if (parsed != buf1len) goto err; - return; - - err: - fprintf(stderr, - "\n*** error in test_no_overflow_long_body %s of length %lu ***\n", - req ? "REQUEST" : "RESPONSE", - (unsigned long)length); - abort(); -} - -void -test_multiple3 (const struct message *r1, const struct message *r2, const struct message *r3) -{ - int message_count = count_parsed_messages(3, r1, r2, r3); - - char total[ strlen(r1->raw) - + strlen(r2->raw) - + strlen(r3->raw) - + 1 - ]; - total[0] = '\0'; - - strcat(total, r1->raw); - strcat(total, r2->raw); - strcat(total, r3->raw); - - parser_init(r1->type); - - size_t read; - - read = parse(total, strlen(total)); - - if (parser->upgrade) { - upgrade_message_fix(total, read, 3, r1, r2, r3); - goto test; - } - - if (read != strlen(total)) { - print_error(total, read); - abort(); - } - - read = parse(NULL, 0); - - if (read != 0) { - print_error(total, read); - abort(); - } - -test: - - if (message_count != num_messages) { - fprintf(stderr, "\n\n*** Parser didn't see 3 messages only %d *** \n", num_messages); - abort(); - } - - if (!message_eq(0, r1)) abort(); - if (message_count > 1 && !message_eq(1, r2)) abort(); - if (message_count > 2 && !message_eq(2, r3)) abort(); - - parser_free(); -} - -/* SCAN through every possible breaking to make sure the - * parser can handle getting the content in any chunks that - * might come from the socket - */ -void -test_scan (const struct message *r1, const struct message *r2, const struct message *r3) -{ - char total[80*1024] = "\0"; - char buf1[80*1024] = "\0"; - char buf2[80*1024] = "\0"; - char buf3[80*1024] = "\0"; - - strcat(total, r1->raw); - strcat(total, r2->raw); - strcat(total, r3->raw); - - size_t read; - - int total_len = strlen(total); - - int total_ops = 2 * (total_len - 1) * (total_len - 2) / 2; - int ops = 0 ; - - size_t buf1_len, buf2_len, buf3_len; - int message_count = count_parsed_messages(3, r1, r2, r3); - - int i,j,type_both; - for (type_both = 0; type_both < 2; type_both ++ ) { - for (j = 2; j < total_len; j ++ ) { - for (i = 1; i < j; i ++ ) { - - if (ops % 1000 == 0) { - printf("\b\b\b\b%3.0f%%", 100 * (float)ops /(float)total_ops); - fflush(stdout); - } - ops += 1; - - parser_init(type_both ? HTTP_BOTH : r1->type); - - buf1_len = i; - strlncpy(buf1, sizeof(buf1), total, buf1_len); - buf1[buf1_len] = 0; - - buf2_len = j - i; - strlncpy(buf2, sizeof(buf1), total+i, buf2_len); - buf2[buf2_len] = 0; - - buf3_len = total_len - j; - strlncpy(buf3, sizeof(buf1), total+j, buf3_len); - buf3[buf3_len] = 0; - - read = parse(buf1, buf1_len); - - if (parser->upgrade) goto test; - - if (read != buf1_len) { - print_error(buf1, read); - goto error; - } - - read += parse(buf2, buf2_len); - - if (parser->upgrade) goto test; - - if (read != buf1_len + buf2_len) { - print_error(buf2, read); - goto error; - } - - read += parse(buf3, buf3_len); - - if (parser->upgrade) goto test; - - if (read != buf1_len + buf2_len + buf3_len) { - print_error(buf3, read); - goto error; - } - - parse(NULL, 0); - -test: - if (parser->upgrade) { - upgrade_message_fix(total, read, 3, r1, r2, r3); - } - - if (message_count != num_messages) { - fprintf(stderr, "\n\nParser didn't see %d messages only %d\n", - message_count, num_messages); - goto error; - } - - if (!message_eq(0, r1)) { - fprintf(stderr, "\n\nError matching messages[0] in test_scan.\n"); - goto error; - } - - if (message_count > 1 && !message_eq(1, r2)) { - fprintf(stderr, "\n\nError matching messages[1] in test_scan.\n"); - goto error; - } - - if (message_count > 2 && !message_eq(2, r3)) { - fprintf(stderr, "\n\nError matching messages[2] in test_scan.\n"); - goto error; - } - - parser_free(); - } - } - } - puts("\b\b\b\b100%"); - return; - - error: - fprintf(stderr, "i=%d j=%d\n", i, j); - fprintf(stderr, "buf1 (%u) %s\n\n", (unsigned int)buf1_len, buf1); - fprintf(stderr, "buf2 (%u) %s\n\n", (unsigned int)buf2_len , buf2); - fprintf(stderr, "buf3 (%u) %s\n", (unsigned int)buf3_len, buf3); - abort(); -} - -// user required to free the result -// string terminated by \0 -char * -create_large_chunked_message (int body_size_in_kb, const char* headers) -{ - int i; - size_t wrote = 0; - size_t headers_len = strlen(headers); - size_t bufsize = headers_len + (5+1024+2)*body_size_in_kb + 6; - char * buf = malloc(bufsize); - - memcpy(buf, headers, headers_len); - wrote += headers_len; - - for (i = 0; i < body_size_in_kb; i++) { - // write 1kb chunk into the body. - memcpy(buf + wrote, "400\r\n", 5); - wrote += 5; - memset(buf + wrote, 'C', 1024); - wrote += 1024; - strcpy(buf + wrote, "\r\n"); - wrote += 2; - } - - memcpy(buf + wrote, "0\r\n\r\n", 6); - wrote += 6; - assert(wrote == bufsize); - - return buf; -} - -void -test_status_complete (void) -{ - parser_init(HTTP_RESPONSE); - parser->data = 0; - http_parser_settings settings = settings_null; - settings.on_status_complete = status_complete_cb; - - char *response = "don't mind me, just a simple response"; - http_parser_execute(parser, &settings, response, strlen(response)); - assert(parser->data == (void*)0); // the status_complete callback was never called - assert(parser->http_errno == HPE_INVALID_CONSTANT); // the errno for an invalid status line -} - -/* Verify that we can pause parsing at any of the bytes in the - * message and still get the result that we're expecting. */ -void -test_message_pause (const struct message *msg) -{ - char *buf = (char*) msg->raw; - size_t buflen = strlen(msg->raw); - size_t nread; - - parser_init(msg->type); - - do { - nread = parse_pause(buf, buflen); - - // We can only set the upgrade buffer once we've gotten our message - // completion callback. - if (messages[0].message_complete_cb_called && - msg->upgrade && - parser->upgrade) { - messages[0].upgrade = buf + nread; - goto test; - } - - if (nread < buflen) { - - // Not much do to if we failed a strict-mode check - if (HTTP_PARSER_ERRNO(parser) == HPE_STRICT) { - parser_free(); - return; - } - - assert (HTTP_PARSER_ERRNO(parser) == HPE_PAUSED); - } - - buf += nread; - buflen -= nread; - http_parser_pause(parser, 0); - } while (buflen > 0); - - nread = parse_pause(NULL, 0); - assert (nread == 0); - -test: - if (num_messages != 1) { - printf("\n*** num_messages != 1 after testing '%s' ***\n\n", msg->name); - abort(); - } - - if(!message_eq(0, msg)) abort(); - - parser_free(); -} - -int -main (void) -{ - parser = NULL; - int i, j, k; - int request_count; - int response_count; - - printf("sizeof(http_parser) = %u\n", (unsigned int)sizeof(http_parser)); - - for (request_count = 0; requests[request_count].name; request_count++); - for (response_count = 0; responses[response_count].name; response_count++); - - //// API - test_preserve_data(); - test_parse_url(); - test_method_str(); - - //// OVERFLOW CONDITIONS - - test_header_overflow_error(HTTP_REQUEST); - test_no_overflow_long_body(HTTP_REQUEST, 1000); - test_no_overflow_long_body(HTTP_REQUEST, 100000); - - test_header_overflow_error(HTTP_RESPONSE); - test_no_overflow_long_body(HTTP_RESPONSE, 1000); - test_no_overflow_long_body(HTTP_RESPONSE, 100000); - - test_header_content_length_overflow_error(); - test_chunk_content_length_overflow_error(); - - //// RESPONSES - - for (i = 0; i < response_count; i++) { - test_message(&responses[i]); - } - - for (i = 0; i < response_count; i++) { - test_message_pause(&responses[i]); - } - - for (i = 0; i < response_count; i++) { - if (!responses[i].should_keep_alive) continue; - for (j = 0; j < response_count; j++) { - if (!responses[j].should_keep_alive) continue; - for (k = 0; k < response_count; k++) { - test_multiple3(&responses[i], &responses[j], &responses[k]); - } - } - } - - test_message_count_body(&responses[NO_HEADERS_NO_BODY_404]); - test_message_count_body(&responses[TRAILING_SPACE_ON_CHUNKED_BODY]); - - // test very large chunked response - { - char * msg = create_large_chunked_message(31337, - "HTTP/1.0 200 OK\r\n" - "Transfer-Encoding: chunked\r\n" - "Content-Type: text/plain\r\n" - "\r\n"); - struct message large_chunked = - {.name= "large chunked" - ,.type= HTTP_RESPONSE - ,.raw= msg - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 0 - ,.status_code= 200 - ,.num_headers= 2 - ,.headers= - { { "Transfer-Encoding", "chunked" } - , { "Content-Type", "text/plain" } - } - ,.body_size= 31337*1024 - }; - test_message_count_body(&large_chunked); - free(msg); - } - - - - printf("response scan 1/2 "); - test_scan( &responses[TRAILING_SPACE_ON_CHUNKED_BODY] - , &responses[NO_BODY_HTTP10_KA_204] - , &responses[NO_REASON_PHRASE] - ); - - printf("response scan 2/2 "); - test_scan( &responses[BONJOUR_MADAME_FR] - , &responses[UNDERSTORE_HEADER_KEY] - , &responses[NO_CARRIAGE_RET] - ); - - puts("responses okay"); - - - /// REQUESTS - - test_simple("hello world", HPE_INVALID_METHOD); - test_simple("GET / HTP/1.1\r\n\r\n", HPE_INVALID_VERSION); - - - test_simple("ASDF / HTTP/1.1\r\n\r\n", HPE_INVALID_METHOD); - test_simple("PROPPATCHA / HTTP/1.1\r\n\r\n", HPE_INVALID_METHOD); - test_simple("GETA / HTTP/1.1\r\n\r\n", HPE_INVALID_METHOD); - - // Well-formed but incomplete - test_simple("GET / HTTP/1.1\r\n" - "Content-Type: text/plain\r\n" - "Content-Length: 6\r\n" - "\r\n" - "fooba", - HPE_OK); - - static const char *all_methods[] = { - "DELETE", - "GET", - "HEAD", - "POST", - "PUT", - //"CONNECT", //CONNECT can't be tested like other methods, it's a tunnel - "OPTIONS", - "TRACE", - "COPY", - "LOCK", - "MKCOL", - "MOVE", - "PROPFIND", - "PROPPATCH", - "UNLOCK", - "REPORT", - "MKACTIVITY", - "CHECKOUT", - "MERGE", - "M-SEARCH", - "NOTIFY", - "SUBSCRIBE", - "UNSUBSCRIBE", - "PATCH", - 0 }; - const char **this_method; - for (this_method = all_methods; *this_method; this_method++) { - char buf[200]; - sprintf(buf, "%s / HTTP/1.1\r\n\r\n", *this_method); - test_simple(buf, HPE_OK); - } - - static const char *bad_methods[] = { - "C******", - "M****", - 0 }; - for (this_method = bad_methods; *this_method; this_method++) { - char buf[200]; - sprintf(buf, "%s / HTTP/1.1\r\n\r\n", *this_method); - test_simple(buf, HPE_UNKNOWN); - } - - const char *dumbfuck2 = - "GET / HTTP/1.1\r\n" - "X-SSL-Bullshit: -----BEGIN CERTIFICATE-----\r\n" - "\tMIIFbTCCBFWgAwIBAgICH4cwDQYJKoZIhvcNAQEFBQAwcDELMAkGA1UEBhMCVUsx\r\n" - "\tETAPBgNVBAoTCGVTY2llbmNlMRIwEAYDVQQLEwlBdXRob3JpdHkxCzAJBgNVBAMT\r\n" - "\tAkNBMS0wKwYJKoZIhvcNAQkBFh5jYS1vcGVyYXRvckBncmlkLXN1cHBvcnQuYWMu\r\n" - "\tdWswHhcNMDYwNzI3MTQxMzI4WhcNMDcwNzI3MTQxMzI4WjBbMQswCQYDVQQGEwJV\r\n" - "\tSzERMA8GA1UEChMIZVNjaWVuY2UxEzARBgNVBAsTCk1hbmNoZXN0ZXIxCzAJBgNV\r\n" - "\tBAcTmrsogriqMWLAk1DMRcwFQYDVQQDEw5taWNoYWVsIHBhcmQYJKoZIhvcNAQEB\r\n" - "\tBQADggEPADCCAQoCggEBANPEQBgl1IaKdSS1TbhF3hEXSl72G9J+WC/1R64fAcEF\r\n" - "\tW51rEyFYiIeZGx/BVzwXbeBoNUK41OK65sxGuflMo5gLflbwJtHBRIEKAfVVp3YR\r\n" - "\tgW7cMA/s/XKgL1GEC7rQw8lIZT8RApukCGqOVHSi/F1SiFlPDxuDfmdiNzL31+sL\r\n" - "\t0iwHDdNkGjy5pyBSB8Y79dsSJtCW/iaLB0/n8Sj7HgvvZJ7x0fr+RQjYOUUfrePP\r\n" - "\tu2MSpFyf+9BbC/aXgaZuiCvSR+8Snv3xApQY+fULK/xY8h8Ua51iXoQ5jrgu2SqR\r\n" - "\twgA7BUi3G8LFzMBl8FRCDYGUDy7M6QaHXx1ZWIPWNKsCAwEAAaOCAiQwggIgMAwG\r\n" - "\tA1UdEwEB/wQCMAAwEQYJYIZIAYb4QgHTTPAQDAgWgMA4GA1UdDwEB/wQEAwID6DAs\r\n" - "\tBglghkgBhvhCAQ0EHxYdVUsgZS1TY2llbmNlIFVzZXIgQ2VydGlmaWNhdGUwHQYD\r\n" - "\tVR0OBBYEFDTt/sf9PeMaZDHkUIldrDYMNTBZMIGaBgNVHSMEgZIwgY+AFAI4qxGj\r\n" - "\tloCLDdMVKwiljjDastqooXSkcjBwMQswCQYDVQQGEwJVSzERMA8GA1UEChMIZVNj\r\n" - "\taWVuY2UxEjAQBgNVBAsTCUF1dGhvcml0eTELMAkGA1UEAxMCQ0ExLTArBgkqhkiG\r\n" - "\t9w0BCQEWHmNhLW9wZXJhdG9yQGdyaWQtc3VwcG9ydC5hYy51a4IBADApBgNVHRIE\r\n" - "\tIjAggR5jYS1vcGVyYXRvckBncmlkLXN1cHBvcnQuYWMudWswGQYDVR0gBBIwEDAO\r\n" - "\tBgwrBgEEAdkvAQEBAQYwPQYJYIZIAYb4QgEEBDAWLmh0dHA6Ly9jYS5ncmlkLXN1\r\n" - "\tcHBvcnQuYWMudmT4sopwqlBWsvcHViL2NybC9jYWNybC5jcmwwPQYJYIZIAYb4QgEDBDAWLmh0\r\n" - "\tdHA6Ly9jYS5ncmlkLXN1cHBvcnQuYWMudWsvcHViL2NybC9jYWNybC5jcmwwPwYD\r\n" - "\tVR0fBDgwNjA0oDKgMIYuaHR0cDovL2NhLmdyaWQt5hYy51ay9wdWIv\r\n" - "\tY3JsL2NhY3JsLmNybDANBgkqhkiG9w0BAQUFAAOCAQEAS/U4iiooBENGW/Hwmmd3\r\n" - "\tXCy6Zrt08YjKCzGNjorT98g8uGsqYjSxv/hmi0qlnlHs+k/3Iobc3LjS5AMYr5L8\r\n" - "\tUO7OSkgFFlLHQyC9JzPfmLCAugvzEbyv4Olnsr8hbxF1MbKZoQxUZtMVu29wjfXk\r\n" - "\thTeApBv7eaKCWpSp7MCbvgzm74izKhu3vlDk9w6qVrxePfGgpKPqfHiOoGhFnbTK\r\n" - "\twTC6o2xq5y0qZ03JonF7OJspEd3I5zKY3E+ov7/ZhW6DqT8UFvsAdjvQbXyhV8Eu\r\n" - "\tYhixw1aKEPzNjNowuIseVogKOLXxWI5vAi5HgXdS0/ES5gDGsABo4fqovUKlgop3\r\n" - "\tRA==\r\n" - "\t-----END CERTIFICATE-----\r\n" - "\r\n"; - test_simple(dumbfuck2, HPE_OK); - -#if 0 - // NOTE(Wed Nov 18 11:57:27 CET 2009) this seems okay. we just read body - // until EOF. - // - // no content-length - // error if there is a body without content length - const char *bad_get_no_headers_no_body = "GET /bad_get_no_headers_no_body/world HTTP/1.1\r\n" - "Accept: */*\r\n" - "\r\n" - "HELLO"; - test_simple(bad_get_no_headers_no_body, 0); -#endif - /* TODO sending junk and large headers gets rejected */ - - - /* check to make sure our predefined requests are okay */ - for (i = 0; requests[i].name; i++) { - test_message(&requests[i]); - } - - for (i = 0; i < request_count; i++) { - test_message_pause(&requests[i]); - } - - for (i = 0; i < request_count; i++) { - if (!requests[i].should_keep_alive) continue; - for (j = 0; j < request_count; j++) { - if (!requests[j].should_keep_alive) continue; - for (k = 0; k < request_count; k++) { - test_multiple3(&requests[i], &requests[j], &requests[k]); - } - } - } - - printf("request scan 1/4 "); - test_scan( &requests[GET_NO_HEADERS_NO_BODY] - , &requests[GET_ONE_HEADER_NO_BODY] - , &requests[GET_NO_HEADERS_NO_BODY] - ); - - printf("request scan 2/4 "); - test_scan( &requests[POST_CHUNKED_ALL_YOUR_BASE] - , &requests[POST_IDENTITY_BODY_WORLD] - , &requests[GET_FUNKY_CONTENT_LENGTH] - ); - - printf("request scan 3/4 "); - test_scan( &requests[TWO_CHUNKS_MULT_ZERO_END] - , &requests[CHUNKED_W_TRAILING_HEADERS] - , &requests[CHUNKED_W_BULLSHIT_AFTER_LENGTH] - ); - - printf("request scan 4/4 "); - test_scan( &requests[QUERY_URL_WITH_QUESTION_MARK_GET] - , &requests[PREFIX_NEWLINE_GET ] - , &requests[CONNECT_REQUEST] - ); - - test_status_complete(); - - puts("requests okay"); - - return 0; -} diff --git a/node/Http.cpp b/node/Http.cpp deleted file mode 100644 index 86ae2d25..00000000 --- a/node/Http.cpp +++ /dev/null @@ -1,323 +0,0 @@ -/* - * ZeroTier One - Global Peer to Peer Ethernet - * Copyright (C) 2012-2013 ZeroTier Networks LLC - * - * This program is free software: you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation, either version 3 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program. If not, see . - * - * -- - * - * ZeroTier may be used and distributed under the terms of the GPLv3, which - * are available at: http://www.gnu.org/licenses/gpl-3.0.html - * - * If you would like to embed ZeroTier into a commercial application or - * redistribute it in a modified binary form, please contact ZeroTier Networks - * LLC. Start here: http://www.zerotier.com/ - */ - -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -#ifndef _WIN32 -#include -#endif - -#include "Http.hpp" -#include "Utils.hpp" -#include "InetAddress.hpp" - -static http_parser_settings _http_parser_settings; - -namespace ZeroTier { - -static bool _sendAll(int fd,const void *buf,unsigned int len) -{ - for(;;) { - int n = (int)::send(fd,buf,len,0); - if ((n < 0)&&(errno == EINTR)) - continue; - return (n == (int)len); - } -} - -const std::map Http::EMPTY_HEADERS; - -Http::Request::Request( - Http::Method m, - const std::string &url, - const std::map &rh, - const std::string &rb, - bool (*handler)(Request *,void *,const std::string &,int,const std::map &,const std::string &), - void *arg) : - _url(url), - _requestHeaders(rh), - _handler(handler), - _arg(arg), - _method(m), - _fd(0) -{ - _http_parser_settings.on_message_begin = &Http::Request::_http_on_message_begin; - _http_parser_settings.on_url = &Http::Request::_http_on_url; - _http_parser_settings.on_status_complete = &Http::Request::_http_on_status_complete; - _http_parser_settings.on_header_field = &Http::Request::_http_on_header_field; - _http_parser_settings.on_header_value = &Http::Request::_http_on_header_value; - _http_parser_settings.on_headers_complete = &Http::Request::_http_on_headers_complete; - _http_parser_settings.on_body = &Http::Request::_http_on_body; - _http_parser_settings.on_message_complete = &Http::Request::_http_on_message_complete; - - start(); -} - -Http::Request::~Request() -{ - if (_fd > 0) - ::close(_fd); - join(); -} - -void Http::Request::main() - throw() -{ - char buf[131072]; - - try { - http_parser_init(&_parser,HTTP_RESPONSE); - _parser.data = this; - - http_parser_url urlParsed; - if (http_parser_parse_url(_url.c_str(),_url.length(),0,&urlParsed)) { - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"URL parse error"); - return; - } - if (!(urlParsed.field_set & (1 << UF_SCHEMA))) { - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"URL specifies no schema"); - return; - } - - std::string schema(_url.substr(urlParsed.field_data[UF_SCHEMA].off,urlParsed.field_data[UF_SCHEMA].len)); - - if (schema == "file") { - const std::string filePath(_url.substr(urlParsed.field_data[UF_PATH].off,urlParsed.field_data[UF_PATH].len)); - - uint64_t lm = Utils::getLastModified(filePath.c_str()); - if (lm) { - const std::map::const_iterator ifModSince(_requestHeaders.find("If-Modified-Since")); - if ((ifModSince != _requestHeaders.end())&&(ifModSince->second.length())) { - uint64_t t64 = Utils::fromRfc1123(ifModSince->second); - if ((t64)&&(lm > t64)) { - suicidalThread = !_handler(this,_arg,_url,304,_responseHeaders,""); - return; - } - } - - if (Utils::readFile(filePath.c_str(),_responseBody)) { - _responseHeaders["Last-Modified"] = Utils::toRfc1123(lm); - suicidalThread = !_handler(this,_arg,_url,200,_responseHeaders,_responseBody); - return; - } - } - - suicidalThread = !_handler(this,_arg,_url,404,_responseHeaders,"file not found or not readable"); - return; - } else if (schema == "http") { - if (!(urlParsed.field_set & (1 << UF_HOST))) { - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"URL contains no host"); - return; - } - std::string host(_url.substr(urlParsed.field_data[UF_HOST].off,urlParsed.field_data[UF_HOST].len)); - - std::list v4,v6; - { - struct addrinfo *res = (struct addrinfo *)0; - if (!getaddrinfo(host.c_str(),(const char *)0,(const struct addrinfo *)0,&res)) { - struct addrinfo *p = res; - do { - if (p->ai_family == AF_INET) - v4.push_back(InetAddress(p->ai_addr)); - else if (p->ai_family == AF_INET6) - v6.push_back(InetAddress(p->ai_addr)); - } while ((p = p->ai_next)); - freeaddrinfo(res); - } - } - - std::list *addrList; - if (v4.empty()&&v6.empty()) { - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"could not find address for host in URL"); - return; - } else if (v4.empty()) { - addrList = &v6; - } else { - addrList = &v4; - } - InetAddress *addr; - { - addrList->sort(); - addrList->unique(); - unsigned int i = 0,k = 0; - k = rand() % addrList->size(); - std::list::iterator a(addrList->begin()); - while (i++ != k) ++a; - addr = &(*a); - } - - int remotePort = ((urlParsed.field_set & (1 << UF_PORT))&&(urlParsed.port)) ? (int)urlParsed.port : (int)80; - if ((remotePort <= 0)||(remotePort > 0xffff)) { - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"URL port out of range"); - return; - } - addr->setPort(remotePort); - - _fd = socket(addr->isV6() ? AF_INET6 : AF_INET,SOCK_STREAM,0); - if (_fd <= 0) { - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"could not open socket"); - return; - } - - for(;;) { - if (connect(_fd,addr->saddr(),addr->saddrLen())) { - if (errno == EINTR) - continue; - ::close(_fd); _fd = 0; - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"connection failed to remote host"); - return; - } else break; - } - - const char *mstr = "GET"; - switch(_method) { - case HTTP_METHOD_HEAD: mstr = "HEAD"; break; - default: break; - } - int mlen = (int)snprintf(buf,sizeof(buf),"%s %s HTTP/1.1\r\nAccept-Encoding: \r\nHost: %s\r\n",mstr,_url.substr(urlParsed.field_data[UF_PATH].off,urlParsed.field_data[UF_PATH].len).c_str(),host.c_str()); - if (mlen >= (int)sizeof(buf)) { - ::close(_fd); _fd = 0; - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"URL too long"); - return; - } - if (!_sendAll(_fd,buf,mlen)) { - ::close(_fd); _fd = 0; - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"write error"); - return; - } - - for(std::map::const_iterator rh(_requestHeaders.begin());rh!=_requestHeaders.end();++rh) { - mlen = (int)snprintf(buf,sizeof(buf),"%s: %s\r\n",rh->first.c_str(),rh->second.c_str()); - if (mlen >= (int)sizeof(buf)) { - ::close(_fd); _fd = 0; - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"header too long"); - return; - } - if (!_sendAll(_fd,buf,mlen)) { - ::close(_fd); _fd = 0; - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"write error"); - return; - } - } - - if (!_sendAll(_fd,"\r\n",2)) { - ::close(_fd); _fd = 0; - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"write error"); - return; - } - - _responseStatusCode = 0; - _messageComplete = false; - for(;;) { - mlen = (int)::recv(_fd,buf,sizeof(buf),0); - if (mlen < 0) { - if (errno != EINTR) - break; - else continue; - } - if (((int)http_parser_execute(&_parser,&_http_parser_settings,buf,mlen) != mlen)||(_parser.upgrade)) { - ::close(_fd); _fd = 0; - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"invalid HTTP response from server"); - return; - } - if (_messageComplete) { - ::close(_fd); _fd = 0; - suicidalThread = !_handler(this,_arg,_url,_responseStatusCode,_responseHeaders,_responseBody); - return; - } - } - - ::close(_fd); _fd = 0; - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"empty HTTP response from server"); - return; - } else { - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"only 'file' and 'http' methods are supported"); - return; - } - } catch ( ... ) { - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"unexpected exception retrieving URL"); - return; - } -} - -int Http::Request::_http_on_message_begin(http_parser *parser) -{ - return 0; -} -int Http::Request::_http_on_url(http_parser *parser,const char *data,size_t length) -{ - return 0; -} -int Http::Request::_http_on_status_complete(http_parser *parser) -{ - Http::Request *r = (Http::Request *)parser->data; - r->_responseStatusCode = parser->status_code; - return 0; -} -int Http::Request::_http_on_header_field(http_parser *parser,const char *data,size_t length) -{ - Http::Request *r = (Http::Request *)parser->data; - if ((r->_currentHeaderField.length())&&(r->_responseHeaders.find(r->_currentHeaderField) != r->_responseHeaders.end())) - r->_currentHeaderField.assign(""); - r->_currentHeaderField.append(data,length); - return 0; -} -int Http::Request::_http_on_header_value(http_parser *parser,const char *data,size_t length) -{ - Http::Request *r = (Http::Request *)parser->data; - if (r->_currentHeaderField.length()) - r->_responseHeaders[r->_currentHeaderField].append(data,length); - return 0; -} -int Http::Request::_http_on_headers_complete(http_parser *parser) -{ - Http::Request *r = (Http::Request *)parser->data; - return ((r->_method == Http::HTTP_METHOD_HEAD) ? 1 : 0); -} -int Http::Request::_http_on_body(http_parser *parser,const char *data,size_t length) -{ - Http::Request *r = (Http::Request *)parser->data; - r->_responseBody.append(data,length); - return 0; -} -int Http::Request::_http_on_message_complete(http_parser *parser) -{ - Http::Request *r = (Http::Request *)parser->data; - r->_messageComplete = true; - return 0; -} - -} // namespace ZeroTier diff --git a/node/Http.hpp b/node/Http.hpp deleted file mode 100644 index a099b15d..00000000 --- a/node/Http.hpp +++ /dev/null @@ -1,129 +0,0 @@ -/* - * ZeroTier One - Global Peer to Peer Ethernet - * Copyright (C) 2012-2013 ZeroTier Networks LLC - * - * This program is free software: you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation, either version 3 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program. If not, see . - * - * -- - * - * ZeroTier may be used and distributed under the terms of the GPLv3, which - * are available at: http://www.gnu.org/licenses/gpl-3.0.html - * - * If you would like to embed ZeroTier into a commercial application or - * redistribute it in a modified binary form, please contact ZeroTier Networks - * LLC. Start here: http://www.zerotier.com/ - */ - -#ifndef _ZT_HTTP_HPP -#define _ZT_HTTP_HPP - -#include -#include -#include -#include "Thread.hpp" - -#include "../ext/http-parser/http_parser.h" - -namespace ZeroTier { - -class Http -{ -public: - /** - * HTTP request methods - */ - enum Method - { - HTTP_METHOD_GET, - HTTP_METHOD_HEAD - }; - - /** - * An empty headers map for convenience - */ - static const std::map EMPTY_HEADERS; - - /** - * HTTP request - */ - class Request : protected Thread - { - public: - /** - * Create and issue an HTTP request - * - * The supplied handler is called when the request is - * complete or if an error occurs. A code of zero indicates - * that the server could not be reached, and a description - * of the error will be in 'body'. If the handler returns - * false the Request object deletes itself. Otherwise the - * object must be deleted by other code. - * - * @param m Request method - * @param url Destination URL - * @param rh Request headers - * @param rb Request body or empty string for none (currently unused) - * @param handler Request handler function - * @param arg First argument to request handler - */ - Request( - Http::Method m, - const std::string &url, - const std::map &rh, - const std::string &rb, - bool (*handler)(Request *,void *,const std::string &,int,const std::map &,const std::string &), - void *arg); - - /** - * Destruction cancels any in-progress request - */ - virtual ~Request(); - - protected: - virtual void main() - throw(); - - private: - // HTTP parser handlers - static int _http_on_message_begin(http_parser *parser); - static int _http_on_url(http_parser *parser,const char *data,size_t length); - static int _http_on_status_complete(http_parser *parser); - static int _http_on_header_field(http_parser *parser,const char *data,size_t length); - static int _http_on_header_value(http_parser *parser,const char *data,size_t length); - static int _http_on_headers_complete(http_parser *parser); - static int _http_on_body(http_parser *parser,const char *data,size_t length); - static int _http_on_message_complete(http_parser *parser); - - http_parser _parser; - std::string _url; - - std::map _requestHeaders; - std::map _responseHeaders; - - std::string _currentHeaderField; - std::string _responseBody; - - bool (*_handler)(Request *,void *,const std::string &,int,const std::map &,const std::string &); - void *_arg; - - Http::Method _method; - int _responseStatusCode; - bool _messageComplete; - volatile int _fd; - }; -}; - -} // namespace ZeroTier - -#endif diff --git a/objects.mk b/objects.mk index f5b88719..07806039 100644 --- a/objects.mk +++ b/objects.mk @@ -1,5 +1,4 @@ OBJS=\ - ext/http-parser/http_parser.o \ ext/huffandpuff/huffman.o \ ext/kissdb/kissdb.o \ ext/lz4/lz4hc.o \ @@ -10,7 +9,6 @@ OBJS=\ node/EthernetTap.o \ node/Filter.o \ node/HMAC.o \ - node/Http.o \ node/Identity.o \ node/InetAddress.o \ node/Logger.o \ -- cgit v1.2.3 From a816f564261911c168be1140e7f01def4ddfc8a7 Mon Sep 17 00:00:00 2001 From: Adam Ierymenko Date: Sat, 27 Jul 2013 14:01:19 -0400 Subject: Dump huffman, doesnt add much and complicates porting to other languages. Also fix compile error in idtool. --- ext/huffandpuff/huffman.c | 252 ---------------------------------------------- ext/huffandpuff/huffman.h | 52 ---------- idtool.cpp | 2 +- node/Utils.hpp | 37 ++----- objects.mk | 1 - selftest.cpp | 28 ------ 6 files changed, 8 insertions(+), 364 deletions(-) delete mode 100644 ext/huffandpuff/huffman.c delete mode 100644 ext/huffandpuff/huffman.h (limited to 'ext') diff --git a/ext/huffandpuff/huffman.c b/ext/huffandpuff/huffman.c deleted file mode 100644 index f4d8314a..00000000 --- a/ext/huffandpuff/huffman.c +++ /dev/null @@ -1,252 +0,0 @@ -/* - * Huffandpuff minimal Huffman coder - * - * (c)2013 Adam Ierymenko - * This code is in the public domain and is distributed with NO WARRANTY. - */ - -#include "huffman.h" - -struct _huffman_node -{ - struct _huffman_node *lr[2]; - struct _huffman_node *qprev,*qnext; - double prob; - unsigned long c; -}; - -struct _huffman_encode_table -{ - unsigned long code; - unsigned long bits; -}; - -static void _huffman_write_tree_and_make_encode_table(unsigned char *out,unsigned long *outbitctr,unsigned long outlen,struct _huffman_encode_table *et,unsigned long code,unsigned int bits,struct _huffman_node *t) -{ - struct _huffman_encode_table *eti; - unsigned int i; - unsigned long byte_index; - - byte_index = (*outbitctr)++ >> 3; - byte_index *= (byte_index < outlen); - if (t->lr[0]) { - out[byte_index] <<= 1; - _huffman_write_tree_and_make_encode_table(out,outbitctr,outlen,et,code,bits + 1,t->lr[0]); - _huffman_write_tree_and_make_encode_table(out,outbitctr,outlen,et,code | (1 << bits),bits + 1,t->lr[1]); - } else { - out[byte_index] = (out[byte_index] << 1) | 1; - for(i=0;i<9;++i) { - byte_index = (*outbitctr)++ >> 3; - if (byte_index >= outlen) return; - out[byte_index] = (out[byte_index] << 1) | ((unsigned char)((t->c >> i) & 1)); - } - eti = &(et[t->c]); - eti->code = code; - eti->bits = bits; - } -} - -static struct _huffman_node *_huffman_read_tree(const unsigned char *in,unsigned long *inbitctr,unsigned long inlen,unsigned char **heapptr,unsigned char *heapend) -{ - struct _huffman_node *n; - unsigned int i; - unsigned long byte_index; - - n = (struct _huffman_node *)(*heapptr); - *heapptr += sizeof(struct _huffman_node); - if (*heapptr > heapend) return (struct _huffman_node *)0; - - byte_index = *inbitctr >> 3; - byte_index *= (byte_index < inlen); - if (((in[byte_index] >> (~((*inbitctr)++) & 7)) & 1)) { - n->lr[0] = (struct _huffman_node *)0; - n->lr[1] = (struct _huffman_node *)0; - n->c = 0; - for(i=0;i<9;++i) { - byte_index = *inbitctr >> 3; - if (byte_index >= inlen) return (struct _huffman_node *)0; - n->c |= (((unsigned int)(in[byte_index] >> (~((*inbitctr)++) & 7))) & 1) << i; - } - } else { - n->lr[0] = _huffman_read_tree(in,inbitctr,inlen,heapptr,heapend); - n->lr[1] = _huffman_read_tree(in,inbitctr,inlen,heapptr,heapend); - if (!((n->lr[0])&&(n->lr[1]))) - return (struct _huffman_node *)0; - } - - return n; -} - -unsigned long huffman_compress(const unsigned char *in,unsigned long inlen,unsigned char *out,unsigned long outlen,void *huffheap) -{ - struct _huffman_encode_table *et,*eti; - struct _huffman_node *t,*n; - struct _huffman_node *pair[2]; - unsigned char *heapptr = (unsigned char *)huffheap; - unsigned long i,code,byte_index,outbitctr; - unsigned int bits,b; - double *counts,lowest_prob,total_symbols; - - counts = (double *)heapptr; - heapptr += (sizeof(double) * 257); - for(i=0;i<256;++i) - counts[i] = 0.0; - counts[256] = 1.0; /* one stop code at end */ - for(i=0;i 0.0) { - n = (struct _huffman_node *)heapptr; - heapptr += sizeof(struct _huffman_node); - if (t) - t->qprev = n; - n->qprev = (struct _huffman_node *)0; - n->qnext = t; - n->lr[0] = (struct _huffman_node *)0; - n->lr[1] = (struct _huffman_node *)0; - n->prob = counts[i] / total_symbols; - n->c = (unsigned int)i; - t = n; - } - } - - while (t->qnext) { - for(i=0;i<2;++i) { - lowest_prob = 1.0; - pair[i] = (struct _huffman_node *)0; - n = t; - while (n) { - if (n->prob <= lowest_prob) { - lowest_prob = n->prob; - pair[i] = n; - } - n = n->qnext; - } - if (pair[i]->qprev) - pair[i]->qprev->qnext = pair[i]->qnext; - else t = pair[i]->qnext; - if (pair[i]->qnext) - pair[i]->qnext->qprev = pair[i]->qprev; - } - n = (struct _huffman_node *)heapptr; - heapptr += sizeof(struct _huffman_node); - n->lr[0] = pair[0]; - n->lr[1] = pair[1]; - n->prob = pair[0]->prob + pair[1]->prob; - if (t) - t->qprev = n; - n->qprev = (struct _huffman_node *)0; - n->qnext = t; - t = n; - } - - et = (struct _huffman_encode_table *)heapptr; - heapptr += (sizeof(struct _huffman_encode_table) * 257); - outbitctr = 0; - _huffman_write_tree_and_make_encode_table(out,&outbitctr,outlen,et,0,0,t); - - for(i=0;icode; - bits = eti->bits; - for(b=0;b> 3; - if (byte_index >= outlen) return 0; - out[byte_index] = (out[byte_index] << 1) | (unsigned char)(code & 1); - code >>= 1; - } - } - code = et[256].code; - bits = et[256].bits; - for(b=0;b> 3; - if (byte_index >= outlen) return 0; - out[byte_index] = (out[byte_index] << 1) | (unsigned char)(code & 1); - code >>= 1; - } - - if (outbitctr > (outlen << 3)) - return 0; - else if ((outbitctr & 7)) { - out[i = (outbitctr >> 3)] <<= 8 - (outbitctr & 7); - return (i + 1); - } else return (outbitctr >> 3); -} - -unsigned long huffman_decompress(const unsigned char *in,unsigned long inlen,unsigned char *out,unsigned long outlen,void *huffheap) -{ - struct _huffman_node *t,*n; - unsigned char *heapptr = (unsigned char *)huffheap; - unsigned long inbitctr,outptr,byte_index = 0; - - inbitctr = 0; - t = _huffman_read_tree(in,&inbitctr,inlen,&heapptr,heapptr + HUFFHEAP_SIZE); - if (!t) return 0; - outptr = 0; - for(;;) { - n = t; - while (n->lr[0]) { - byte_index = inbitctr >> 3; - if (byte_index >= inlen) return 0; - n = n->lr[((unsigned long)(in[byte_index] >> (~(inbitctr++) & 7))) & 1]; - } - if (n->c == 256) return outptr; - if (outptr == outlen) return 0; - out[outptr++] = (unsigned char)n->c; - } -} - -#ifdef HUFFANDPUFF_TEST - -#include -#include -#include -#include - -#define HUFFANDPUFF_TEST_MAXLEN 1048576 -#define HUFFANDPUFF_TEST_ITER 1024 - -static unsigned char testin[HUFFANDPUFF_TEST_MAXLEN]; -static unsigned char testout[HUFFANDPUFF_TEST_MAXLEN * 2]; -static unsigned char testver[HUFFANDPUFF_TEST_MAXLEN]; -static unsigned char huffbuf[HUFFHEAP_SIZE]; - -int main(int argc,char **argv) -{ - unsigned long i,k,l,cl,dcl; - int v; - unsigned char mask; - - srand(time(0)); - - for(k=0;k - * This code is in the public domain and is distributed with NO WARRANTY. - */ - -#ifndef ____HUFFMAN_H -#define ____HUFFMAN_H - -#ifdef __cplusplus -extern "C" { -#endif - -/** - * Required size of huffheap parameter to compress and decompress - * - * Note: if you change any of the data types in the _huffman_node - * or _huffman_encode_table structs in huffman.c, this also must be - * changed. - */ -#define HUFFHEAP_SIZE ((sizeof(double) * 257) + (((sizeof(void *) * 4) + sizeof(double) + sizeof(unsigned long)) * (257 * 3)) + ((sizeof(unsigned long) + sizeof(unsigned long)) * 257)) - -/** - * Huffman encode a block of data - * - * @param in Input data - * @param inlen Input data length - * @param out Output buffer - * @param outlen Output buffer length - * @param huffheap Heap memory to use for compression (must be HUFFHEAP_SIZE in size) - * @return Size of encoded result or 0 on out buffer overrun - */ -extern unsigned long huffman_compress(const unsigned char *in,unsigned long inlen,unsigned char *out,unsigned long outlen,void *huffheap); - -/** - * Huffman decode a block of data - * - * @param in Input data - * @param inlen Length of input data - * @param out Output buffer - * @param outlen Length of output buffer - * @param huffheap Heap memory to use for decompression (must be HUFFHEAP_SIZE in size) - * @return Size of decoded result or 0 on out buffer overrun or corrupt input data - */ -extern unsigned long huffman_decompress(const unsigned char *in,unsigned long inlen,unsigned char *out,unsigned long outlen,void *huffheap); - -#ifdef __cplusplus -} -#endif - -#endif diff --git a/idtool.cpp b/idtool.cpp index 8068fc14..00511a0f 100644 --- a/idtool.cpp +++ b/idtool.cpp @@ -159,7 +159,7 @@ int main(int argc,char **argv) } std::string signature(Utils::base64Decode(argv[4],strlen(argv[4]))); - if ((signature.length() > ZEROTIER_ADDRESS_LENGTH)&&(id.verifySignature(inf.data(),inf.length(),signature))) { + if ((signature.length() > ZT_ADDRESS_LENGTH)&&(id.verifySignature(inf.data(),inf.length(),signature.data(),signature.length()))) { std::cout << argv[3] << " signature valid" << std::endl; } else { std::cerr << argv[3] << " signature check FAILED" << std::endl; diff --git a/node/Utils.hpp b/node/Utils.hpp index fbabe3d7..de7acf0e 100644 --- a/node/Utils.hpp +++ b/node/Utils.hpp @@ -42,7 +42,6 @@ #include "../ext/lz4/lz4.h" #include "../ext/lz4/lz4hc.h" -#include "../ext/huffandpuff/huffman.h" #include "Constants.hpp" @@ -176,7 +175,6 @@ public: template static inline void compress(I begin,I end,O out) { - char huffheap[HUFFHEAP_SIZE]; unsigned int bufLen = LZ4_compressBound(ZT_COMPRESSION_BLOCK_SIZE); char *buf = new char[bufLen * 2]; char *buf2 = buf + bufLen; @@ -210,16 +208,9 @@ public: continue; } - unsigned long huffCompressedLen = huffman_compress((const unsigned char *)buf2,lz4CompressedLen,(unsigned char *)buf,bufLen,huffheap); - if ((!huffCompressedLen)||((int)huffCompressedLen >= lz4CompressedLen)) { - l = hton((uint32_t)lz4CompressedLen); // lz4 only - out((const void *)&l,4); - out((const void *)buf2,(unsigned int)lz4CompressedLen); - } else { - l = hton((uint32_t)0x80000000 | (uint32_t)huffCompressedLen); // lz4 with huffman - out((const void *)&l,4); - out((const void *)buf,(unsigned int)huffCompressedLen); - } + l = hton((uint32_t)lz4CompressedLen); // lz4 only + out((const void *)&l,4); + out((const void *)buf2,(unsigned int)lz4CompressedLen); } delete [] buf; @@ -242,7 +233,6 @@ public: template static inline bool decompress(I begin,I end,O out) { - char huffheap[HUFFHEAP_SIZE]; volatile char i32c[4]; void *const i32cp = (void *)i32c; unsigned int bufLen = LZ4_compressBound(ZT_COMPRESSION_BLOCK_SIZE); @@ -279,23 +269,10 @@ public: return false; } - if ((_compressedSize & 0x80000000)) { // lz4 and huffman - unsigned long lz4CompressedSize = huffman_decompress((const unsigned char *)buf,compressedSize,(unsigned char *)buf2,bufLen,huffheap); - if (lz4CompressedSize) { - if (LZ4_uncompress_unknownOutputSize(buf2,buf,lz4CompressedSize,bufLen) != (int)originalSize) { - delete [] buf; - return false; - } else out((const void *)buf,(unsigned int)originalSize); - } else { - delete [] buf; - return false; - } - } else { // lz4 only - if (LZ4_uncompress_unknownOutputSize(buf,buf2,compressedSize,bufLen) != (int)originalSize) { - delete [] buf; - return false; - } else out((const void *)buf2,(unsigned int)originalSize); - } + if (LZ4_uncompress_unknownOutputSize(buf,buf2,compressedSize,bufLen) != (int)originalSize) { + delete [] buf; + return false; + } else out((const void *)buf2,(unsigned int)originalSize); } else { // stored if (originalSize > bufLen) { delete [] buf; diff --git a/objects.mk b/objects.mk index 07806039..6ad6f5ec 100644 --- a/objects.mk +++ b/objects.mk @@ -1,5 +1,4 @@ OBJS=\ - ext/huffandpuff/huffman.o \ ext/kissdb/kissdb.o \ ext/lz4/lz4hc.o \ ext/lz4/lz4.o \ diff --git a/selftest.cpp b/selftest.cpp index dffd22cd..4377d837 100644 --- a/selftest.cpp +++ b/selftest.cpp @@ -43,7 +43,6 @@ #include "node/HMAC.hpp" #include "node/MAC.hpp" #include "node/Peer.hpp" -#include "node/Http.hpp" #include "node/Condition.hpp" #include "node/NodeConfig.hpp" @@ -302,39 +301,12 @@ static int testOther() return 0; } -static Condition testHttpDoneCondition; - -static bool testHttpHandler(Http::Request *req,void *arg,const std::string &url,int code,const std::map &headers,const std::string &body) -{ - if (code) - std::cout << "[net] " << url << " " << code << " bytes: " << body.length() << std::endl; - else std::cout << "[net] " << url << " FAILED: " << body << std::endl; - testHttpDoneCondition.signal(); - return false; -} - -static int testNet() -{ - std::cout << "[net] GET http://www.uc.edu/" << std::endl; - new Http::Request(Http::HTTP_METHOD_GET,"http://www.uc.edu/",Http::EMPTY_HEADERS,std::string(),&testHttpHandler,(void *)0); - testHttpDoneCondition.wait(); - std::cout << "[net] GET http://zerotier.com/" << std::endl; - new Http::Request(Http::HTTP_METHOD_GET,"http://zerotier.com/",Http::EMPTY_HEADERS,std::string(),&testHttpHandler,(void *)0); - testHttpDoneCondition.wait(); - std::cout << "[net] GET http://www.google.com/" << std::endl; - new Http::Request(Http::HTTP_METHOD_GET,"http://www.google.com/",Http::EMPTY_HEADERS,std::string(),&testHttpHandler,(void *)0); - testHttpDoneCondition.wait(); - - return 0; -} - int main(int argc,char **argv) { int r = 0; srand(time(0)); - r |= testNet(); r |= testCrypto(); r |= testPacket(); r |= testOther(); -- cgit v1.2.3 From dd203f006574276be8eccedbcd1ddb8452b69eb1 Mon Sep 17 00:00:00 2001 From: Adam Ierymenko Date: Sat, 27 Jul 2013 15:45:01 -0400 Subject: Revert removal of Http. Witness my indecisiveness. --- AUTHORS.txt | 6 + ext/http-parser/AUTHORS | 41 + ext/http-parser/CONTRIBUTIONS | 4 + ext/http-parser/LICENSE-MIT | 23 + ext/http-parser/README.md | 180 ++ ext/http-parser/contrib/parsertrace.c | 156 ++ ext/http-parser/contrib/url_parser.c | 44 + ext/http-parser/http_parser.c | 2175 +++++++++++++++++++++ ext/http-parser/http_parser.gyp | 111 ++ ext/http-parser/http_parser.h | 304 +++ ext/http-parser/test.c | 3425 +++++++++++++++++++++++++++++++++ node/Http.cpp | 323 ++++ node/Http.hpp | 129 ++ objects.mk | 2 + 14 files changed, 6923 insertions(+) create mode 100644 ext/http-parser/AUTHORS create mode 100644 ext/http-parser/CONTRIBUTIONS create mode 100644 ext/http-parser/LICENSE-MIT create mode 100644 ext/http-parser/README.md create mode 100644 ext/http-parser/contrib/parsertrace.c create mode 100644 ext/http-parser/contrib/url_parser.c create mode 100644 ext/http-parser/http_parser.c create mode 100644 ext/http-parser/http_parser.gyp create mode 100644 ext/http-parser/http_parser.h create mode 100644 ext/http-parser/test.c create mode 100644 node/Http.cpp create mode 100644 node/Http.hpp (limited to 'ext') diff --git a/AUTHORS.txt b/AUTHORS.txt index f304f6af..7374b7c8 100644 --- a/AUTHORS.txt +++ b/AUTHORS.txt @@ -16,6 +16,12 @@ ZeroTier One includes or links with the following third party software: * LZ4 compression algorithm by Yann Collet (BSD license) http://code.google.com/p/lz4/ + * JsonCPP by Baptiste Lepilleur (public domain) + http://jsoncpp.sourceforge.net + + * http-parser, a simple C http parser library (MIT license) + https://github.com/joyent/http-parser + * OpenSSL libcrypto (BSD-style OpenSSL license) http://www.openssl.org/ diff --git a/ext/http-parser/AUTHORS b/ext/http-parser/AUTHORS new file mode 100644 index 00000000..92ee45ca --- /dev/null +++ b/ext/http-parser/AUTHORS @@ -0,0 +1,41 @@ +# Authors ordered by first contribution. +Ryan Dahl +Jeremy Hinegardner +Sergey Shepelev +Joe Damato +tomika +Phoenix Sol +Cliff Frey +Ewen Cheslack-Postava +Santiago Gala +Tim Becker +Jeff Terrace +Ben Noordhuis +Nathan Rajlich +Mark Nottingham +Aman Gupta +Tim Becker +Sean Cunningham +Peter Griess +Salman Haq +Cliff Frey +Jon Kolb +Fouad Mardini +Paul Querna +Felix Geisendörfer +koichik +Andre Caron +Ivo Raisr +James McLaughlin +David Gwynne +Thomas LE ROUX +Randy Rizun +Andre Louis Caron +Simon Zimmermann +Erik Dubbelboer +Martell Malone +Bertrand Paquet +BogDan Vatra +Peter Faiman +Corey Richardson +Tóth Tamás diff --git a/ext/http-parser/CONTRIBUTIONS b/ext/http-parser/CONTRIBUTIONS new file mode 100644 index 00000000..11ba31e4 --- /dev/null +++ b/ext/http-parser/CONTRIBUTIONS @@ -0,0 +1,4 @@ +Contributors must agree to the Contributor License Agreement before patches +can be accepted. + +http://spreadsheets2.google.com/viewform?hl=en&formkey=dDJXOGUwbzlYaWM4cHN1MERwQS1CSnc6MQ diff --git a/ext/http-parser/LICENSE-MIT b/ext/http-parser/LICENSE-MIT new file mode 100644 index 00000000..58010b38 --- /dev/null +++ b/ext/http-parser/LICENSE-MIT @@ -0,0 +1,23 @@ +http_parser.c is based on src/http/ngx_http_parse.c from NGINX copyright +Igor Sysoev. + +Additional changes are licensed under the same terms as NGINX and +copyright Joyent, Inc. and other Node contributors. All rights reserved. + +Permission is hereby granted, free of charge, to any person obtaining a copy +of this software and associated documentation files (the "Software"), to +deal in the Software without restriction, including without limitation the +rights to use, copy, modify, merge, publish, distribute, sublicense, and/or +sell copies of the Software, and to permit persons to whom the Software is +furnished to do so, subject to the following conditions: + +The above copyright notice and this permission notice shall be included in +all copies or substantial portions of the Software. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING +FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS +IN THE SOFTWARE. diff --git a/ext/http-parser/README.md b/ext/http-parser/README.md new file mode 100644 index 00000000..b63418af --- /dev/null +++ b/ext/http-parser/README.md @@ -0,0 +1,180 @@ +HTTP Parser +=========== + +[![Build Status](https://travis-ci.org/joyent/http-parser.png?branch=master)](https://travis-ci.org/joyent/http-parser) + +This is a parser for HTTP messages written in C. It parses both requests and +responses. The parser is designed to be used in performance HTTP +applications. It does not make any syscalls nor allocations, it does not +buffer data, it can be interrupted at anytime. Depending on your +architecture, it only requires about 40 bytes of data per message +stream (in a web server that is per connection). + +Features: + + * No dependencies + * Handles persistent streams (keep-alive). + * Decodes chunked encoding. + * Upgrade support + * Defends against buffer overflow attacks. + +The parser extracts the following information from HTTP messages: + + * Header fields and values + * Content-Length + * Request method + * Response status code + * Transfer-Encoding + * HTTP version + * Request URL + * Message body + + +Usage +----- + +One `http_parser` object is used per TCP connection. Initialize the struct +using `http_parser_init()` and set the callbacks. That might look something +like this for a request parser: + + http_parser_settings settings; + settings.on_url = my_url_callback; + settings.on_header_field = my_header_field_callback; + /* ... */ + + http_parser *parser = malloc(sizeof(http_parser)); + http_parser_init(parser, HTTP_REQUEST); + parser->data = my_socket; + +When data is received on the socket execute the parser and check for errors. + + size_t len = 80*1024, nparsed; + char buf[len]; + ssize_t recved; + + recved = recv(fd, buf, len, 0); + + if (recved < 0) { + /* Handle error. */ + } + + /* Start up / continue the parser. + * Note we pass recved==0 to signal that EOF has been recieved. + */ + nparsed = http_parser_execute(parser, &settings, buf, recved); + + if (parser->upgrade) { + /* handle new protocol */ + } else if (nparsed != recved) { + /* Handle error. Usually just close the connection. */ + } + +HTTP needs to know where the end of the stream is. For example, sometimes +servers send responses without Content-Length and expect the client to +consume input (for the body) until EOF. To tell http_parser about EOF, give +`0` as the forth parameter to `http_parser_execute()`. Callbacks and errors +can still be encountered during an EOF, so one must still be prepared +to receive them. + +Scalar valued message information such as `status_code`, `method`, and the +HTTP version are stored in the parser structure. This data is only +temporally stored in `http_parser` and gets reset on each new message. If +this information is needed later, copy it out of the structure during the +`headers_complete` callback. + +The parser decodes the transfer-encoding for both requests and responses +transparently. That is, a chunked encoding is decoded before being sent to +the on_body callback. + + +The Special Problem of Upgrade +------------------------------ + +HTTP supports upgrading the connection to a different protocol. An +increasingly common example of this is the Web Socket protocol which sends +a request like + + GET /demo HTTP/1.1 + Upgrade: WebSocket + Connection: Upgrade + Host: example.com + Origin: http://example.com + WebSocket-Protocol: sample + +followed by non-HTTP data. + +(See http://tools.ietf.org/html/draft-hixie-thewebsocketprotocol-75 for more +information the Web Socket protocol.) + +To support this, the parser will treat this as a normal HTTP message without a +body. Issuing both on_headers_complete and on_message_complete callbacks. However +http_parser_execute() will stop parsing at the end of the headers and return. + +The user is expected to check if `parser->upgrade` has been set to 1 after +`http_parser_execute()` returns. Non-HTTP data begins at the buffer supplied +offset by the return value of `http_parser_execute()`. + + +Callbacks +--------- + +During the `http_parser_execute()` call, the callbacks set in +`http_parser_settings` will be executed. The parser maintains state and +never looks behind, so buffering the data is not necessary. If you need to +save certain data for later usage, you can do that from the callbacks. + +There are two types of callbacks: + +* notification `typedef int (*http_cb) (http_parser*);` + Callbacks: on_message_begin, on_headers_complete, on_message_complete. +* data `typedef int (*http_data_cb) (http_parser*, const char *at, size_t length);` + Callbacks: (requests only) on_uri, + (common) on_header_field, on_header_value, on_body; + +Callbacks must return 0 on success. Returning a non-zero value indicates +error to the parser, making it exit immediately. + +In case you parse HTTP message in chunks (i.e. `read()` request line +from socket, parse, read half headers, parse, etc) your data callbacks +may be called more than once. Http-parser guarantees that data pointer is only +valid for the lifetime of callback. You can also `read()` into a heap allocated +buffer to avoid copying memory around if this fits your application. + +Reading headers may be a tricky task if you read/parse headers partially. +Basically, you need to remember whether last header callback was field or value +and apply following logic: + + (on_header_field and on_header_value shortened to on_h_*) + ------------------------ ------------ -------------------------------------------- + | State (prev. callback) | Callback | Description/action | + ------------------------ ------------ -------------------------------------------- + | nothing (first call) | on_h_field | Allocate new buffer and copy callback data | + | | | into it | + ------------------------ ------------ -------------------------------------------- + | value | on_h_field | New header started. | + | | | Copy current name,value buffers to headers | + | | | list and allocate new buffer for new name | + ------------------------ ------------ -------------------------------------------- + | field | on_h_field | Previous name continues. Reallocate name | + | | | buffer and append callback data to it | + ------------------------ ------------ -------------------------------------------- + | field | on_h_value | Value for current header started. Allocate | + | | | new buffer and copy callback data to it | + ------------------------ ------------ -------------------------------------------- + | value | on_h_value | Value continues. Reallocate value buffer | + | | | and append callback data to it | + ------------------------ ------------ -------------------------------------------- + + +Parsing URLs +------------ + +A simplistic zero-copy URL parser is provided as `http_parser_parse_url()`. +Users of this library may wish to use it to parse URLs constructed from +consecutive `on_url` callbacks. + +See examples of reading in headers: + +* [partial example](http://gist.github.com/155877) in C +* [from http-parser tests](http://github.com/joyent/http-parser/blob/37a0ff8/test.c#L403) in C +* [from Node library](http://github.com/joyent/node/blob/842eaf4/src/http.js#L284) in Javascript diff --git a/ext/http-parser/contrib/parsertrace.c b/ext/http-parser/contrib/parsertrace.c new file mode 100644 index 00000000..c9bc71ec --- /dev/null +++ b/ext/http-parser/contrib/parsertrace.c @@ -0,0 +1,156 @@ +/* Based on src/http/ngx_http_parse.c from NGINX copyright Igor Sysoev + * + * Additional changes are licensed under the same terms as NGINX and + * copyright Joyent, Inc. and other Node contributors. All rights reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +/* Dump what the parser finds to stdout as it happen */ + +#include "http_parser.h" +#include +#include +#include + +int on_message_begin(http_parser* _) { + (void)_; + printf("\n***MESSAGE BEGIN***\n\n"); + return 0; +} + +int on_headers_complete(http_parser* _) { + (void)_; + printf("\n***HEADERS COMPLETE***\n\n"); + return 0; +} + +int on_message_complete(http_parser* _) { + (void)_; + printf("\n***MESSAGE COMPLETE***\n\n"); + return 0; +} + +int on_url(http_parser* _, const char* at, size_t length) { + (void)_; + printf("Url: %.*s\n", (int)length, at); + return 0; +} + +int on_header_field(http_parser* _, const char* at, size_t length) { + (void)_; + printf("Header field: %.*s\n", (int)length, at); + return 0; +} + +int on_header_value(http_parser* _, const char* at, size_t length) { + (void)_; + printf("Header value: %.*s\n", (int)length, at); + return 0; +} + +int on_body(http_parser* _, const char* at, size_t length) { + (void)_; + printf("Body: %.*s\n", (int)length, at); + return 0; +} + +void usage(const char* name) { + fprintf(stderr, + "Usage: %s $type $filename\n" + " type: -x, where x is one of {r,b,q}\n" + " parses file as a Response, reQuest, or Both\n", + name); + exit(EXIT_FAILURE); +} + +int main(int argc, char* argv[]) { + enum http_parser_type file_type; + + if (argc != 3) { + usage(argv[0]); + } + + char* type = argv[1]; + if (type[0] != '-') { + usage(argv[0]); + } + + switch (type[1]) { + /* in the case of "-", type[1] will be NUL */ + case 'r': + file_type = HTTP_RESPONSE; + break; + case 'q': + file_type = HTTP_REQUEST; + break; + case 'b': + file_type = HTTP_BOTH; + break; + default: + usage(argv[0]); + } + + char* filename = argv[2]; + FILE* file = fopen(filename, "r"); + if (file == NULL) { + perror("fopen"); + return EXIT_FAILURE; + } + + fseek(file, 0, SEEK_END); + long file_length = ftell(file); + if (file_length == -1) { + perror("ftell"); + return EXIT_FAILURE; + } + fseek(file, 0, SEEK_SET); + + char* data = malloc(file_length); + if (fread(data, 1, file_length, file) != (size_t)file_length) { + fprintf(stderr, "couldn't read entire file\n"); + free(data); + return EXIT_FAILURE; + } + + http_parser_settings settings; + memset(&settings, 0, sizeof(settings)); + settings.on_message_begin = on_message_begin; + settings.on_url = on_url; + settings.on_header_field = on_header_field; + settings.on_header_value = on_header_value; + settings.on_headers_complete = on_headers_complete; + settings.on_body = on_body; + settings.on_message_complete = on_message_complete; + + http_parser parser; + http_parser_init(&parser, file_type); + size_t nparsed = http_parser_execute(&parser, &settings, data, file_length); + free(data); + + if (nparsed != (size_t)file_length) { + fprintf(stderr, + "Error: %s (%s)\n", + http_errno_description(HTTP_PARSER_ERRNO(&parser)), + http_errno_name(HTTP_PARSER_ERRNO(&parser))); + return EXIT_FAILURE; + } + + return EXIT_SUCCESS; +} diff --git a/ext/http-parser/contrib/url_parser.c b/ext/http-parser/contrib/url_parser.c new file mode 100644 index 00000000..b1f9c979 --- /dev/null +++ b/ext/http-parser/contrib/url_parser.c @@ -0,0 +1,44 @@ +#include "http_parser.h" +#include +#include + +void +dump_url (const char *url, const struct http_parser_url *u) +{ + unsigned int i; + + printf("\tfield_set: 0x%x, port: %u\n", u->field_set, u->port); + for (i = 0; i < UF_MAX; i++) { + if ((u->field_set & (1 << i)) == 0) { + printf("\tfield_data[%u]: unset\n", i); + continue; + } + + printf("\tfield_data[%u]: off: %u len: %u part: \"%.*s\n", + i, + u->field_data[i].off, + u->field_data[i].len, + u->field_data[i].len, + url + u->field_data[i].off); + } +} + +int main(int argc, char ** argv) { + if (argc != 3) { + printf("Syntax : %s connect|get url\n", argv[0]); + return 1; + } + struct http_parser_url u; + int len = strlen(argv[2]); + int connect = strcmp("connect", argv[1]) == 0 ? 1 : 0; + printf("Parsing %s, connect %d\n", argv[2], connect); + + int result = http_parser_parse_url(argv[2], len, connect, &u); + if (result != 0) { + printf("Parse error : %d\n", result); + return result; + } + printf("Parse ok, result : \n"); + dump_url(argv[2], &u); + return 0; +} \ No newline at end of file diff --git a/ext/http-parser/http_parser.c b/ext/http-parser/http_parser.c new file mode 100644 index 00000000..ed3a9232 --- /dev/null +++ b/ext/http-parser/http_parser.c @@ -0,0 +1,2175 @@ +/* Based on src/http/ngx_http_parse.c from NGINX copyright Igor Sysoev + * + * Additional changes are licensed under the same terms as NGINX and + * copyright Joyent, Inc. and other Node contributors. All rights reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ +#include "http_parser.h" +#include +#include +#include +#include +#include +#include + +#ifndef ULLONG_MAX +# define ULLONG_MAX ((uint64_t) -1) /* 2^64-1 */ +#endif + +#ifndef MIN +# define MIN(a,b) ((a) < (b) ? (a) : (b)) +#endif + +#ifndef ARRAY_SIZE +# define ARRAY_SIZE(a) (sizeof(a) / sizeof((a)[0])) +#endif + +#ifndef BIT_AT +# define BIT_AT(a, i) \ + (!!((unsigned int) (a)[(unsigned int) (i) >> 3] & \ + (1 << ((unsigned int) (i) & 7)))) +#endif + +#ifndef ELEM_AT +# define ELEM_AT(a, i, v) ((unsigned int) (i) < ARRAY_SIZE(a) ? (a)[(i)] : (v)) +#endif + +#define SET_ERRNO(e) \ +do { \ + parser->http_errno = (e); \ +} while(0) + + +/* Run the notify callback FOR, returning ER if it fails */ +#define CALLBACK_NOTIFY_(FOR, ER) \ +do { \ + assert(HTTP_PARSER_ERRNO(parser) == HPE_OK); \ + \ + if (settings->on_##FOR) { \ + if (0 != settings->on_##FOR(parser)) { \ + SET_ERRNO(HPE_CB_##FOR); \ + } \ + \ + /* We either errored above or got paused; get out */ \ + if (HTTP_PARSER_ERRNO(parser) != HPE_OK) { \ + return (ER); \ + } \ + } \ +} while (0) + +/* Run the notify callback FOR and consume the current byte */ +#define CALLBACK_NOTIFY(FOR) CALLBACK_NOTIFY_(FOR, p - data + 1) + +/* Run the notify callback FOR and don't consume the current byte */ +#define CALLBACK_NOTIFY_NOADVANCE(FOR) CALLBACK_NOTIFY_(FOR, p - data) + +/* Run data callback FOR with LEN bytes, returning ER if it fails */ +#define CALLBACK_DATA_(FOR, LEN, ER) \ +do { \ + assert(HTTP_PARSER_ERRNO(parser) == HPE_OK); \ + \ + if (FOR##_mark) { \ + if (settings->on_##FOR) { \ + if (0 != settings->on_##FOR(parser, FOR##_mark, (LEN))) { \ + SET_ERRNO(HPE_CB_##FOR); \ + } \ + \ + /* We either errored above or got paused; get out */ \ + if (HTTP_PARSER_ERRNO(parser) != HPE_OK) { \ + return (ER); \ + } \ + } \ + FOR##_mark = NULL; \ + } \ +} while (0) + +/* Run the data callback FOR and consume the current byte */ +#define CALLBACK_DATA(FOR) \ + CALLBACK_DATA_(FOR, p - FOR##_mark, p - data + 1) + +/* Run the data callback FOR and don't consume the current byte */ +#define CALLBACK_DATA_NOADVANCE(FOR) \ + CALLBACK_DATA_(FOR, p - FOR##_mark, p - data) + +/* Set the mark FOR; non-destructive if mark is already set */ +#define MARK(FOR) \ +do { \ + if (!FOR##_mark) { \ + FOR##_mark = p; \ + } \ +} while (0) + + +#define PROXY_CONNECTION "proxy-connection" +#define CONNECTION "connection" +#define CONTENT_LENGTH "content-length" +#define TRANSFER_ENCODING "transfer-encoding" +#define UPGRADE "upgrade" +#define CHUNKED "chunked" +#define KEEP_ALIVE "keep-alive" +#define CLOSE "close" + + +static const char *method_strings[] = + { +#define XX(num, name, string) #string, + HTTP_METHOD_MAP(XX) +#undef XX + }; + + +/* Tokens as defined by rfc 2616. Also lowercases them. + * token = 1* + * separators = "(" | ")" | "<" | ">" | "@" + * | "," | ";" | ":" | "\" | <"> + * | "/" | "[" | "]" | "?" | "=" + * | "{" | "}" | SP | HT + */ +static const char tokens[256] = { +/* 0 nul 1 soh 2 stx 3 etx 4 eot 5 enq 6 ack 7 bel */ + 0, 0, 0, 0, 0, 0, 0, 0, +/* 8 bs 9 ht 10 nl 11 vt 12 np 13 cr 14 so 15 si */ + 0, 0, 0, 0, 0, 0, 0, 0, +/* 16 dle 17 dc1 18 dc2 19 dc3 20 dc4 21 nak 22 syn 23 etb */ + 0, 0, 0, 0, 0, 0, 0, 0, +/* 24 can 25 em 26 sub 27 esc 28 fs 29 gs 30 rs 31 us */ + 0, 0, 0, 0, 0, 0, 0, 0, +/* 32 sp 33 ! 34 " 35 # 36 $ 37 % 38 & 39 ' */ + 0, '!', 0, '#', '$', '%', '&', '\'', +/* 40 ( 41 ) 42 * 43 + 44 , 45 - 46 . 47 / */ + 0, 0, '*', '+', 0, '-', '.', 0, +/* 48 0 49 1 50 2 51 3 52 4 53 5 54 6 55 7 */ + '0', '1', '2', '3', '4', '5', '6', '7', +/* 56 8 57 9 58 : 59 ; 60 < 61 = 62 > 63 ? */ + '8', '9', 0, 0, 0, 0, 0, 0, +/* 64 @ 65 A 66 B 67 C 68 D 69 E 70 F 71 G */ + 0, 'a', 'b', 'c', 'd', 'e', 'f', 'g', +/* 72 H 73 I 74 J 75 K 76 L 77 M 78 N 79 O */ + 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', +/* 80 P 81 Q 82 R 83 S 84 T 85 U 86 V 87 W */ + 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', +/* 88 X 89 Y 90 Z 91 [ 92 \ 93 ] 94 ^ 95 _ */ + 'x', 'y', 'z', 0, 0, 0, '^', '_', +/* 96 ` 97 a 98 b 99 c 100 d 101 e 102 f 103 g */ + '`', 'a', 'b', 'c', 'd', 'e', 'f', 'g', +/* 104 h 105 i 106 j 107 k 108 l 109 m 110 n 111 o */ + 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', +/* 112 p 113 q 114 r 115 s 116 t 117 u 118 v 119 w */ + 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', +/* 120 x 121 y 122 z 123 { 124 | 125 } 126 ~ 127 del */ + 'x', 'y', 'z', 0, '|', 0, '~', 0 }; + + +static const int8_t unhex[256] = + {-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 + ,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 + ,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 + , 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,-1,-1,-1,-1,-1,-1 + ,-1,10,11,12,13,14,15,-1,-1,-1,-1,-1,-1,-1,-1,-1 + ,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 + ,-1,10,11,12,13,14,15,-1,-1,-1,-1,-1,-1,-1,-1,-1 + ,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 + }; + + +#if HTTP_PARSER_STRICT +# define T(v) 0 +#else +# define T(v) v +#endif + + +static const uint8_t normal_url_char[32] = { +/* 0 nul 1 soh 2 stx 3 etx 4 eot 5 enq 6 ack 7 bel */ + 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0, +/* 8 bs 9 ht 10 nl 11 vt 12 np 13 cr 14 so 15 si */ + 0 | T(2) | 0 | 0 | T(16) | 0 | 0 | 0, +/* 16 dle 17 dc1 18 dc2 19 dc3 20 dc4 21 nak 22 syn 23 etb */ + 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0, +/* 24 can 25 em 26 sub 27 esc 28 fs 29 gs 30 rs 31 us */ + 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0, +/* 32 sp 33 ! 34 " 35 # 36 $ 37 % 38 & 39 ' */ + 0 | 2 | 4 | 0 | 16 | 32 | 64 | 128, +/* 40 ( 41 ) 42 * 43 + 44 , 45 - 46 . 47 / */ + 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, +/* 48 0 49 1 50 2 51 3 52 4 53 5 54 6 55 7 */ + 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, +/* 56 8 57 9 58 : 59 ; 60 < 61 = 62 > 63 ? */ + 1 | 2 | 4 | 8 | 16 | 32 | 64 | 0, +/* 64 @ 65 A 66 B 67 C 68 D 69 E 70 F 71 G */ + 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, +/* 72 H 73 I 74 J 75 K 76 L 77 M 78 N 79 O */ + 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, +/* 80 P 81 Q 82 R 83 S 84 T 85 U 86 V 87 W */ + 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, +/* 88 X 89 Y 90 Z 91 [ 92 \ 93 ] 94 ^ 95 _ */ + 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, +/* 96 ` 97 a 98 b 99 c 100 d 101 e 102 f 103 g */ + 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, +/* 104 h 105 i 106 j 107 k 108 l 109 m 110 n 111 o */ + 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, +/* 112 p 113 q 114 r 115 s 116 t 117 u 118 v 119 w */ + 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, +/* 120 x 121 y 122 z 123 { 124 | 125 } 126 ~ 127 del */ + 1 | 2 | 4 | 8 | 16 | 32 | 64 | 0, }; + +#undef T + +enum state + { s_dead = 1 /* important that this is > 0 */ + + , s_start_req_or_res + , s_res_or_resp_H + , s_start_res + , s_res_H + , s_res_HT + , s_res_HTT + , s_res_HTTP + , s_res_first_http_major + , s_res_http_major + , s_res_first_http_minor + , s_res_http_minor + , s_res_first_status_code + , s_res_status_code + , s_res_status + , s_res_line_almost_done + + , s_start_req + + , s_req_method + , s_req_spaces_before_url + , s_req_schema + , s_req_schema_slash + , s_req_schema_slash_slash + , s_req_server_start + , s_req_server + , s_req_server_with_at + , s_req_path + , s_req_query_string_start + , s_req_query_string + , s_req_fragment_start + , s_req_fragment + , s_req_http_start + , s_req_http_H + , s_req_http_HT + , s_req_http_HTT + , s_req_http_HTTP + , s_req_first_http_major + , s_req_http_major + , s_req_first_http_minor + , s_req_http_minor + , s_req_line_almost_done + + , s_header_field_start + , s_header_field + , s_header_value_start + , s_header_value + , s_header_value_lws + + , s_header_almost_done + + , s_chunk_size_start + , s_chunk_size + , s_chunk_parameters + , s_chunk_size_almost_done + + , s_headers_almost_done + , s_headers_done + + /* Important: 's_headers_done' must be the last 'header' state. All + * states beyond this must be 'body' states. It is used for overflow + * checking. See the PARSING_HEADER() macro. + */ + + , s_chunk_data + , s_chunk_data_almost_done + , s_chunk_data_done + + , s_body_identity + , s_body_identity_eof + + , s_message_done + }; + + +#define PARSING_HEADER(state) (state <= s_headers_done) + + +enum header_states + { h_general = 0 + , h_C + , h_CO + , h_CON + + , h_matching_connection + , h_matching_proxy_connection + , h_matching_content_length + , h_matching_transfer_encoding + , h_matching_upgrade + + , h_connection + , h_content_length + , h_transfer_encoding + , h_upgrade + + , h_matching_transfer_encoding_chunked + , h_matching_connection_keep_alive + , h_matching_connection_close + + , h_transfer_encoding_chunked + , h_connection_keep_alive + , h_connection_close + }; + +enum http_host_state + { + s_http_host_dead = 1 + , s_http_userinfo_start + , s_http_userinfo + , s_http_host_start + , s_http_host_v6_start + , s_http_host + , s_http_host_v6 + , s_http_host_v6_end + , s_http_host_port_start + , s_http_host_port +}; + +/* Macros for character classes; depends on strict-mode */ +#define CR '\r' +#define LF '\n' +#define LOWER(c) (unsigned char)(c | 0x20) +#define IS_ALPHA(c) (LOWER(c) >= 'a' && LOWER(c) <= 'z') +#define IS_NUM(c) ((c) >= '0' && (c) <= '9') +#define IS_ALPHANUM(c) (IS_ALPHA(c) || IS_NUM(c)) +#define IS_HEX(c) (IS_NUM(c) || (LOWER(c) >= 'a' && LOWER(c) <= 'f')) +#define IS_MARK(c) ((c) == '-' || (c) == '_' || (c) == '.' || \ + (c) == '!' || (c) == '~' || (c) == '*' || (c) == '\'' || (c) == '(' || \ + (c) == ')') +#define IS_USERINFO_CHAR(c) (IS_ALPHANUM(c) || IS_MARK(c) || (c) == '%' || \ + (c) == ';' || (c) == ':' || (c) == '&' || (c) == '=' || (c) == '+' || \ + (c) == '$' || (c) == ',') + +#if HTTP_PARSER_STRICT +#define TOKEN(c) (tokens[(unsigned char)c]) +#define IS_URL_CHAR(c) (BIT_AT(normal_url_char, (unsigned char)c)) +#define IS_HOST_CHAR(c) (IS_ALPHANUM(c) || (c) == '.' || (c) == '-') +#else +#define TOKEN(c) ((c == ' ') ? ' ' : tokens[(unsigned char)c]) +#define IS_URL_CHAR(c) \ + (BIT_AT(normal_url_char, (unsigned char)c) || ((c) & 0x80)) +#define IS_HOST_CHAR(c) \ + (IS_ALPHANUM(c) || (c) == '.' || (c) == '-' || (c) == '_') +#endif + + +#define start_state (parser->type == HTTP_REQUEST ? s_start_req : s_start_res) + + +#if HTTP_PARSER_STRICT +# define STRICT_CHECK(cond) \ +do { \ + if (cond) { \ + SET_ERRNO(HPE_STRICT); \ + goto error; \ + } \ +} while (0) +# define NEW_MESSAGE() (http_should_keep_alive(parser) ? start_state : s_dead) +#else +# define STRICT_CHECK(cond) +# define NEW_MESSAGE() start_state +#endif + + +/* Map errno values to strings for human-readable output */ +#define HTTP_STRERROR_GEN(n, s) { "HPE_" #n, s }, +static struct { + const char *name; + const char *description; +} http_strerror_tab[] = { + HTTP_ERRNO_MAP(HTTP_STRERROR_GEN) +}; +#undef HTTP_STRERROR_GEN + +int http_message_needs_eof(const http_parser *parser); + +/* Our URL parser. + * + * This is designed to be shared by http_parser_execute() for URL validation, + * hence it has a state transition + byte-for-byte interface. In addition, it + * is meant to be embedded in http_parser_parse_url(), which does the dirty + * work of turning state transitions URL components for its API. + * + * This function should only be invoked with non-space characters. It is + * assumed that the caller cares about (and can detect) the transition between + * URL and non-URL states by looking for these. + */ +static enum state +parse_url_char(enum state s, const char ch) +{ + if (ch == ' ' || ch == '\r' || ch == '\n') { + return s_dead; + } + +#if HTTP_PARSER_STRICT + if (ch == '\t' || ch == '\f') { + return s_dead; + } +#endif + + switch (s) { + case s_req_spaces_before_url: + /* Proxied requests are followed by scheme of an absolute URI (alpha). + * All methods except CONNECT are followed by '/' or '*'. + */ + + if (ch == '/' || ch == '*') { + return s_req_path; + } + + if (IS_ALPHA(ch)) { + return s_req_schema; + } + + break; + + case s_req_schema: + if (IS_ALPHA(ch)) { + return s; + } + + if (ch == ':') { + return s_req_schema_slash; + } + + break; + + case s_req_schema_slash: + if (ch == '/') { + return s_req_schema_slash_slash; + } + + break; + + case s_req_schema_slash_slash: + if (ch == '/') { + return s_req_server_start; + } + + break; + + case s_req_server_with_at: + if (ch == '@') { + return s_dead; + } + + /* FALLTHROUGH */ + case s_req_server_start: + case s_req_server: + if (ch == '/') { + return s_req_path; + } + + if (ch == '?') { + return s_req_query_string_start; + } + + if (ch == '@') { + return s_req_server_with_at; + } + + if (IS_USERINFO_CHAR(ch) || ch == '[' || ch == ']') { + return s_req_server; + } + + break; + + case s_req_path: + if (IS_URL_CHAR(ch)) { + return s; + } + + switch (ch) { + case '?': + return s_req_query_string_start; + + case '#': + return s_req_fragment_start; + } + + break; + + case s_req_query_string_start: + case s_req_query_string: + if (IS_URL_CHAR(ch)) { + return s_req_query_string; + } + + switch (ch) { + case '?': + /* allow extra '?' in query string */ + return s_req_query_string; + + case '#': + return s_req_fragment_start; + } + + break; + + case s_req_fragment_start: + if (IS_URL_CHAR(ch)) { + return s_req_fragment; + } + + switch (ch) { + case '?': + return s_req_fragment; + + case '#': + return s; + } + + break; + + case s_req_fragment: + if (IS_URL_CHAR(ch)) { + return s; + } + + switch (ch) { + case '?': + case '#': + return s; + } + + break; + + default: + break; + } + + /* We should never fall out of the switch above unless there's an error */ + return s_dead; +} + +size_t http_parser_execute (http_parser *parser, + const http_parser_settings *settings, + const char *data, + size_t len) +{ + char c, ch; + int8_t unhex_val; + const char *p = data; + const char *header_field_mark = 0; + const char *header_value_mark = 0; + const char *url_mark = 0; + const char *body_mark = 0; + + /* We're in an error state. Don't bother doing anything. */ + if (HTTP_PARSER_ERRNO(parser) != HPE_OK) { + return 0; + } + + if (len == 0) { + switch (parser->state) { + case s_body_identity_eof: + /* Use of CALLBACK_NOTIFY() here would erroneously return 1 byte read if + * we got paused. + */ + CALLBACK_NOTIFY_NOADVANCE(message_complete); + return 0; + + case s_dead: + case s_start_req_or_res: + case s_start_res: + case s_start_req: + return 0; + + default: + SET_ERRNO(HPE_INVALID_EOF_STATE); + return 1; + } + } + + + if (parser->state == s_header_field) + header_field_mark = data; + if (parser->state == s_header_value) + header_value_mark = data; + switch (parser->state) { + case s_req_path: + case s_req_schema: + case s_req_schema_slash: + case s_req_schema_slash_slash: + case s_req_server_start: + case s_req_server: + case s_req_server_with_at: + case s_req_query_string_start: + case s_req_query_string: + case s_req_fragment_start: + case s_req_fragment: + url_mark = data; + break; + } + + for (p=data; p != data + len; p++) { + ch = *p; + + if (PARSING_HEADER(parser->state)) { + ++parser->nread; + /* Buffer overflow attack */ + if (parser->nread > HTTP_MAX_HEADER_SIZE) { + SET_ERRNO(HPE_HEADER_OVERFLOW); + goto error; + } + } + + reexecute_byte: + switch (parser->state) { + + case s_dead: + /* this state is used after a 'Connection: close' message + * the parser will error out if it reads another message + */ + if (ch == CR || ch == LF) + break; + + SET_ERRNO(HPE_CLOSED_CONNECTION); + goto error; + + case s_start_req_or_res: + { + if (ch == CR || ch == LF) + break; + parser->flags = 0; + parser->content_length = ULLONG_MAX; + + if (ch == 'H') { + parser->state = s_res_or_resp_H; + + CALLBACK_NOTIFY(message_begin); + } else { + parser->type = HTTP_REQUEST; + parser->state = s_start_req; + goto reexecute_byte; + } + + break; + } + + case s_res_or_resp_H: + if (ch == 'T') { + parser->type = HTTP_RESPONSE; + parser->state = s_res_HT; + } else { + if (ch != 'E') { + SET_ERRNO(HPE_INVALID_CONSTANT); + goto error; + } + + parser->type = HTTP_REQUEST; + parser->method = HTTP_HEAD; + parser->index = 2; + parser->state = s_req_method; + } + break; + + case s_start_res: + { + parser->flags = 0; + parser->content_length = ULLONG_MAX; + + switch (ch) { + case 'H': + parser->state = s_res_H; + break; + + case CR: + case LF: + break; + + default: + SET_ERRNO(HPE_INVALID_CONSTANT); + goto error; + } + + CALLBACK_NOTIFY(message_begin); + break; + } + + case s_res_H: + STRICT_CHECK(ch != 'T'); + parser->state = s_res_HT; + break; + + case s_res_HT: + STRICT_CHECK(ch != 'T'); + parser->state = s_res_HTT; + break; + + case s_res_HTT: + STRICT_CHECK(ch != 'P'); + parser->state = s_res_HTTP; + break; + + case s_res_HTTP: + STRICT_CHECK(ch != '/'); + parser->state = s_res_first_http_major; + break; + + case s_res_first_http_major: + if (ch < '0' || ch > '9') { + SET_ERRNO(HPE_INVALID_VERSION); + goto error; + } + + parser->http_major = ch - '0'; + parser->state = s_res_http_major; + break; + + /* major HTTP version or dot */ + case s_res_http_major: + { + if (ch == '.') { + parser->state = s_res_first_http_minor; + break; + } + + if (!IS_NUM(ch)) { + SET_ERRNO(HPE_INVALID_VERSION); + goto error; + } + + parser->http_major *= 10; + parser->http_major += ch - '0'; + + if (parser->http_major > 999) { + SET_ERRNO(HPE_INVALID_VERSION); + goto error; + } + + break; + } + + /* first digit of minor HTTP version */ + case s_res_first_http_minor: + if (!IS_NUM(ch)) { + SET_ERRNO(HPE_INVALID_VERSION); + goto error; + } + + parser->http_minor = ch - '0'; + parser->state = s_res_http_minor; + break; + + /* minor HTTP version or end of request line */ + case s_res_http_minor: + { + if (ch == ' ') { + parser->state = s_res_first_status_code; + break; + } + + if (!IS_NUM(ch)) { + SET_ERRNO(HPE_INVALID_VERSION); + goto error; + } + + parser->http_minor *= 10; + parser->http_minor += ch - '0'; + + if (parser->http_minor > 999) { + SET_ERRNO(HPE_INVALID_VERSION); + goto error; + } + + break; + } + + case s_res_first_status_code: + { + if (!IS_NUM(ch)) { + if (ch == ' ') { + break; + } + + SET_ERRNO(HPE_INVALID_STATUS); + goto error; + } + parser->status_code = ch - '0'; + parser->state = s_res_status_code; + break; + } + + case s_res_status_code: + { + if (!IS_NUM(ch)) { + switch (ch) { + case ' ': + parser->state = s_res_status; + break; + case CR: + parser->state = s_res_line_almost_done; + break; + case LF: + parser->state = s_header_field_start; + break; + default: + SET_ERRNO(HPE_INVALID_STATUS); + goto error; + } + break; + } + + parser->status_code *= 10; + parser->status_code += ch - '0'; + + if (parser->status_code > 999) { + SET_ERRNO(HPE_INVALID_STATUS); + goto error; + } + + break; + } + + case s_res_status: + /* the human readable status. e.g. "NOT FOUND" + * we are not humans so just ignore this */ + if (ch == CR) { + parser->state = s_res_line_almost_done; + break; + } + + if (ch == LF) { + parser->state = s_header_field_start; + break; + } + break; + + case s_res_line_almost_done: + STRICT_CHECK(ch != LF); + parser->state = s_header_field_start; + CALLBACK_NOTIFY(status_complete); + break; + + case s_start_req: + { + if (ch == CR || ch == LF) + break; + parser->flags = 0; + parser->content_length = ULLONG_MAX; + + if (!IS_ALPHA(ch)) { + SET_ERRNO(HPE_INVALID_METHOD); + goto error; + } + + parser->method = (enum http_method) 0; + parser->index = 1; + switch (ch) { + case 'C': parser->method = HTTP_CONNECT; /* or COPY, CHECKOUT */ break; + case 'D': parser->method = HTTP_DELETE; break; + case 'G': parser->method = HTTP_GET; break; + case 'H': parser->method = HTTP_HEAD; break; + case 'L': parser->method = HTTP_LOCK; break; + case 'M': parser->method = HTTP_MKCOL; /* or MOVE, MKACTIVITY, MERGE, M-SEARCH */ break; + case 'N': parser->method = HTTP_NOTIFY; break; + case 'O': parser->method = HTTP_OPTIONS; break; + case 'P': parser->method = HTTP_POST; + /* or PROPFIND|PROPPATCH|PUT|PATCH|PURGE */ + break; + case 'R': parser->method = HTTP_REPORT; break; + case 'S': parser->method = HTTP_SUBSCRIBE; /* or SEARCH */ break; + case 'T': parser->method = HTTP_TRACE; break; + case 'U': parser->method = HTTP_UNLOCK; /* or UNSUBSCRIBE */ break; + default: + SET_ERRNO(HPE_INVALID_METHOD); + goto error; + } + parser->state = s_req_method; + + CALLBACK_NOTIFY(message_begin); + + break; + } + + case s_req_method: + { + const char *matcher; + if (ch == '\0') { + SET_ERRNO(HPE_INVALID_METHOD); + goto error; + } + + matcher = method_strings[parser->method]; + if (ch == ' ' && matcher[parser->index] == '\0') { + parser->state = s_req_spaces_before_url; + } else if (ch == matcher[parser->index]) { + ; /* nada */ + } else if (parser->method == HTTP_CONNECT) { + if (parser->index == 1 && ch == 'H') { + parser->method = HTTP_CHECKOUT; + } else if (parser->index == 2 && ch == 'P') { + parser->method = HTTP_COPY; + } else { + goto error; + } + } else if (parser->method == HTTP_MKCOL) { + if (parser->index == 1 && ch == 'O') { + parser->method = HTTP_MOVE; + } else if (parser->index == 1 && ch == 'E') { + parser->method = HTTP_MERGE; + } else if (parser->index == 1 && ch == '-') { + parser->method = HTTP_MSEARCH; + } else if (parser->index == 2 && ch == 'A') { + parser->method = HTTP_MKACTIVITY; + } else { + goto error; + } + } else if (parser->method == HTTP_SUBSCRIBE) { + if (parser->index == 1 && ch == 'E') { + parser->method = HTTP_SEARCH; + } else { + goto error; + } + } else if (parser->index == 1 && parser->method == HTTP_POST) { + if (ch == 'R') { + parser->method = HTTP_PROPFIND; /* or HTTP_PROPPATCH */ + } else if (ch == 'U') { + parser->method = HTTP_PUT; /* or HTTP_PURGE */ + } else if (ch == 'A') { + parser->method = HTTP_PATCH; + } else { + goto error; + } + } else if (parser->index == 2) { + if (parser->method == HTTP_PUT) { + if (ch == 'R') parser->method = HTTP_PURGE; + } else if (parser->method == HTTP_UNLOCK) { + if (ch == 'S') parser->method = HTTP_UNSUBSCRIBE; + } + } else if (parser->index == 4 && parser->method == HTTP_PROPFIND && ch == 'P') { + parser->method = HTTP_PROPPATCH; + } else { + SET_ERRNO(HPE_INVALID_METHOD); + goto error; + } + + ++parser->index; + break; + } + + case s_req_spaces_before_url: + { + if (ch == ' ') break; + + MARK(url); + if (parser->method == HTTP_CONNECT) { + parser->state = s_req_server_start; + } + + parser->state = parse_url_char((enum state)parser->state, ch); + if (parser->state == s_dead) { + SET_ERRNO(HPE_INVALID_URL); + goto error; + } + + break; + } + + case s_req_schema: + case s_req_schema_slash: + case s_req_schema_slash_slash: + case s_req_server_start: + { + switch (ch) { + /* No whitespace allowed here */ + case ' ': + case CR: + case LF: + SET_ERRNO(HPE_INVALID_URL); + goto error; + default: + parser->state = parse_url_char((enum state)parser->state, ch); + if (parser->state == s_dead) { + SET_ERRNO(HPE_INVALID_URL); + goto error; + } + } + + break; + } + + case s_req_server: + case s_req_server_with_at: + case s_req_path: + case s_req_query_string_start: + case s_req_query_string: + case s_req_fragment_start: + case s_req_fragment: + { + switch (ch) { + case ' ': + parser->state = s_req_http_start; + CALLBACK_DATA(url); + break; + case CR: + case LF: + parser->http_major = 0; + parser->http_minor = 9; + parser->state = (ch == CR) ? + s_req_line_almost_done : + s_header_field_start; + CALLBACK_DATA(url); + break; + default: + parser->state = parse_url_char((enum state)parser->state, ch); + if (parser->state == s_dead) { + SET_ERRNO(HPE_INVALID_URL); + goto error; + } + } + break; + } + + case s_req_http_start: + switch (ch) { + case 'H': + parser->state = s_req_http_H; + break; + case ' ': + break; + default: + SET_ERRNO(HPE_INVALID_CONSTANT); + goto error; + } + break; + + case s_req_http_H: + STRICT_CHECK(ch != 'T'); + parser->state = s_req_http_HT; + break; + + case s_req_http_HT: + STRICT_CHECK(ch != 'T'); + parser->state = s_req_http_HTT; + break; + + case s_req_http_HTT: + STRICT_CHECK(ch != 'P'); + parser->state = s_req_http_HTTP; + break; + + case s_req_http_HTTP: + STRICT_CHECK(ch != '/'); + parser->state = s_req_first_http_major; + break; + + /* first digit of major HTTP version */ + case s_req_first_http_major: + if (ch < '1' || ch > '9') { + SET_ERRNO(HPE_INVALID_VERSION); + goto error; + } + + parser->http_major = ch - '0'; + parser->state = s_req_http_major; + break; + + /* major HTTP version or dot */ + case s_req_http_major: + { + if (ch == '.') { + parser->state = s_req_first_http_minor; + break; + } + + if (!IS_NUM(ch)) { + SET_ERRNO(HPE_INVALID_VERSION); + goto error; + } + + parser->http_major *= 10; + parser->http_major += ch - '0'; + + if (parser->http_major > 999) { + SET_ERRNO(HPE_INVALID_VERSION); + goto error; + } + + break; + } + + /* first digit of minor HTTP version */ + case s_req_first_http_minor: + if (!IS_NUM(ch)) { + SET_ERRNO(HPE_INVALID_VERSION); + goto error; + } + + parser->http_minor = ch - '0'; + parser->state = s_req_http_minor; + break; + + /* minor HTTP version or end of request line */ + case s_req_http_minor: + { + if (ch == CR) { + parser->state = s_req_line_almost_done; + break; + } + + if (ch == LF) { + parser->state = s_header_field_start; + break; + } + + /* XXX allow spaces after digit? */ + + if (!IS_NUM(ch)) { + SET_ERRNO(HPE_INVALID_VERSION); + goto error; + } + + parser->http_minor *= 10; + parser->http_minor += ch - '0'; + + if (parser->http_minor > 999) { + SET_ERRNO(HPE_INVALID_VERSION); + goto error; + } + + break; + } + + /* end of request line */ + case s_req_line_almost_done: + { + if (ch != LF) { + SET_ERRNO(HPE_LF_EXPECTED); + goto error; + } + + parser->state = s_header_field_start; + break; + } + + case s_header_field_start: + { + if (ch == CR) { + parser->state = s_headers_almost_done; + break; + } + + if (ch == LF) { + /* they might be just sending \n instead of \r\n so this would be + * the second \n to denote the end of headers*/ + parser->state = s_headers_almost_done; + goto reexecute_byte; + } + + c = TOKEN(ch); + + if (!c) { + SET_ERRNO(HPE_INVALID_HEADER_TOKEN); + goto error; + } + + MARK(header_field); + + parser->index = 0; + parser->state = s_header_field; + + switch (c) { + case 'c': + parser->header_state = h_C; + break; + + case 'p': + parser->header_state = h_matching_proxy_connection; + break; + + case 't': + parser->header_state = h_matching_transfer_encoding; + break; + + case 'u': + parser->header_state = h_matching_upgrade; + break; + + default: + parser->header_state = h_general; + break; + } + break; + } + + case s_header_field: + { + c = TOKEN(ch); + + if (c) { + switch (parser->header_state) { + case h_general: + break; + + case h_C: + parser->index++; + parser->header_state = (c == 'o' ? h_CO : h_general); + break; + + case h_CO: + parser->index++; + parser->header_state = (c == 'n' ? h_CON : h_general); + break; + + case h_CON: + parser->index++; + switch (c) { + case 'n': + parser->header_state = h_matching_connection; + break; + case 't': + parser->header_state = h_matching_content_length; + break; + default: + parser->header_state = h_general; + break; + } + break; + + /* connection */ + + case h_matching_connection: + parser->index++; + if (parser->index > sizeof(CONNECTION)-1 + || c != CONNECTION[parser->index]) { + parser->header_state = h_general; + } else if (parser->index == sizeof(CONNECTION)-2) { + parser->header_state = h_connection; + } + break; + + /* proxy-connection */ + + case h_matching_proxy_connection: + parser->index++; + if (parser->index > sizeof(PROXY_CONNECTION)-1 + || c != PROXY_CONNECTION[parser->index]) { + parser->header_state = h_general; + } else if (parser->index == sizeof(PROXY_CONNECTION)-2) { + parser->header_state = h_connection; + } + break; + + /* content-length */ + + case h_matching_content_length: + parser->index++; + if (parser->index > sizeof(CONTENT_LENGTH)-1 + || c != CONTENT_LENGTH[parser->index]) { + parser->header_state = h_general; + } else if (parser->index == sizeof(CONTENT_LENGTH)-2) { + parser->header_state = h_content_length; + } + break; + + /* transfer-encoding */ + + case h_matching_transfer_encoding: + parser->index++; + if (parser->index > sizeof(TRANSFER_ENCODING)-1 + || c != TRANSFER_ENCODING[parser->index]) { + parser->header_state = h_general; + } else if (parser->index == sizeof(TRANSFER_ENCODING)-2) { + parser->header_state = h_transfer_encoding; + } + break; + + /* upgrade */ + + case h_matching_upgrade: + parser->index++; + if (parser->index > sizeof(UPGRADE)-1 + || c != UPGRADE[parser->index]) { + parser->header_state = h_general; + } else if (parser->index == sizeof(UPGRADE)-2) { + parser->header_state = h_upgrade; + } + break; + + case h_connection: + case h_content_length: + case h_transfer_encoding: + case h_upgrade: + if (ch != ' ') parser->header_state = h_general; + break; + + default: + assert(0 && "Unknown header_state"); + break; + } + break; + } + + if (ch == ':') { + parser->state = s_header_value_start; + CALLBACK_DATA(header_field); + break; + } + + if (ch == CR) { + parser->state = s_header_almost_done; + CALLBACK_DATA(header_field); + break; + } + + if (ch == LF) { + parser->state = s_header_field_start; + CALLBACK_DATA(header_field); + break; + } + + SET_ERRNO(HPE_INVALID_HEADER_TOKEN); + goto error; + } + + case s_header_value_start: + { + if (ch == ' ' || ch == '\t') break; + + MARK(header_value); + + parser->state = s_header_value; + parser->index = 0; + + if (ch == CR) { + parser->header_state = h_general; + parser->state = s_header_almost_done; + CALLBACK_DATA(header_value); + break; + } + + if (ch == LF) { + parser->state = s_header_field_start; + CALLBACK_DATA(header_value); + break; + } + + c = LOWER(ch); + + switch (parser->header_state) { + case h_upgrade: + parser->flags |= F_UPGRADE; + parser->header_state = h_general; + break; + + case h_transfer_encoding: + /* looking for 'Transfer-Encoding: chunked' */ + if ('c' == c) { + parser->header_state = h_matching_transfer_encoding_chunked; + } else { + parser->header_state = h_general; + } + break; + + case h_content_length: + if (!IS_NUM(ch)) { + SET_ERRNO(HPE_INVALID_CONTENT_LENGTH); + goto error; + } + + parser->content_length = ch - '0'; + break; + + case h_connection: + /* looking for 'Connection: keep-alive' */ + if (c == 'k') { + parser->header_state = h_matching_connection_keep_alive; + /* looking for 'Connection: close' */ + } else if (c == 'c') { + parser->header_state = h_matching_connection_close; + } else { + parser->header_state = h_general; + } + break; + + default: + parser->header_state = h_general; + break; + } + break; + } + + case s_header_value: + { + + if (ch == CR) { + parser->state = s_header_almost_done; + CALLBACK_DATA(header_value); + break; + } + + if (ch == LF) { + parser->state = s_header_almost_done; + CALLBACK_DATA_NOADVANCE(header_value); + goto reexecute_byte; + } + + c = LOWER(ch); + + switch (parser->header_state) { + case h_general: + break; + + case h_connection: + case h_transfer_encoding: + assert(0 && "Shouldn't get here."); + break; + + case h_content_length: + { + uint64_t t; + + if (ch == ' ') break; + + if (!IS_NUM(ch)) { + SET_ERRNO(HPE_INVALID_CONTENT_LENGTH); + goto error; + } + + t = parser->content_length; + t *= 10; + t += ch - '0'; + + /* Overflow? */ + if (t < parser->content_length || t == ULLONG_MAX) { + SET_ERRNO(HPE_INVALID_CONTENT_LENGTH); + goto error; + } + + parser->content_length = t; + break; + } + + /* Transfer-Encoding: chunked */ + case h_matching_transfer_encoding_chunked: + parser->index++; + if (parser->index > sizeof(CHUNKED)-1 + || c != CHUNKED[parser->index]) { + parser->header_state = h_general; + } else if (parser->index == sizeof(CHUNKED)-2) { + parser->header_state = h_transfer_encoding_chunked; + } + break; + + /* looking for 'Connection: keep-alive' */ + case h_matching_connection_keep_alive: + parser->index++; + if (parser->index > sizeof(KEEP_ALIVE)-1 + || c != KEEP_ALIVE[parser->index]) { + parser->header_state = h_general; + } else if (parser->index == sizeof(KEEP_ALIVE)-2) { + parser->header_state = h_connection_keep_alive; + } + break; + + /* looking for 'Connection: close' */ + case h_matching_connection_close: + parser->index++; + if (parser->index > sizeof(CLOSE)-1 || c != CLOSE[parser->index]) { + parser->header_state = h_general; + } else if (parser->index == sizeof(CLOSE)-2) { + parser->header_state = h_connection_close; + } + break; + + case h_transfer_encoding_chunked: + case h_connection_keep_alive: + case h_connection_close: + if (ch != ' ') parser->header_state = h_general; + break; + + default: + parser->state = s_header_value; + parser->header_state = h_general; + break; + } + break; + } + + case s_header_almost_done: + { + STRICT_CHECK(ch != LF); + + parser->state = s_header_value_lws; + + switch (parser->header_state) { + case h_connection_keep_alive: + parser->flags |= F_CONNECTION_KEEP_ALIVE; + break; + case h_connection_close: + parser->flags |= F_CONNECTION_CLOSE; + break; + case h_transfer_encoding_chunked: + parser->flags |= F_CHUNKED; + break; + default: + break; + } + + break; + } + + case s_header_value_lws: + { + if (ch == ' ' || ch == '\t') + parser->state = s_header_value_start; + else + { + parser->state = s_header_field_start; + goto reexecute_byte; + } + break; + } + + case s_headers_almost_done: + { + STRICT_CHECK(ch != LF); + + if (parser->flags & F_TRAILING) { + /* End of a chunked request */ + parser->state = NEW_MESSAGE(); + CALLBACK_NOTIFY(message_complete); + break; + } + + parser->state = s_headers_done; + + /* Set this here so that on_headers_complete() callbacks can see it */ + parser->upgrade = + (parser->flags & F_UPGRADE || parser->method == HTTP_CONNECT); + + /* Here we call the headers_complete callback. This is somewhat + * different than other callbacks because if the user returns 1, we + * will interpret that as saying that this message has no body. This + * is needed for the annoying case of recieving a response to a HEAD + * request. + * + * We'd like to use CALLBACK_NOTIFY_NOADVANCE() here but we cannot, so + * we have to simulate it by handling a change in errno below. + */ + if (settings->on_headers_complete) { + switch (settings->on_headers_complete(parser)) { + case 0: + break; + + case 1: + parser->flags |= F_SKIPBODY; + break; + + default: + SET_ERRNO(HPE_CB_headers_complete); + return p - data; /* Error */ + } + } + + if (HTTP_PARSER_ERRNO(parser) != HPE_OK) { + return p - data; + } + + goto reexecute_byte; + } + + case s_headers_done: + { + STRICT_CHECK(ch != LF); + + parser->nread = 0; + + /* Exit, the rest of the connect is in a different protocol. */ + if (parser->upgrade) { + parser->state = NEW_MESSAGE(); + CALLBACK_NOTIFY(message_complete); + return (p - data) + 1; + } + + if (parser->flags & F_SKIPBODY) { + parser->state = NEW_MESSAGE(); + CALLBACK_NOTIFY(message_complete); + } else if (parser->flags & F_CHUNKED) { + /* chunked encoding - ignore Content-Length header */ + parser->state = s_chunk_size_start; + } else { + if (parser->content_length == 0) { + /* Content-Length header given but zero: Content-Length: 0\r\n */ + parser->state = NEW_MESSAGE(); + CALLBACK_NOTIFY(message_complete); + } else if (parser->content_length != ULLONG_MAX) { + /* Content-Length header given and non-zero */ + parser->state = s_body_identity; + } else { + if (parser->type == HTTP_REQUEST || + !http_message_needs_eof(parser)) { + /* Assume content-length 0 - read the next */ + parser->state = NEW_MESSAGE(); + CALLBACK_NOTIFY(message_complete); + } else { + /* Read body until EOF */ + parser->state = s_body_identity_eof; + } + } + } + + break; + } + + case s_body_identity: + { + uint64_t to_read = MIN(parser->content_length, + (uint64_t) ((data + len) - p)); + + assert(parser->content_length != 0 + && parser->content_length != ULLONG_MAX); + + /* The difference between advancing content_length and p is because + * the latter will automaticaly advance on the next loop iteration. + * Further, if content_length ends up at 0, we want to see the last + * byte again for our message complete callback. + */ + MARK(body); + parser->content_length -= to_read; + p += to_read - 1; + + if (parser->content_length == 0) { + parser->state = s_message_done; + + /* Mimic CALLBACK_DATA_NOADVANCE() but with one extra byte. + * + * The alternative to doing this is to wait for the next byte to + * trigger the data callback, just as in every other case. The + * problem with this is that this makes it difficult for the test + * harness to distinguish between complete-on-EOF and + * complete-on-length. It's not clear that this distinction is + * important for applications, but let's keep it for now. + */ + CALLBACK_DATA_(body, p - body_mark + 1, p - data); + goto reexecute_byte; + } + + break; + } + + /* read until EOF */ + case s_body_identity_eof: + MARK(body); + p = data + len - 1; + + break; + + case s_message_done: + parser->state = NEW_MESSAGE(); + CALLBACK_NOTIFY(message_complete); + break; + + case s_chunk_size_start: + { + assert(parser->nread == 1); + assert(parser->flags & F_CHUNKED); + + unhex_val = unhex[(unsigned char)ch]; + if (unhex_val == -1) { + SET_ERRNO(HPE_INVALID_CHUNK_SIZE); + goto error; + } + + parser->content_length = unhex_val; + parser->state = s_chunk_size; + break; + } + + case s_chunk_size: + { + uint64_t t; + + assert(parser->flags & F_CHUNKED); + + if (ch == CR) { + parser->state = s_chunk_size_almost_done; + break; + } + + unhex_val = unhex[(unsigned char)ch]; + + if (unhex_val == -1) { + if (ch == ';' || ch == ' ') { + parser->state = s_chunk_parameters; + break; + } + + SET_ERRNO(HPE_INVALID_CHUNK_SIZE); + goto error; + } + + t = parser->content_length; + t *= 16; + t += unhex_val; + + /* Overflow? */ + if (t < parser->content_length || t == ULLONG_MAX) { + SET_ERRNO(HPE_INVALID_CONTENT_LENGTH); + goto error; + } + + parser->content_length = t; + break; + } + + case s_chunk_parameters: + { + assert(parser->flags & F_CHUNKED); + /* just ignore this shit. TODO check for overflow */ + if (ch == CR) { + parser->state = s_chunk_size_almost_done; + break; + } + break; + } + + case s_chunk_size_almost_done: + { + assert(parser->flags & F_CHUNKED); + STRICT_CHECK(ch != LF); + + parser->nread = 0; + + if (parser->content_length == 0) { + parser->flags |= F_TRAILING; + parser->state = s_header_field_start; + } else { + parser->state = s_chunk_data; + } + break; + } + + case s_chunk_data: + { + uint64_t to_read = MIN(parser->content_length, + (uint64_t) ((data + len) - p)); + + assert(parser->flags & F_CHUNKED); + assert(parser->content_length != 0 + && parser->content_length != ULLONG_MAX); + + /* See the explanation in s_body_identity for why the content + * length and data pointers are managed this way. + */ + MARK(body); + parser->content_length -= to_read; + p += to_read - 1; + + if (parser->content_length == 0) { + parser->state = s_chunk_data_almost_done; + } + + break; + } + + case s_chunk_data_almost_done: + assert(parser->flags & F_CHUNKED); + assert(parser->content_length == 0); + STRICT_CHECK(ch != CR); + parser->state = s_chunk_data_done; + CALLBACK_DATA(body); + break; + + case s_chunk_data_done: + assert(parser->flags & F_CHUNKED); + STRICT_CHECK(ch != LF); + parser->nread = 0; + parser->state = s_chunk_size_start; + break; + + default: + assert(0 && "unhandled state"); + SET_ERRNO(HPE_INVALID_INTERNAL_STATE); + goto error; + } + } + + /* Run callbacks for any marks that we have leftover after we ran our of + * bytes. There should be at most one of these set, so it's OK to invoke + * them in series (unset marks will not result in callbacks). + * + * We use the NOADVANCE() variety of callbacks here because 'p' has already + * overflowed 'data' and this allows us to correct for the off-by-one that + * we'd otherwise have (since CALLBACK_DATA() is meant to be run with a 'p' + * value that's in-bounds). + */ + + assert(((header_field_mark ? 1 : 0) + + (header_value_mark ? 1 : 0) + + (url_mark ? 1 : 0) + + (body_mark ? 1 : 0)) <= 1); + + CALLBACK_DATA_NOADVANCE(header_field); + CALLBACK_DATA_NOADVANCE(header_value); + CALLBACK_DATA_NOADVANCE(url); + CALLBACK_DATA_NOADVANCE(body); + + return len; + +error: + if (HTTP_PARSER_ERRNO(parser) == HPE_OK) { + SET_ERRNO(HPE_UNKNOWN); + } + + return (p - data); +} + + +/* Does the parser need to see an EOF to find the end of the message? */ +int +http_message_needs_eof (const http_parser *parser) +{ + if (parser->type == HTTP_REQUEST) { + return 0; + } + + /* See RFC 2616 section 4.4 */ + if (parser->status_code / 100 == 1 || /* 1xx e.g. Continue */ + parser->status_code == 204 || /* No Content */ + parser->status_code == 304 || /* Not Modified */ + parser->flags & F_SKIPBODY) { /* response to a HEAD request */ + return 0; + } + + if ((parser->flags & F_CHUNKED) || parser->content_length != ULLONG_MAX) { + return 0; + } + + return 1; +} + + +int +http_should_keep_alive (const http_parser *parser) +{ + if (parser->http_major > 0 && parser->http_minor > 0) { + /* HTTP/1.1 */ + if (parser->flags & F_CONNECTION_CLOSE) { + return 0; + } + } else { + /* HTTP/1.0 or earlier */ + if (!(parser->flags & F_CONNECTION_KEEP_ALIVE)) { + return 0; + } + } + + return !http_message_needs_eof(parser); +} + + +const char * +http_method_str (enum http_method m) +{ + return ELEM_AT(method_strings, m, ""); +} + + +void +http_parser_init (http_parser *parser, enum http_parser_type t) +{ + void *data = parser->data; /* preserve application data */ + memset(parser, 0, sizeof(*parser)); + parser->data = data; + parser->type = t; + parser->state = (t == HTTP_REQUEST ? s_start_req : (t == HTTP_RESPONSE ? s_start_res : s_start_req_or_res)); + parser->http_errno = HPE_OK; +} + +const char * +http_errno_name(enum http_errno err) { + assert(err < (sizeof(http_strerror_tab)/sizeof(http_strerror_tab[0]))); + return http_strerror_tab[err].name; +} + +const char * +http_errno_description(enum http_errno err) { + assert(err < (sizeof(http_strerror_tab)/sizeof(http_strerror_tab[0]))); + return http_strerror_tab[err].description; +} + +static enum http_host_state +http_parse_host_char(enum http_host_state s, const char ch) { + switch(s) { + case s_http_userinfo: + case s_http_userinfo_start: + if (ch == '@') { + return s_http_host_start; + } + + if (IS_USERINFO_CHAR(ch)) { + return s_http_userinfo; + } + break; + + case s_http_host_start: + if (ch == '[') { + return s_http_host_v6_start; + } + + if (IS_HOST_CHAR(ch)) { + return s_http_host; + } + + break; + + case s_http_host: + if (IS_HOST_CHAR(ch)) { + return s_http_host; + } + + /* FALLTHROUGH */ + case s_http_host_v6_end: + if (ch == ':') { + return s_http_host_port_start; + } + + break; + + case s_http_host_v6: + if (ch == ']') { + return s_http_host_v6_end; + } + + /* FALLTHROUGH */ + case s_http_host_v6_start: + if (IS_HEX(ch) || ch == ':' || ch == '.') { + return s_http_host_v6; + } + + break; + + case s_http_host_port: + case s_http_host_port_start: + if (IS_NUM(ch)) { + return s_http_host_port; + } + + break; + + default: + break; + } + return s_http_host_dead; +} + +static int +http_parse_host(const char * buf, struct http_parser_url *u, int found_at) { + enum http_host_state s; + + const char *p; + size_t buflen = u->field_data[UF_HOST].off + u->field_data[UF_HOST].len; + + u->field_data[UF_HOST].len = 0; + + s = found_at ? s_http_userinfo_start : s_http_host_start; + + for (p = buf + u->field_data[UF_HOST].off; p < buf + buflen; p++) { + enum http_host_state new_s = http_parse_host_char(s, *p); + + if (new_s == s_http_host_dead) { + return 1; + } + + switch(new_s) { + case s_http_host: + if (s != s_http_host) { + u->field_data[UF_HOST].off = p - buf; + } + u->field_data[UF_HOST].len++; + break; + + case s_http_host_v6: + if (s != s_http_host_v6) { + u->field_data[UF_HOST].off = p - buf; + } + u->field_data[UF_HOST].len++; + break; + + case s_http_host_port: + if (s != s_http_host_port) { + u->field_data[UF_PORT].off = p - buf; + u->field_data[UF_PORT].len = 0; + u->field_set |= (1 << UF_PORT); + } + u->field_data[UF_PORT].len++; + break; + + case s_http_userinfo: + if (s != s_http_userinfo) { + u->field_data[UF_USERINFO].off = p - buf ; + u->field_data[UF_USERINFO].len = 0; + u->field_set |= (1 << UF_USERINFO); + } + u->field_data[UF_USERINFO].len++; + break; + + default: + break; + } + s = new_s; + } + + /* Make sure we don't end somewhere unexpected */ + switch (s) { + case s_http_host_start: + case s_http_host_v6_start: + case s_http_host_v6: + case s_http_host_port_start: + case s_http_userinfo: + case s_http_userinfo_start: + return 1; + default: + break; + } + + return 0; +} + +int +http_parser_parse_url(const char *buf, size_t buflen, int is_connect, + struct http_parser_url *u) +{ + enum state s; + const char *p; + enum http_parser_url_fields uf, old_uf; + int found_at = 0; + + u->port = u->field_set = 0; + s = is_connect ? s_req_server_start : s_req_spaces_before_url; + uf = old_uf = UF_MAX; + + for (p = buf; p < buf + buflen; p++) { + s = parse_url_char(s, *p); + + /* Figure out the next field that we're operating on */ + switch (s) { + case s_dead: + return 1; + + /* Skip delimeters */ + case s_req_schema_slash: + case s_req_schema_slash_slash: + case s_req_server_start: + case s_req_query_string_start: + case s_req_fragment_start: + continue; + + case s_req_schema: + uf = UF_SCHEMA; + break; + + case s_req_server_with_at: + found_at = 1; + + /* FALLTROUGH */ + case s_req_server: + uf = UF_HOST; + break; + + case s_req_path: + uf = UF_PATH; + break; + + case s_req_query_string: + uf = UF_QUERY; + break; + + case s_req_fragment: + uf = UF_FRAGMENT; + break; + + default: + assert(!"Unexpected state"); + return 1; + } + + /* Nothing's changed; soldier on */ + if (uf == old_uf) { + u->field_data[uf].len++; + continue; + } + + u->field_data[uf].off = p - buf; + u->field_data[uf].len = 1; + + u->field_set |= (1 << uf); + old_uf = uf; + } + + /* host must be present if there is a schema */ + /* parsing http:///toto will fail */ + if ((u->field_set & ((1 << UF_SCHEMA) | (1 << UF_HOST))) != 0) { + if (http_parse_host(buf, u, found_at) != 0) { + return 1; + } + } + + /* CONNECT requests can only contain "hostname:port" */ + if (is_connect && u->field_set != ((1 << UF_HOST)|(1 << UF_PORT))) { + return 1; + } + + if (u->field_set & (1 << UF_PORT)) { + /* Don't bother with endp; we've already validated the string */ + unsigned long v = strtoul(buf + u->field_data[UF_PORT].off, NULL, 10); + + /* Ports have a max value of 2^16 */ + if (v > 0xffff) { + return 1; + } + + u->port = (uint16_t) v; + } + + return 0; +} + +void +http_parser_pause(http_parser *parser, int paused) { + /* Users should only be pausing/unpausing a parser that is not in an error + * state. In non-debug builds, there's not much that we can do about this + * other than ignore it. + */ + if (HTTP_PARSER_ERRNO(parser) == HPE_OK || + HTTP_PARSER_ERRNO(parser) == HPE_PAUSED) { + SET_ERRNO((paused) ? HPE_PAUSED : HPE_OK); + } else { + assert(0 && "Attempting to pause parser in error state"); + } +} + +int +http_body_is_final(const struct http_parser *parser) { + return parser->state == s_message_done; +} diff --git a/ext/http-parser/http_parser.gyp b/ext/http-parser/http_parser.gyp new file mode 100644 index 00000000..ef34ecae --- /dev/null +++ b/ext/http-parser/http_parser.gyp @@ -0,0 +1,111 @@ +# This file is used with the GYP meta build system. +# http://code.google.com/p/gyp/ +# To build try this: +# svn co http://gyp.googlecode.com/svn/trunk gyp +# ./gyp/gyp -f make --depth=`pwd` http_parser.gyp +# ./out/Debug/test +{ + 'target_defaults': { + 'default_configuration': 'Debug', + 'configurations': { + # TODO: hoist these out and put them somewhere common, because + # RuntimeLibrary MUST MATCH across the entire project + 'Debug': { + 'defines': [ 'DEBUG', '_DEBUG' ], + 'cflags': [ '-Wall', '-Wextra', '-O0', '-g', '-ftrapv' ], + 'msvs_settings': { + 'VCCLCompilerTool': { + 'RuntimeLibrary': 1, # static debug + }, + }, + }, + 'Release': { + 'defines': [ 'NDEBUG' ], + 'cflags': [ '-Wall', '-Wextra', '-O3' ], + 'msvs_settings': { + 'VCCLCompilerTool': { + 'RuntimeLibrary': 0, # static release + }, + }, + } + }, + 'msvs_settings': { + 'VCCLCompilerTool': { + }, + 'VCLibrarianTool': { + }, + 'VCLinkerTool': { + 'GenerateDebugInformation': 'true', + }, + }, + 'conditions': [ + ['OS == "win"', { + 'defines': [ + 'WIN32' + ], + }] + ], + }, + + 'targets': [ + { + 'target_name': 'http_parser', + 'type': 'static_library', + 'include_dirs': [ '.' ], + 'direct_dependent_settings': { + 'defines': [ 'HTTP_PARSER_STRICT=0' ], + 'include_dirs': [ '.' ], + }, + 'defines': [ 'HTTP_PARSER_STRICT=0' ], + 'sources': [ './http_parser.c', ], + 'conditions': [ + ['OS=="win"', { + 'msvs_settings': { + 'VCCLCompilerTool': { + # Compile as C++. http_parser.c is actually C99, but C++ is + # close enough in this case. + 'CompileAs': 2, + }, + }, + }] + ], + }, + + { + 'target_name': 'http_parser_strict', + 'type': 'static_library', + 'include_dirs': [ '.' ], + 'direct_dependent_settings': { + 'defines': [ 'HTTP_PARSER_STRICT=1' ], + 'include_dirs': [ '.' ], + }, + 'defines': [ 'HTTP_PARSER_STRICT=1' ], + 'sources': [ './http_parser.c', ], + 'conditions': [ + ['OS=="win"', { + 'msvs_settings': { + 'VCCLCompilerTool': { + # Compile as C++. http_parser.c is actually C99, but C++ is + # close enough in this case. + 'CompileAs': 2, + }, + }, + }] + ], + }, + + { + 'target_name': 'test-nonstrict', + 'type': 'executable', + 'dependencies': [ 'http_parser' ], + 'sources': [ 'test.c' ] + }, + + { + 'target_name': 'test-strict', + 'type': 'executable', + 'dependencies': [ 'http_parser_strict' ], + 'sources': [ 'test.c' ] + } + ] +} diff --git a/ext/http-parser/http_parser.h b/ext/http-parser/http_parser.h new file mode 100644 index 00000000..a992c742 --- /dev/null +++ b/ext/http-parser/http_parser.h @@ -0,0 +1,304 @@ +/* Copyright Joyent, Inc. and other Node contributors. All rights reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ +#ifndef http_parser_h +#define http_parser_h +#ifdef __cplusplus +extern "C" { +#endif + +#define HTTP_PARSER_VERSION_MAJOR 2 +#define HTTP_PARSER_VERSION_MINOR 0 + +#include +#if defined(_WIN32) && !defined(__MINGW32__) && (!defined(_MSC_VER) || _MSC_VER<1600) +#include +#include +typedef __int8 int8_t; +typedef unsigned __int8 uint8_t; +typedef __int16 int16_t; +typedef unsigned __int16 uint16_t; +typedef __int32 int32_t; +typedef unsigned __int32 uint32_t; +typedef __int64 int64_t; +typedef unsigned __int64 uint64_t; +#else +#include +#endif + +/* Compile with -DHTTP_PARSER_STRICT=0 to make less checks, but run + * faster + */ +#ifndef HTTP_PARSER_STRICT +# define HTTP_PARSER_STRICT 1 +#endif + +/* Maximium header size allowed */ +#define HTTP_MAX_HEADER_SIZE (80*1024) + + +typedef struct http_parser http_parser; +typedef struct http_parser_settings http_parser_settings; + + +/* Callbacks should return non-zero to indicate an error. The parser will + * then halt execution. + * + * The one exception is on_headers_complete. In a HTTP_RESPONSE parser + * returning '1' from on_headers_complete will tell the parser that it + * should not expect a body. This is used when receiving a response to a + * HEAD request which may contain 'Content-Length' or 'Transfer-Encoding: + * chunked' headers that indicate the presence of a body. + * + * http_data_cb does not return data chunks. It will be call arbitrarally + * many times for each string. E.G. you might get 10 callbacks for "on_url" + * each providing just a few characters more data. + */ +typedef int (*http_data_cb) (http_parser*, const char *at, size_t length); +typedef int (*http_cb) (http_parser*); + + +/* Request Methods */ +#define HTTP_METHOD_MAP(XX) \ + XX(0, DELETE, DELETE) \ + XX(1, GET, GET) \ + XX(2, HEAD, HEAD) \ + XX(3, POST, POST) \ + XX(4, PUT, PUT) \ + /* pathological */ \ + XX(5, CONNECT, CONNECT) \ + XX(6, OPTIONS, OPTIONS) \ + XX(7, TRACE, TRACE) \ + /* webdav */ \ + XX(8, COPY, COPY) \ + XX(9, LOCK, LOCK) \ + XX(10, MKCOL, MKCOL) \ + XX(11, MOVE, MOVE) \ + XX(12, PROPFIND, PROPFIND) \ + XX(13, PROPPATCH, PROPPATCH) \ + XX(14, SEARCH, SEARCH) \ + XX(15, UNLOCK, UNLOCK) \ + /* subversion */ \ + XX(16, REPORT, REPORT) \ + XX(17, MKACTIVITY, MKACTIVITY) \ + XX(18, CHECKOUT, CHECKOUT) \ + XX(19, MERGE, MERGE) \ + /* upnp */ \ + XX(20, MSEARCH, M-SEARCH) \ + XX(21, NOTIFY, NOTIFY) \ + XX(22, SUBSCRIBE, SUBSCRIBE) \ + XX(23, UNSUBSCRIBE, UNSUBSCRIBE) \ + /* RFC-5789 */ \ + XX(24, PATCH, PATCH) \ + XX(25, PURGE, PURGE) \ + +enum http_method + { +#define XX(num, name, string) HTTP_##name = num, + HTTP_METHOD_MAP(XX) +#undef XX + }; + + +enum http_parser_type { HTTP_REQUEST, HTTP_RESPONSE, HTTP_BOTH }; + + +/* Flag values for http_parser.flags field */ +enum flags + { F_CHUNKED = 1 << 0 + , F_CONNECTION_KEEP_ALIVE = 1 << 1 + , F_CONNECTION_CLOSE = 1 << 2 + , F_TRAILING = 1 << 3 + , F_UPGRADE = 1 << 4 + , F_SKIPBODY = 1 << 5 + }; + + +/* Map for errno-related constants + * + * The provided argument should be a macro that takes 2 arguments. + */ +#define HTTP_ERRNO_MAP(XX) \ + /* No error */ \ + XX(OK, "success") \ + \ + /* Callback-related errors */ \ + XX(CB_message_begin, "the on_message_begin callback failed") \ + XX(CB_status_complete, "the on_status_complete callback failed") \ + XX(CB_url, "the on_url callback failed") \ + XX(CB_header_field, "the on_header_field callback failed") \ + XX(CB_header_value, "the on_header_value callback failed") \ + XX(CB_headers_complete, "the on_headers_complete callback failed") \ + XX(CB_body, "the on_body callback failed") \ + XX(CB_message_complete, "the on_message_complete callback failed") \ + \ + /* Parsing-related errors */ \ + XX(INVALID_EOF_STATE, "stream ended at an unexpected time") \ + XX(HEADER_OVERFLOW, \ + "too many header bytes seen; overflow detected") \ + XX(CLOSED_CONNECTION, \ + "data received after completed connection: close message") \ + XX(INVALID_VERSION, "invalid HTTP version") \ + XX(INVALID_STATUS, "invalid HTTP status code") \ + XX(INVALID_METHOD, "invalid HTTP method") \ + XX(INVALID_URL, "invalid URL") \ + XX(INVALID_HOST, "invalid host") \ + XX(INVALID_PORT, "invalid port") \ + XX(INVALID_PATH, "invalid path") \ + XX(INVALID_QUERY_STRING, "invalid query string") \ + XX(INVALID_FRAGMENT, "invalid fragment") \ + XX(LF_EXPECTED, "LF character expected") \ + XX(INVALID_HEADER_TOKEN, "invalid character in header") \ + XX(INVALID_CONTENT_LENGTH, \ + "invalid character in content-length header") \ + XX(INVALID_CHUNK_SIZE, \ + "invalid character in chunk size header") \ + XX(INVALID_CONSTANT, "invalid constant string") \ + XX(INVALID_INTERNAL_STATE, "encountered unexpected internal state")\ + XX(STRICT, "strict mode assertion failed") \ + XX(PAUSED, "parser is paused") \ + XX(UNKNOWN, "an unknown error occurred") + + +/* Define HPE_* values for each errno value above */ +#define HTTP_ERRNO_GEN(n, s) HPE_##n, +enum http_errno { + HTTP_ERRNO_MAP(HTTP_ERRNO_GEN) +}; +#undef HTTP_ERRNO_GEN + + +/* Get an http_errno value from an http_parser */ +#define HTTP_PARSER_ERRNO(p) ((enum http_errno) (p)->http_errno) + + +struct http_parser { + /** PRIVATE **/ + unsigned char type : 2; /* enum http_parser_type */ + unsigned char flags : 6; /* F_* values from 'flags' enum; semi-public */ + unsigned char state; /* enum state from http_parser.c */ + unsigned char header_state; /* enum header_state from http_parser.c */ + unsigned char index; /* index into current matcher */ + + uint32_t nread; /* # bytes read in various scenarios */ + uint64_t content_length; /* # bytes in body (0 if no Content-Length header) */ + + /** READ-ONLY **/ + unsigned short http_major; + unsigned short http_minor; + unsigned short status_code; /* responses only */ + unsigned char method; /* requests only */ + unsigned char http_errno : 7; + + /* 1 = Upgrade header was present and the parser has exited because of that. + * 0 = No upgrade header present. + * Should be checked when http_parser_execute() returns in addition to + * error checking. + */ + unsigned char upgrade : 1; + + /** PUBLIC **/ + void *data; /* A pointer to get hook to the "connection" or "socket" object */ +}; + + +struct http_parser_settings { + http_cb on_message_begin; + http_data_cb on_url; + http_cb on_status_complete; + http_data_cb on_header_field; + http_data_cb on_header_value; + http_cb on_headers_complete; + http_data_cb on_body; + http_cb on_message_complete; +}; + + +enum http_parser_url_fields + { UF_SCHEMA = 0 + , UF_HOST = 1 + , UF_PORT = 2 + , UF_PATH = 3 + , UF_QUERY = 4 + , UF_FRAGMENT = 5 + , UF_USERINFO = 6 + , UF_MAX = 7 + }; + + +/* Result structure for http_parser_parse_url(). + * + * Callers should index into field_data[] with UF_* values iff field_set + * has the relevant (1 << UF_*) bit set. As a courtesy to clients (and + * because we probably have padding left over), we convert any port to + * a uint16_t. + */ +struct http_parser_url { + uint16_t field_set; /* Bitmask of (1 << UF_*) values */ + uint16_t port; /* Converted UF_PORT string */ + + struct { + uint16_t off; /* Offset into buffer in which field starts */ + uint16_t len; /* Length of run in buffer */ + } field_data[UF_MAX]; +}; + + +void http_parser_init(http_parser *parser, enum http_parser_type type); + + +size_t http_parser_execute(http_parser *parser, + const http_parser_settings *settings, + const char *data, + size_t len); + + +/* If http_should_keep_alive() in the on_headers_complete or + * on_message_complete callback returns 0, then this should be + * the last message on the connection. + * If you are the server, respond with the "Connection: close" header. + * If you are the client, close the connection. + */ +int http_should_keep_alive(const http_parser *parser); + +/* Returns a string version of the HTTP method. */ +const char *http_method_str(enum http_method m); + +/* Return a string name of the given error */ +const char *http_errno_name(enum http_errno err); + +/* Return a string description of the given error */ +const char *http_errno_description(enum http_errno err); + +/* Parse a URL; return nonzero on failure */ +int http_parser_parse_url(const char *buf, size_t buflen, + int is_connect, + struct http_parser_url *u); + +/* Pause or un-pause the parser; a nonzero value pauses */ +void http_parser_pause(http_parser *parser, int paused); + +/* Checks if this is the final chunk of the body. */ +int http_body_is_final(const http_parser *parser); + +#ifdef __cplusplus +} +#endif +#endif diff --git a/ext/http-parser/test.c b/ext/http-parser/test.c new file mode 100644 index 00000000..83723b7f --- /dev/null +++ b/ext/http-parser/test.c @@ -0,0 +1,3425 @@ +/* Copyright Joyent, Inc. and other Node contributors. All rights reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ +#include "http_parser.h" +#include +#include +#include +#include /* rand */ +#include +#include + +#undef TRUE +#define TRUE 1 +#undef FALSE +#define FALSE 0 + +#define MAX_HEADERS 13 +#define MAX_ELEMENT_SIZE 2048 + +#define MIN(a,b) ((a) < (b) ? (a) : (b)) + +static http_parser *parser; + +struct message { + const char *name; // for debugging purposes + const char *raw; + enum http_parser_type type; + enum http_method method; + int status_code; + char request_path[MAX_ELEMENT_SIZE]; + char request_url[MAX_ELEMENT_SIZE]; + char fragment[MAX_ELEMENT_SIZE]; + char query_string[MAX_ELEMENT_SIZE]; + char body[MAX_ELEMENT_SIZE]; + size_t body_size; + const char *host; + const char *userinfo; + uint16_t port; + int num_headers; + enum { NONE=0, FIELD, VALUE } last_header_element; + char headers [MAX_HEADERS][2][MAX_ELEMENT_SIZE]; + int should_keep_alive; + + const char *upgrade; // upgraded body + + unsigned short http_major; + unsigned short http_minor; + + int message_begin_cb_called; + int headers_complete_cb_called; + int message_complete_cb_called; + int message_complete_on_eof; + int body_is_final; +}; + +static int currently_parsing_eof; + +static struct message messages[5]; +static int num_messages; +static http_parser_settings *current_pause_parser; + +/* * R E Q U E S T S * */ +const struct message requests[] = +#define CURL_GET 0 +{ {.name= "curl get" + ,.type= HTTP_REQUEST + ,.raw= "GET /test HTTP/1.1\r\n" + "User-Agent: curl/7.18.0 (i486-pc-linux-gnu) libcurl/7.18.0 OpenSSL/0.9.8g zlib/1.2.3.3 libidn/1.1\r\n" + "Host: 0.0.0.0=5000\r\n" + "Accept: */*\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_GET + ,.query_string= "" + ,.fragment= "" + ,.request_path= "/test" + ,.request_url= "/test" + ,.num_headers= 3 + ,.headers= + { { "User-Agent", "curl/7.18.0 (i486-pc-linux-gnu) libcurl/7.18.0 OpenSSL/0.9.8g zlib/1.2.3.3 libidn/1.1" } + , { "Host", "0.0.0.0=5000" } + , { "Accept", "*/*" } + } + ,.body= "" + } + +#define FIREFOX_GET 1 +, {.name= "firefox get" + ,.type= HTTP_REQUEST + ,.raw= "GET /favicon.ico HTTP/1.1\r\n" + "Host: 0.0.0.0=5000\r\n" + "User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9) Gecko/2008061015 Firefox/3.0\r\n" + "Accept: text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8\r\n" + "Accept-Language: en-us,en;q=0.5\r\n" + "Accept-Encoding: gzip,deflate\r\n" + "Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7\r\n" + "Keep-Alive: 300\r\n" + "Connection: keep-alive\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_GET + ,.query_string= "" + ,.fragment= "" + ,.request_path= "/favicon.ico" + ,.request_url= "/favicon.ico" + ,.num_headers= 8 + ,.headers= + { { "Host", "0.0.0.0=5000" } + , { "User-Agent", "Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9) Gecko/2008061015 Firefox/3.0" } + , { "Accept", "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8" } + , { "Accept-Language", "en-us,en;q=0.5" } + , { "Accept-Encoding", "gzip,deflate" } + , { "Accept-Charset", "ISO-8859-1,utf-8;q=0.7,*;q=0.7" } + , { "Keep-Alive", "300" } + , { "Connection", "keep-alive" } + } + ,.body= "" + } + +#define DUMBFUCK 2 +, {.name= "dumbfuck" + ,.type= HTTP_REQUEST + ,.raw= "GET /dumbfuck HTTP/1.1\r\n" + "aaaaaaaaaaaaa:++++++++++\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_GET + ,.query_string= "" + ,.fragment= "" + ,.request_path= "/dumbfuck" + ,.request_url= "/dumbfuck" + ,.num_headers= 1 + ,.headers= + { { "aaaaaaaaaaaaa", "++++++++++" } + } + ,.body= "" + } + +#define FRAGMENT_IN_URI 3 +, {.name= "fragment in url" + ,.type= HTTP_REQUEST + ,.raw= "GET /forums/1/topics/2375?page=1#posts-17408 HTTP/1.1\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_GET + ,.query_string= "page=1" + ,.fragment= "posts-17408" + ,.request_path= "/forums/1/topics/2375" + /* XXX request url does include fragment? */ + ,.request_url= "/forums/1/topics/2375?page=1#posts-17408" + ,.num_headers= 0 + ,.body= "" + } + +#define GET_NO_HEADERS_NO_BODY 4 +, {.name= "get no headers no body" + ,.type= HTTP_REQUEST + ,.raw= "GET /get_no_headers_no_body/world HTTP/1.1\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE /* would need Connection: close */ + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_GET + ,.query_string= "" + ,.fragment= "" + ,.request_path= "/get_no_headers_no_body/world" + ,.request_url= "/get_no_headers_no_body/world" + ,.num_headers= 0 + ,.body= "" + } + +#define GET_ONE_HEADER_NO_BODY 5 +, {.name= "get one header no body" + ,.type= HTTP_REQUEST + ,.raw= "GET /get_one_header_no_body HTTP/1.1\r\n" + "Accept: */*\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE /* would need Connection: close */ + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_GET + ,.query_string= "" + ,.fragment= "" + ,.request_path= "/get_one_header_no_body" + ,.request_url= "/get_one_header_no_body" + ,.num_headers= 1 + ,.headers= + { { "Accept" , "*/*" } + } + ,.body= "" + } + +#define GET_FUNKY_CONTENT_LENGTH 6 +, {.name= "get funky content length body hello" + ,.type= HTTP_REQUEST + ,.raw= "GET /get_funky_content_length_body_hello HTTP/1.0\r\n" + "conTENT-Length: 5\r\n" + "\r\n" + "HELLO" + ,.should_keep_alive= FALSE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 0 + ,.method= HTTP_GET + ,.query_string= "" + ,.fragment= "" + ,.request_path= "/get_funky_content_length_body_hello" + ,.request_url= "/get_funky_content_length_body_hello" + ,.num_headers= 1 + ,.headers= + { { "conTENT-Length" , "5" } + } + ,.body= "HELLO" + } + +#define POST_IDENTITY_BODY_WORLD 7 +, {.name= "post identity body world" + ,.type= HTTP_REQUEST + ,.raw= "POST /post_identity_body_world?q=search#hey HTTP/1.1\r\n" + "Accept: */*\r\n" + "Transfer-Encoding: identity\r\n" + "Content-Length: 5\r\n" + "\r\n" + "World" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_POST + ,.query_string= "q=search" + ,.fragment= "hey" + ,.request_path= "/post_identity_body_world" + ,.request_url= "/post_identity_body_world?q=search#hey" + ,.num_headers= 3 + ,.headers= + { { "Accept", "*/*" } + , { "Transfer-Encoding", "identity" } + , { "Content-Length", "5" } + } + ,.body= "World" + } + +#define POST_CHUNKED_ALL_YOUR_BASE 8 +, {.name= "post - chunked body: all your base are belong to us" + ,.type= HTTP_REQUEST + ,.raw= "POST /post_chunked_all_your_base HTTP/1.1\r\n" + "Transfer-Encoding: chunked\r\n" + "\r\n" + "1e\r\nall your base are belong to us\r\n" + "0\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_POST + ,.query_string= "" + ,.fragment= "" + ,.request_path= "/post_chunked_all_your_base" + ,.request_url= "/post_chunked_all_your_base" + ,.num_headers= 1 + ,.headers= + { { "Transfer-Encoding" , "chunked" } + } + ,.body= "all your base are belong to us" + } + +#define TWO_CHUNKS_MULT_ZERO_END 9 +, {.name= "two chunks ; triple zero ending" + ,.type= HTTP_REQUEST + ,.raw= "POST /two_chunks_mult_zero_end HTTP/1.1\r\n" + "Transfer-Encoding: chunked\r\n" + "\r\n" + "5\r\nhello\r\n" + "6\r\n world\r\n" + "000\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_POST + ,.query_string= "" + ,.fragment= "" + ,.request_path= "/two_chunks_mult_zero_end" + ,.request_url= "/two_chunks_mult_zero_end" + ,.num_headers= 1 + ,.headers= + { { "Transfer-Encoding", "chunked" } + } + ,.body= "hello world" + } + +#define CHUNKED_W_TRAILING_HEADERS 10 +, {.name= "chunked with trailing headers. blech." + ,.type= HTTP_REQUEST + ,.raw= "POST /chunked_w_trailing_headers HTTP/1.1\r\n" + "Transfer-Encoding: chunked\r\n" + "\r\n" + "5\r\nhello\r\n" + "6\r\n world\r\n" + "0\r\n" + "Vary: *\r\n" + "Content-Type: text/plain\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_POST + ,.query_string= "" + ,.fragment= "" + ,.request_path= "/chunked_w_trailing_headers" + ,.request_url= "/chunked_w_trailing_headers" + ,.num_headers= 3 + ,.headers= + { { "Transfer-Encoding", "chunked" } + , { "Vary", "*" } + , { "Content-Type", "text/plain" } + } + ,.body= "hello world" + } + +#define CHUNKED_W_BULLSHIT_AFTER_LENGTH 11 +, {.name= "with bullshit after the length" + ,.type= HTTP_REQUEST + ,.raw= "POST /chunked_w_bullshit_after_length HTTP/1.1\r\n" + "Transfer-Encoding: chunked\r\n" + "\r\n" + "5; ihatew3;whatthefuck=aretheseparametersfor\r\nhello\r\n" + "6; blahblah; blah\r\n world\r\n" + "0\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_POST + ,.query_string= "" + ,.fragment= "" + ,.request_path= "/chunked_w_bullshit_after_length" + ,.request_url= "/chunked_w_bullshit_after_length" + ,.num_headers= 1 + ,.headers= + { { "Transfer-Encoding", "chunked" } + } + ,.body= "hello world" + } + +#define WITH_QUOTES 12 +, {.name= "with quotes" + ,.type= HTTP_REQUEST + ,.raw= "GET /with_\"stupid\"_quotes?foo=\"bar\" HTTP/1.1\r\n\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_GET + ,.query_string= "foo=\"bar\"" + ,.fragment= "" + ,.request_path= "/with_\"stupid\"_quotes" + ,.request_url= "/with_\"stupid\"_quotes?foo=\"bar\"" + ,.num_headers= 0 + ,.headers= { } + ,.body= "" + } + +#define APACHEBENCH_GET 13 +/* The server receiving this request SHOULD NOT wait for EOF + * to know that content-length == 0. + * How to represent this in a unit test? message_complete_on_eof + * Compare with NO_CONTENT_LENGTH_RESPONSE. + */ +, {.name = "apachebench get" + ,.type= HTTP_REQUEST + ,.raw= "GET /test HTTP/1.0\r\n" + "Host: 0.0.0.0:5000\r\n" + "User-Agent: ApacheBench/2.3\r\n" + "Accept: */*\r\n\r\n" + ,.should_keep_alive= FALSE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 0 + ,.method= HTTP_GET + ,.query_string= "" + ,.fragment= "" + ,.request_path= "/test" + ,.request_url= "/test" + ,.num_headers= 3 + ,.headers= { { "Host", "0.0.0.0:5000" } + , { "User-Agent", "ApacheBench/2.3" } + , { "Accept", "*/*" } + } + ,.body= "" + } + +#define QUERY_URL_WITH_QUESTION_MARK_GET 14 +/* Some clients include '?' characters in query strings. + */ +, {.name = "query url with question mark" + ,.type= HTTP_REQUEST + ,.raw= "GET /test.cgi?foo=bar?baz HTTP/1.1\r\n\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_GET + ,.query_string= "foo=bar?baz" + ,.fragment= "" + ,.request_path= "/test.cgi" + ,.request_url= "/test.cgi?foo=bar?baz" + ,.num_headers= 0 + ,.headers= {} + ,.body= "" + } + +#define PREFIX_NEWLINE_GET 15 +/* Some clients, especially after a POST in a keep-alive connection, + * will send an extra CRLF before the next request + */ +, {.name = "newline prefix get" + ,.type= HTTP_REQUEST + ,.raw= "\r\nGET /test HTTP/1.1\r\n\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_GET + ,.query_string= "" + ,.fragment= "" + ,.request_path= "/test" + ,.request_url= "/test" + ,.num_headers= 0 + ,.headers= { } + ,.body= "" + } + +#define UPGRADE_REQUEST 16 +, {.name = "upgrade request" + ,.type= HTTP_REQUEST + ,.raw= "GET /demo HTTP/1.1\r\n" + "Host: example.com\r\n" + "Connection: Upgrade\r\n" + "Sec-WebSocket-Key2: 12998 5 Y3 1 .P00\r\n" + "Sec-WebSocket-Protocol: sample\r\n" + "Upgrade: WebSocket\r\n" + "Sec-WebSocket-Key1: 4 @1 46546xW%0l 1 5\r\n" + "Origin: http://example.com\r\n" + "\r\n" + "Hot diggity dogg" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_GET + ,.query_string= "" + ,.fragment= "" + ,.request_path= "/demo" + ,.request_url= "/demo" + ,.num_headers= 7 + ,.upgrade="Hot diggity dogg" + ,.headers= { { "Host", "example.com" } + , { "Connection", "Upgrade" } + , { "Sec-WebSocket-Key2", "12998 5 Y3 1 .P00" } + , { "Sec-WebSocket-Protocol", "sample" } + , { "Upgrade", "WebSocket" } + , { "Sec-WebSocket-Key1", "4 @1 46546xW%0l 1 5" } + , { "Origin", "http://example.com" } + } + ,.body= "" + } + +#define CONNECT_REQUEST 17 +, {.name = "connect request" + ,.type= HTTP_REQUEST + ,.raw= "CONNECT 0-home0.netscape.com:443 HTTP/1.0\r\n" + "User-agent: Mozilla/1.1N\r\n" + "Proxy-authorization: basic aGVsbG86d29ybGQ=\r\n" + "\r\n" + "some data\r\n" + "and yet even more data" + ,.should_keep_alive= FALSE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 0 + ,.method= HTTP_CONNECT + ,.query_string= "" + ,.fragment= "" + ,.request_path= "" + ,.request_url= "0-home0.netscape.com:443" + ,.num_headers= 2 + ,.upgrade="some data\r\nand yet even more data" + ,.headers= { { "User-agent", "Mozilla/1.1N" } + , { "Proxy-authorization", "basic aGVsbG86d29ybGQ=" } + } + ,.body= "" + } + +#define REPORT_REQ 18 +, {.name= "report request" + ,.type= HTTP_REQUEST + ,.raw= "REPORT /test HTTP/1.1\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_REPORT + ,.query_string= "" + ,.fragment= "" + ,.request_path= "/test" + ,.request_url= "/test" + ,.num_headers= 0 + ,.headers= {} + ,.body= "" + } + +#define NO_HTTP_VERSION 19 +, {.name= "request with no http version" + ,.type= HTTP_REQUEST + ,.raw= "GET /\r\n" + "\r\n" + ,.should_keep_alive= FALSE + ,.message_complete_on_eof= FALSE + ,.http_major= 0 + ,.http_minor= 9 + ,.method= HTTP_GET + ,.query_string= "" + ,.fragment= "" + ,.request_path= "/" + ,.request_url= "/" + ,.num_headers= 0 + ,.headers= {} + ,.body= "" + } + +#define MSEARCH_REQ 20 +, {.name= "m-search request" + ,.type= HTTP_REQUEST + ,.raw= "M-SEARCH * HTTP/1.1\r\n" + "HOST: 239.255.255.250:1900\r\n" + "MAN: \"ssdp:discover\"\r\n" + "ST: \"ssdp:all\"\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_MSEARCH + ,.query_string= "" + ,.fragment= "" + ,.request_path= "*" + ,.request_url= "*" + ,.num_headers= 3 + ,.headers= { { "HOST", "239.255.255.250:1900" } + , { "MAN", "\"ssdp:discover\"" } + , { "ST", "\"ssdp:all\"" } + } + ,.body= "" + } + +#define LINE_FOLDING_IN_HEADER 21 +, {.name= "line folding in header value" + ,.type= HTTP_REQUEST + ,.raw= "GET / HTTP/1.1\r\n" + "Line1: abc\r\n" + "\tdef\r\n" + " ghi\r\n" + "\t\tjkl\r\n" + " mno \r\n" + "\t \tqrs\r\n" + "Line2: \t line2\t\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_GET + ,.query_string= "" + ,.fragment= "" + ,.request_path= "/" + ,.request_url= "/" + ,.num_headers= 2 + ,.headers= { { "Line1", "abcdefghijklmno qrs" } + , { "Line2", "line2\t" } + } + ,.body= "" + } + + +#define QUERY_TERMINATED_HOST 22 +, {.name= "host terminated by a query string" + ,.type= HTTP_REQUEST + ,.raw= "GET http://hypnotoad.org?hail=all HTTP/1.1\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_GET + ,.query_string= "hail=all" + ,.fragment= "" + ,.request_path= "" + ,.request_url= "http://hypnotoad.org?hail=all" + ,.host= "hypnotoad.org" + ,.num_headers= 0 + ,.headers= { } + ,.body= "" + } + +#define QUERY_TERMINATED_HOSTPORT 23 +, {.name= "host:port terminated by a query string" + ,.type= HTTP_REQUEST + ,.raw= "GET http://hypnotoad.org:1234?hail=all HTTP/1.1\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_GET + ,.query_string= "hail=all" + ,.fragment= "" + ,.request_path= "" + ,.request_url= "http://hypnotoad.org:1234?hail=all" + ,.host= "hypnotoad.org" + ,.port= 1234 + ,.num_headers= 0 + ,.headers= { } + ,.body= "" + } + +#define SPACE_TERMINATED_HOSTPORT 24 +, {.name= "host:port terminated by a space" + ,.type= HTTP_REQUEST + ,.raw= "GET http://hypnotoad.org:1234 HTTP/1.1\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_GET + ,.query_string= "" + ,.fragment= "" + ,.request_path= "" + ,.request_url= "http://hypnotoad.org:1234" + ,.host= "hypnotoad.org" + ,.port= 1234 + ,.num_headers= 0 + ,.headers= { } + ,.body= "" + } + +#define PATCH_REQ 25 +, {.name = "PATCH request" + ,.type= HTTP_REQUEST + ,.raw= "PATCH /file.txt HTTP/1.1\r\n" + "Host: www.example.com\r\n" + "Content-Type: application/example\r\n" + "If-Match: \"e0023aa4e\"\r\n" + "Content-Length: 10\r\n" + "\r\n" + "cccccccccc" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_PATCH + ,.query_string= "" + ,.fragment= "" + ,.request_path= "/file.txt" + ,.request_url= "/file.txt" + ,.num_headers= 4 + ,.headers= { { "Host", "www.example.com" } + , { "Content-Type", "application/example" } + , { "If-Match", "\"e0023aa4e\"" } + , { "Content-Length", "10" } + } + ,.body= "cccccccccc" + } + +#define CONNECT_CAPS_REQUEST 26 +, {.name = "connect caps request" + ,.type= HTTP_REQUEST + ,.raw= "CONNECT HOME0.NETSCAPE.COM:443 HTTP/1.0\r\n" + "User-agent: Mozilla/1.1N\r\n" + "Proxy-authorization: basic aGVsbG86d29ybGQ=\r\n" + "\r\n" + ,.should_keep_alive= FALSE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 0 + ,.method= HTTP_CONNECT + ,.query_string= "" + ,.fragment= "" + ,.request_path= "" + ,.request_url= "HOME0.NETSCAPE.COM:443" + ,.num_headers= 2 + ,.upgrade="" + ,.headers= { { "User-agent", "Mozilla/1.1N" } + , { "Proxy-authorization", "basic aGVsbG86d29ybGQ=" } + } + ,.body= "" + } + +#if !HTTP_PARSER_STRICT +#define UTF8_PATH_REQ 27 +, {.name= "utf-8 path request" + ,.type= HTTP_REQUEST + ,.raw= "GET /δ¶/δt/pope?q=1#narf HTTP/1.1\r\n" + "Host: github.com\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_GET + ,.query_string= "q=1" + ,.fragment= "narf" + ,.request_path= "/δ¶/δt/pope" + ,.request_url= "/δ¶/δt/pope?q=1#narf" + ,.num_headers= 1 + ,.headers= { {"Host", "github.com" } + } + ,.body= "" + } + +#define HOSTNAME_UNDERSCORE 28 +, {.name = "hostname underscore" + ,.type= HTTP_REQUEST + ,.raw= "CONNECT home_0.netscape.com:443 HTTP/1.0\r\n" + "User-agent: Mozilla/1.1N\r\n" + "Proxy-authorization: basic aGVsbG86d29ybGQ=\r\n" + "\r\n" + ,.should_keep_alive= FALSE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 0 + ,.method= HTTP_CONNECT + ,.query_string= "" + ,.fragment= "" + ,.request_path= "" + ,.request_url= "home_0.netscape.com:443" + ,.num_headers= 2 + ,.upgrade="" + ,.headers= { { "User-agent", "Mozilla/1.1N" } + , { "Proxy-authorization", "basic aGVsbG86d29ybGQ=" } + } + ,.body= "" + } +#endif /* !HTTP_PARSER_STRICT */ + +/* see https://github.com/ry/http-parser/issues/47 */ +#define EAT_TRAILING_CRLF_NO_CONNECTION_CLOSE 29 +, {.name = "eat CRLF between requests, no \"Connection: close\" header" + ,.raw= "POST / HTTP/1.1\r\n" + "Host: www.example.com\r\n" + "Content-Type: application/x-www-form-urlencoded\r\n" + "Content-Length: 4\r\n" + "\r\n" + "q=42\r\n" /* note the trailing CRLF */ + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_POST + ,.query_string= "" + ,.fragment= "" + ,.request_path= "/" + ,.request_url= "/" + ,.num_headers= 3 + ,.upgrade= 0 + ,.headers= { { "Host", "www.example.com" } + , { "Content-Type", "application/x-www-form-urlencoded" } + , { "Content-Length", "4" } + } + ,.body= "q=42" + } + +/* see https://github.com/ry/http-parser/issues/47 */ +#define EAT_TRAILING_CRLF_WITH_CONNECTION_CLOSE 30 +, {.name = "eat CRLF between requests even if \"Connection: close\" is set" + ,.raw= "POST / HTTP/1.1\r\n" + "Host: www.example.com\r\n" + "Content-Type: application/x-www-form-urlencoded\r\n" + "Content-Length: 4\r\n" + "Connection: close\r\n" + "\r\n" + "q=42\r\n" /* note the trailing CRLF */ + ,.should_keep_alive= FALSE + ,.message_complete_on_eof= FALSE /* input buffer isn't empty when on_message_complete is called */ + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_POST + ,.query_string= "" + ,.fragment= "" + ,.request_path= "/" + ,.request_url= "/" + ,.num_headers= 4 + ,.upgrade= 0 + ,.headers= { { "Host", "www.example.com" } + , { "Content-Type", "application/x-www-form-urlencoded" } + , { "Content-Length", "4" } + , { "Connection", "close" } + } + ,.body= "q=42" + } + +#define PURGE_REQ 31 +, {.name = "PURGE request" + ,.type= HTTP_REQUEST + ,.raw= "PURGE /file.txt HTTP/1.1\r\n" + "Host: www.example.com\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_PURGE + ,.query_string= "" + ,.fragment= "" + ,.request_path= "/file.txt" + ,.request_url= "/file.txt" + ,.num_headers= 1 + ,.headers= { { "Host", "www.example.com" } } + ,.body= "" + } + +#define SEARCH_REQ 32 +, {.name = "SEARCH request" + ,.type= HTTP_REQUEST + ,.raw= "SEARCH / HTTP/1.1\r\n" + "Host: www.example.com\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_SEARCH + ,.query_string= "" + ,.fragment= "" + ,.request_path= "/" + ,.request_url= "/" + ,.num_headers= 1 + ,.headers= { { "Host", "www.example.com" } } + ,.body= "" + } + +#define PROXY_WITH_BASIC_AUTH 33 +, {.name= "host:port and basic_auth" + ,.type= HTTP_REQUEST + ,.raw= "GET http://a%12:b!&*$@hypnotoad.org:1234/toto HTTP/1.1\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.method= HTTP_GET + ,.fragment= "" + ,.request_path= "/toto" + ,.request_url= "http://a%12:b!&*$@hypnotoad.org:1234/toto" + ,.host= "hypnotoad.org" + ,.userinfo= "a%12:b!&*$" + ,.port= 1234 + ,.num_headers= 0 + ,.headers= { } + ,.body= "" + } + + +, {.name= NULL } /* sentinel */ +}; + +/* * R E S P O N S E S * */ +const struct message responses[] = +#define GOOGLE_301 0 +{ {.name= "google 301" + ,.type= HTTP_RESPONSE + ,.raw= "HTTP/1.1 301 Moved Permanently\r\n" + "Location: http://www.google.com/\r\n" + "Content-Type: text/html; charset=UTF-8\r\n" + "Date: Sun, 26 Apr 2009 11:11:49 GMT\r\n" + "Expires: Tue, 26 May 2009 11:11:49 GMT\r\n" + "X-$PrototypeBI-Version: 1.6.0.3\r\n" /* $ char in header field */ + "Cache-Control: public, max-age=2592000\r\n" + "Server: gws\r\n" + "Content-Length: 219 \r\n" + "\r\n" + "\n" + "301 Moved\n" + "

301 Moved

\n" + "The document has moved\n" + "here.\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.status_code= 301 + ,.num_headers= 8 + ,.headers= + { { "Location", "http://www.google.com/" } + , { "Content-Type", "text/html; charset=UTF-8" } + , { "Date", "Sun, 26 Apr 2009 11:11:49 GMT" } + , { "Expires", "Tue, 26 May 2009 11:11:49 GMT" } + , { "X-$PrototypeBI-Version", "1.6.0.3" } + , { "Cache-Control", "public, max-age=2592000" } + , { "Server", "gws" } + , { "Content-Length", "219 " } + } + ,.body= "\n" + "301 Moved\n" + "

301 Moved

\n" + "The document has moved\n" + "here.\r\n" + "\r\n" + } + +#define NO_CONTENT_LENGTH_RESPONSE 1 +/* The client should wait for the server's EOF. That is, when content-length + * is not specified, and "Connection: close", the end of body is specified + * by the EOF. + * Compare with APACHEBENCH_GET + */ +, {.name= "no content-length response" + ,.type= HTTP_RESPONSE + ,.raw= "HTTP/1.1 200 OK\r\n" + "Date: Tue, 04 Aug 2009 07:59:32 GMT\r\n" + "Server: Apache\r\n" + "X-Powered-By: Servlet/2.5 JSP/2.1\r\n" + "Content-Type: text/xml; charset=utf-8\r\n" + "Connection: close\r\n" + "\r\n" + "\n" + "\n" + " \n" + " \n" + " SOAP-ENV:Client\n" + " Client Error\n" + " \n" + " \n" + "" + ,.should_keep_alive= FALSE + ,.message_complete_on_eof= TRUE + ,.http_major= 1 + ,.http_minor= 1 + ,.status_code= 200 + ,.num_headers= 5 + ,.headers= + { { "Date", "Tue, 04 Aug 2009 07:59:32 GMT" } + , { "Server", "Apache" } + , { "X-Powered-By", "Servlet/2.5 JSP/2.1" } + , { "Content-Type", "text/xml; charset=utf-8" } + , { "Connection", "close" } + } + ,.body= "\n" + "\n" + " \n" + " \n" + " SOAP-ENV:Client\n" + " Client Error\n" + " \n" + " \n" + "" + } + +#define NO_HEADERS_NO_BODY_404 2 +, {.name= "404 no headers no body" + ,.type= HTTP_RESPONSE + ,.raw= "HTTP/1.1 404 Not Found\r\n\r\n" + ,.should_keep_alive= FALSE + ,.message_complete_on_eof= TRUE + ,.http_major= 1 + ,.http_minor= 1 + ,.status_code= 404 + ,.num_headers= 0 + ,.headers= {} + ,.body_size= 0 + ,.body= "" + } + +#define NO_REASON_PHRASE 3 +, {.name= "301 no response phrase" + ,.type= HTTP_RESPONSE + ,.raw= "HTTP/1.1 301\r\n\r\n" + ,.should_keep_alive = FALSE + ,.message_complete_on_eof= TRUE + ,.http_major= 1 + ,.http_minor= 1 + ,.status_code= 301 + ,.num_headers= 0 + ,.headers= {} + ,.body= "" + } + +#define TRAILING_SPACE_ON_CHUNKED_BODY 4 +, {.name="200 trailing space on chunked body" + ,.type= HTTP_RESPONSE + ,.raw= "HTTP/1.1 200 OK\r\n" + "Content-Type: text/plain\r\n" + "Transfer-Encoding: chunked\r\n" + "\r\n" + "25 \r\n" + "This is the data in the first chunk\r\n" + "\r\n" + "1C\r\n" + "and this is the second one\r\n" + "\r\n" + "0 \r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.status_code= 200 + ,.num_headers= 2 + ,.headers= + { {"Content-Type", "text/plain" } + , {"Transfer-Encoding", "chunked" } + } + ,.body_size = 37+28 + ,.body = + "This is the data in the first chunk\r\n" + "and this is the second one\r\n" + + } + +#define NO_CARRIAGE_RET 5 +, {.name="no carriage ret" + ,.type= HTTP_RESPONSE + ,.raw= "HTTP/1.1 200 OK\n" + "Content-Type: text/html; charset=utf-8\n" + "Connection: close\n" + "\n" + "these headers are from http://news.ycombinator.com/" + ,.should_keep_alive= FALSE + ,.message_complete_on_eof= TRUE + ,.http_major= 1 + ,.http_minor= 1 + ,.status_code= 200 + ,.num_headers= 2 + ,.headers= + { {"Content-Type", "text/html; charset=utf-8" } + , {"Connection", "close" } + } + ,.body= "these headers are from http://news.ycombinator.com/" + } + +#define PROXY_CONNECTION 6 +, {.name="proxy connection" + ,.type= HTTP_RESPONSE + ,.raw= "HTTP/1.1 200 OK\r\n" + "Content-Type: text/html; charset=UTF-8\r\n" + "Content-Length: 11\r\n" + "Proxy-Connection: close\r\n" + "Date: Thu, 31 Dec 2009 20:55:48 +0000\r\n" + "\r\n" + "hello world" + ,.should_keep_alive= FALSE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.status_code= 200 + ,.num_headers= 4 + ,.headers= + { {"Content-Type", "text/html; charset=UTF-8" } + , {"Content-Length", "11" } + , {"Proxy-Connection", "close" } + , {"Date", "Thu, 31 Dec 2009 20:55:48 +0000"} + } + ,.body= "hello world" + } + +#define UNDERSTORE_HEADER_KEY 7 + // shown by + // curl -o /dev/null -v "http://ad.doubleclick.net/pfadx/DARTSHELLCONFIGXML;dcmt=text/xml;" +, {.name="underscore header key" + ,.type= HTTP_RESPONSE + ,.raw= "HTTP/1.1 200 OK\r\n" + "Server: DCLK-AdSvr\r\n" + "Content-Type: text/xml\r\n" + "Content-Length: 0\r\n" + "DCLK_imp: v7;x;114750856;0-0;0;17820020;0/0;21603567/21621457/1;;~okv=;dcmt=text/xml;;~cs=o\r\n\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.status_code= 200 + ,.num_headers= 4 + ,.headers= + { {"Server", "DCLK-AdSvr" } + , {"Content-Type", "text/xml" } + , {"Content-Length", "0" } + , {"DCLK_imp", "v7;x;114750856;0-0;0;17820020;0/0;21603567/21621457/1;;~okv=;dcmt=text/xml;;~cs=o" } + } + ,.body= "" + } + +#define BONJOUR_MADAME_FR 8 +/* The client should not merge two headers fields when the first one doesn't + * have a value. + */ +, {.name= "bonjourmadame.fr" + ,.type= HTTP_RESPONSE + ,.raw= "HTTP/1.0 301 Moved Permanently\r\n" + "Date: Thu, 03 Jun 2010 09:56:32 GMT\r\n" + "Server: Apache/2.2.3 (Red Hat)\r\n" + "Cache-Control: public\r\n" + "Pragma: \r\n" + "Location: http://www.bonjourmadame.fr/\r\n" + "Vary: Accept-Encoding\r\n" + "Content-Length: 0\r\n" + "Content-Type: text/html; charset=UTF-8\r\n" + "Connection: keep-alive\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 0 + ,.status_code= 301 + ,.num_headers= 9 + ,.headers= + { { "Date", "Thu, 03 Jun 2010 09:56:32 GMT" } + , { "Server", "Apache/2.2.3 (Red Hat)" } + , { "Cache-Control", "public" } + , { "Pragma", "" } + , { "Location", "http://www.bonjourmadame.fr/" } + , { "Vary", "Accept-Encoding" } + , { "Content-Length", "0" } + , { "Content-Type", "text/html; charset=UTF-8" } + , { "Connection", "keep-alive" } + } + ,.body= "" + } + +#define RES_FIELD_UNDERSCORE 9 +/* Should handle spaces in header fields */ +, {.name= "field underscore" + ,.type= HTTP_RESPONSE + ,.raw= "HTTP/1.1 200 OK\r\n" + "Date: Tue, 28 Sep 2010 01:14:13 GMT\r\n" + "Server: Apache\r\n" + "Cache-Control: no-cache, must-revalidate\r\n" + "Expires: Mon, 26 Jul 1997 05:00:00 GMT\r\n" + ".et-Cookie: PlaxoCS=1274804622353690521; path=/; domain=.plaxo.com\r\n" + "Vary: Accept-Encoding\r\n" + "_eep-Alive: timeout=45\r\n" /* semantic value ignored */ + "_onnection: Keep-Alive\r\n" /* semantic value ignored */ + "Transfer-Encoding: chunked\r\n" + "Content-Type: text/html\r\n" + "Connection: close\r\n" + "\r\n" + "0\r\n\r\n" + ,.should_keep_alive= FALSE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.status_code= 200 + ,.num_headers= 11 + ,.headers= + { { "Date", "Tue, 28 Sep 2010 01:14:13 GMT" } + , { "Server", "Apache" } + , { "Cache-Control", "no-cache, must-revalidate" } + , { "Expires", "Mon, 26 Jul 1997 05:00:00 GMT" } + , { ".et-Cookie", "PlaxoCS=1274804622353690521; path=/; domain=.plaxo.com" } + , { "Vary", "Accept-Encoding" } + , { "_eep-Alive", "timeout=45" } + , { "_onnection", "Keep-Alive" } + , { "Transfer-Encoding", "chunked" } + , { "Content-Type", "text/html" } + , { "Connection", "close" } + } + ,.body= "" + } + +#define NON_ASCII_IN_STATUS_LINE 10 +/* Should handle non-ASCII in status line */ +, {.name= "non-ASCII in status line" + ,.type= HTTP_RESPONSE + ,.raw= "HTTP/1.1 500 Oriëntatieprobleem\r\n" + "Date: Fri, 5 Nov 2010 23:07:12 GMT+2\r\n" + "Content-Length: 0\r\n" + "Connection: close\r\n" + "\r\n" + ,.should_keep_alive= FALSE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.status_code= 500 + ,.num_headers= 3 + ,.headers= + { { "Date", "Fri, 5 Nov 2010 23:07:12 GMT+2" } + , { "Content-Length", "0" } + , { "Connection", "close" } + } + ,.body= "" + } + +#define HTTP_VERSION_0_9 11 +/* Should handle HTTP/0.9 */ +, {.name= "http version 0.9" + ,.type= HTTP_RESPONSE + ,.raw= "HTTP/0.9 200 OK\r\n" + "\r\n" + ,.should_keep_alive= FALSE + ,.message_complete_on_eof= TRUE + ,.http_major= 0 + ,.http_minor= 9 + ,.status_code= 200 + ,.num_headers= 0 + ,.headers= + {} + ,.body= "" + } + +#define NO_CONTENT_LENGTH_NO_TRANSFER_ENCODING_RESPONSE 12 +/* The client should wait for the server's EOF. That is, when neither + * content-length nor transfer-encoding is specified, the end of body + * is specified by the EOF. + */ +, {.name= "neither content-length nor transfer-encoding response" + ,.type= HTTP_RESPONSE + ,.raw= "HTTP/1.1 200 OK\r\n" + "Content-Type: text/plain\r\n" + "\r\n" + "hello world" + ,.should_keep_alive= FALSE + ,.message_complete_on_eof= TRUE + ,.http_major= 1 + ,.http_minor= 1 + ,.status_code= 200 + ,.num_headers= 1 + ,.headers= + { { "Content-Type", "text/plain" } + } + ,.body= "hello world" + } + +#define NO_BODY_HTTP10_KA_200 13 +, {.name= "HTTP/1.0 with keep-alive and EOF-terminated 200 status" + ,.type= HTTP_RESPONSE + ,.raw= "HTTP/1.0 200 OK\r\n" + "Connection: keep-alive\r\n" + "\r\n" + ,.should_keep_alive= FALSE + ,.message_complete_on_eof= TRUE + ,.http_major= 1 + ,.http_minor= 0 + ,.status_code= 200 + ,.num_headers= 1 + ,.headers= + { { "Connection", "keep-alive" } + } + ,.body_size= 0 + ,.body= "" + } + +#define NO_BODY_HTTP10_KA_204 14 +, {.name= "HTTP/1.0 with keep-alive and a 204 status" + ,.type= HTTP_RESPONSE + ,.raw= "HTTP/1.0 204 No content\r\n" + "Connection: keep-alive\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 0 + ,.status_code= 204 + ,.num_headers= 1 + ,.headers= + { { "Connection", "keep-alive" } + } + ,.body_size= 0 + ,.body= "" + } + +#define NO_BODY_HTTP11_KA_200 15 +, {.name= "HTTP/1.1 with an EOF-terminated 200 status" + ,.type= HTTP_RESPONSE + ,.raw= "HTTP/1.1 200 OK\r\n" + "\r\n" + ,.should_keep_alive= FALSE + ,.message_complete_on_eof= TRUE + ,.http_major= 1 + ,.http_minor= 1 + ,.status_code= 200 + ,.num_headers= 0 + ,.headers={} + ,.body_size= 0 + ,.body= "" + } + +#define NO_BODY_HTTP11_KA_204 16 +, {.name= "HTTP/1.1 with a 204 status" + ,.type= HTTP_RESPONSE + ,.raw= "HTTP/1.1 204 No content\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.status_code= 204 + ,.num_headers= 0 + ,.headers={} + ,.body_size= 0 + ,.body= "" + } + +#define NO_BODY_HTTP11_NOKA_204 17 +, {.name= "HTTP/1.1 with a 204 status and keep-alive disabled" + ,.type= HTTP_RESPONSE + ,.raw= "HTTP/1.1 204 No content\r\n" + "Connection: close\r\n" + "\r\n" + ,.should_keep_alive= FALSE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.status_code= 204 + ,.num_headers= 1 + ,.headers= + { { "Connection", "close" } + } + ,.body_size= 0 + ,.body= "" + } + +#define NO_BODY_HTTP11_KA_CHUNKED_200 18 +, {.name= "HTTP/1.1 with chunked endocing and a 200 response" + ,.type= HTTP_RESPONSE + ,.raw= "HTTP/1.1 200 OK\r\n" + "Transfer-Encoding: chunked\r\n" + "\r\n" + "0\r\n" + "\r\n" + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.status_code= 200 + ,.num_headers= 1 + ,.headers= + { { "Transfer-Encoding", "chunked" } + } + ,.body_size= 0 + ,.body= "" + } + +#if !HTTP_PARSER_STRICT +#define SPACE_IN_FIELD_RES 19 +/* Should handle spaces in header fields */ +, {.name= "field space" + ,.type= HTTP_RESPONSE + ,.raw= "HTTP/1.1 200 OK\r\n" + "Server: Microsoft-IIS/6.0\r\n" + "X-Powered-By: ASP.NET\r\n" + "en-US Content-Type: text/xml\r\n" /* this is the problem */ + "Content-Type: text/xml\r\n" + "Content-Length: 16\r\n" + "Date: Fri, 23 Jul 2010 18:45:38 GMT\r\n" + "Connection: keep-alive\r\n" + "\r\n" + "hello" /* fake body */ + ,.should_keep_alive= TRUE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 1 + ,.status_code= 200 + ,.num_headers= 7 + ,.headers= + { { "Server", "Microsoft-IIS/6.0" } + , { "X-Powered-By", "ASP.NET" } + , { "en-US Content-Type", "text/xml" } + , { "Content-Type", "text/xml" } + , { "Content-Length", "16" } + , { "Date", "Fri, 23 Jul 2010 18:45:38 GMT" } + , { "Connection", "keep-alive" } + } + ,.body= "hello" + } +#endif /* !HTTP_PARSER_STRICT */ + +, {.name= NULL } /* sentinel */ +}; + +/* strnlen() is a POSIX.2008 addition. Can't rely on it being available so + * define it ourselves. + */ +size_t +strnlen(const char *s, size_t maxlen) +{ + const char *p; + + p = memchr(s, '\0', maxlen); + if (p == NULL) + return maxlen; + + return p - s; +} + +size_t +strlncat(char *dst, size_t len, const char *src, size_t n) +{ + size_t slen; + size_t dlen; + size_t rlen; + size_t ncpy; + + slen = strnlen(src, n); + dlen = strnlen(dst, len); + + if (dlen < len) { + rlen = len - dlen; + ncpy = slen < rlen ? slen : (rlen - 1); + memcpy(dst + dlen, src, ncpy); + dst[dlen + ncpy] = '\0'; + } + + assert(len > slen + dlen); + return slen + dlen; +} + +size_t +strlcat(char *dst, const char *src, size_t len) +{ + return strlncat(dst, len, src, (size_t) -1); +} + +size_t +strlncpy(char *dst, size_t len, const char *src, size_t n) +{ + size_t slen; + size_t ncpy; + + slen = strnlen(src, n); + + if (len > 0) { + ncpy = slen < len ? slen : (len - 1); + memcpy(dst, src, ncpy); + dst[ncpy] = '\0'; + } + + assert(len > slen); + return slen; +} + +size_t +strlcpy(char *dst, const char *src, size_t len) +{ + return strlncpy(dst, len, src, (size_t) -1); +} + +int +request_url_cb (http_parser *p, const char *buf, size_t len) +{ + assert(p == parser); + strlncat(messages[num_messages].request_url, + sizeof(messages[num_messages].request_url), + buf, + len); + return 0; +} + +int +status_complete_cb (http_parser *p) { + assert(p == parser); + p->data++; + return 0; +} + +int +header_field_cb (http_parser *p, const char *buf, size_t len) +{ + assert(p == parser); + struct message *m = &messages[num_messages]; + + if (m->last_header_element != FIELD) + m->num_headers++; + + strlncat(m->headers[m->num_headers-1][0], + sizeof(m->headers[m->num_headers-1][0]), + buf, + len); + + m->last_header_element = FIELD; + + return 0; +} + +int +header_value_cb (http_parser *p, const char *buf, size_t len) +{ + assert(p == parser); + struct message *m = &messages[num_messages]; + + strlncat(m->headers[m->num_headers-1][1], + sizeof(m->headers[m->num_headers-1][1]), + buf, + len); + + m->last_header_element = VALUE; + + return 0; +} + +void +check_body_is_final (const http_parser *p) +{ + if (messages[num_messages].body_is_final) { + fprintf(stderr, "\n\n *** Error http_body_is_final() should return 1 " + "on last on_body callback call " + "but it doesn't! ***\n\n"); + assert(0); + abort(); + } + messages[num_messages].body_is_final = http_body_is_final(p); +} + +int +body_cb (http_parser *p, const char *buf, size_t len) +{ + assert(p == parser); + strlncat(messages[num_messages].body, + sizeof(messages[num_messages].body), + buf, + len); + messages[num_messages].body_size += len; + check_body_is_final(p); + // printf("body_cb: '%s'\n", requests[num_messages].body); + return 0; +} + +int +count_body_cb (http_parser *p, const char *buf, size_t len) +{ + assert(p == parser); + assert(buf); + messages[num_messages].body_size += len; + check_body_is_final(p); + return 0; +} + +int +message_begin_cb (http_parser *p) +{ + assert(p == parser); + messages[num_messages].message_begin_cb_called = TRUE; + return 0; +} + +int +headers_complete_cb (http_parser *p) +{ + assert(p == parser); + messages[num_messages].method = parser->method; + messages[num_messages].status_code = parser->status_code; + messages[num_messages].http_major = parser->http_major; + messages[num_messages].http_minor = parser->http_minor; + messages[num_messages].headers_complete_cb_called = TRUE; + messages[num_messages].should_keep_alive = http_should_keep_alive(parser); + return 0; +} + +int +message_complete_cb (http_parser *p) +{ + assert(p == parser); + if (messages[num_messages].should_keep_alive != http_should_keep_alive(parser)) + { + fprintf(stderr, "\n\n *** Error http_should_keep_alive() should have same " + "value in both on_message_complete and on_headers_complete " + "but it doesn't! ***\n\n"); + assert(0); + abort(); + } + + if (messages[num_messages].body_size && + http_body_is_final(p) && + !messages[num_messages].body_is_final) + { + fprintf(stderr, "\n\n *** Error http_body_is_final() should return 1 " + "on last on_body callback call " + "but it doesn't! ***\n\n"); + assert(0); + abort(); + } + + messages[num_messages].message_complete_cb_called = TRUE; + + messages[num_messages].message_complete_on_eof = currently_parsing_eof; + + num_messages++; + return 0; +} + +/* These dontcall_* callbacks exist so that we can verify that when we're + * paused, no additional callbacks are invoked */ +int +dontcall_message_begin_cb (http_parser *p) +{ + if (p) { } // gcc + fprintf(stderr, "\n\n*** on_message_begin() called on paused parser ***\n\n"); + abort(); +} + +int +dontcall_header_field_cb (http_parser *p, const char *buf, size_t len) +{ + if (p || buf || len) { } // gcc + fprintf(stderr, "\n\n*** on_header_field() called on paused parser ***\n\n"); + abort(); +} + +int +dontcall_header_value_cb (http_parser *p, const char *buf, size_t len) +{ + if (p || buf || len) { } // gcc + fprintf(stderr, "\n\n*** on_header_value() called on paused parser ***\n\n"); + abort(); +} + +int +dontcall_request_url_cb (http_parser *p, const char *buf, size_t len) +{ + if (p || buf || len) { } // gcc + fprintf(stderr, "\n\n*** on_request_url() called on paused parser ***\n\n"); + abort(); +} + +int +dontcall_body_cb (http_parser *p, const char *buf, size_t len) +{ + if (p || buf || len) { } // gcc + fprintf(stderr, "\n\n*** on_body_cb() called on paused parser ***\n\n"); + abort(); +} + +int +dontcall_headers_complete_cb (http_parser *p) +{ + if (p) { } // gcc + fprintf(stderr, "\n\n*** on_headers_complete() called on paused " + "parser ***\n\n"); + abort(); +} + +int +dontcall_message_complete_cb (http_parser *p) +{ + if (p) { } // gcc + fprintf(stderr, "\n\n*** on_message_complete() called on paused " + "parser ***\n\n"); + abort(); +} + +static http_parser_settings settings_dontcall = + {.on_message_begin = dontcall_message_begin_cb + ,.on_header_field = dontcall_header_field_cb + ,.on_header_value = dontcall_header_value_cb + ,.on_url = dontcall_request_url_cb + ,.on_body = dontcall_body_cb + ,.on_headers_complete = dontcall_headers_complete_cb + ,.on_message_complete = dontcall_message_complete_cb + }; + +/* These pause_* callbacks always pause the parser and just invoke the regular + * callback that tracks content. Before returning, we overwrite the parser + * settings to point to the _dontcall variety so that we can verify that + * the pause actually did, you know, pause. */ +int +pause_message_begin_cb (http_parser *p) +{ + http_parser_pause(p, 1); + *current_pause_parser = settings_dontcall; + return message_begin_cb(p); +} + +int +pause_header_field_cb (http_parser *p, const char *buf, size_t len) +{ + http_parser_pause(p, 1); + *current_pause_parser = settings_dontcall; + return header_field_cb(p, buf, len); +} + +int +pause_header_value_cb (http_parser *p, const char *buf, size_t len) +{ + http_parser_pause(p, 1); + *current_pause_parser = settings_dontcall; + return header_value_cb(p, buf, len); +} + +int +pause_request_url_cb (http_parser *p, const char *buf, size_t len) +{ + http_parser_pause(p, 1); + *current_pause_parser = settings_dontcall; + return request_url_cb(p, buf, len); +} + +int +pause_body_cb (http_parser *p, const char *buf, size_t len) +{ + http_parser_pause(p, 1); + *current_pause_parser = settings_dontcall; + return body_cb(p, buf, len); +} + +int +pause_headers_complete_cb (http_parser *p) +{ + http_parser_pause(p, 1); + *current_pause_parser = settings_dontcall; + return headers_complete_cb(p); +} + +int +pause_message_complete_cb (http_parser *p) +{ + http_parser_pause(p, 1); + *current_pause_parser = settings_dontcall; + return message_complete_cb(p); +} + +static http_parser_settings settings_pause = + {.on_message_begin = pause_message_begin_cb + ,.on_header_field = pause_header_field_cb + ,.on_header_value = pause_header_value_cb + ,.on_url = pause_request_url_cb + ,.on_body = pause_body_cb + ,.on_headers_complete = pause_headers_complete_cb + ,.on_message_complete = pause_message_complete_cb + }; + +static http_parser_settings settings = + {.on_message_begin = message_begin_cb + ,.on_header_field = header_field_cb + ,.on_header_value = header_value_cb + ,.on_url = request_url_cb + ,.on_body = body_cb + ,.on_headers_complete = headers_complete_cb + ,.on_message_complete = message_complete_cb + }; + +static http_parser_settings settings_count_body = + {.on_message_begin = message_begin_cb + ,.on_header_field = header_field_cb + ,.on_header_value = header_value_cb + ,.on_url = request_url_cb + ,.on_body = count_body_cb + ,.on_headers_complete = headers_complete_cb + ,.on_message_complete = message_complete_cb + }; + +static http_parser_settings settings_null = + {.on_message_begin = 0 + ,.on_header_field = 0 + ,.on_header_value = 0 + ,.on_url = 0 + ,.on_body = 0 + ,.on_headers_complete = 0 + ,.on_message_complete = 0 + }; + +void +parser_init (enum http_parser_type type) +{ + num_messages = 0; + + assert(parser == NULL); + + parser = malloc(sizeof(http_parser)); + + http_parser_init(parser, type); + + memset(&messages, 0, sizeof messages); + +} + +void +parser_free () +{ + assert(parser); + free(parser); + parser = NULL; +} + +size_t parse (const char *buf, size_t len) +{ + size_t nparsed; + currently_parsing_eof = (len == 0); + nparsed = http_parser_execute(parser, &settings, buf, len); + return nparsed; +} + +size_t parse_count_body (const char *buf, size_t len) +{ + size_t nparsed; + currently_parsing_eof = (len == 0); + nparsed = http_parser_execute(parser, &settings_count_body, buf, len); + return nparsed; +} + +size_t parse_pause (const char *buf, size_t len) +{ + size_t nparsed; + http_parser_settings s = settings_pause; + + currently_parsing_eof = (len == 0); + current_pause_parser = &s; + nparsed = http_parser_execute(parser, current_pause_parser, buf, len); + return nparsed; +} + +static inline int +check_str_eq (const struct message *m, + const char *prop, + const char *expected, + const char *found) { + if ((expected == NULL) != (found == NULL)) { + printf("\n*** Error: %s in '%s' ***\n\n", prop, m->name); + printf("expected %s\n", (expected == NULL) ? "NULL" : expected); + printf(" found %s\n", (found == NULL) ? "NULL" : found); + return 0; + } + if (expected != NULL && 0 != strcmp(expected, found)) { + printf("\n*** Error: %s in '%s' ***\n\n", prop, m->name); + printf("expected '%s'\n", expected); + printf(" found '%s'\n", found); + return 0; + } + return 1; +} + +static inline int +check_num_eq (const struct message *m, + const char *prop, + int expected, + int found) { + if (expected != found) { + printf("\n*** Error: %s in '%s' ***\n\n", prop, m->name); + printf("expected %d\n", expected); + printf(" found %d\n", found); + return 0; + } + return 1; +} + +#define MESSAGE_CHECK_STR_EQ(expected, found, prop) \ + if (!check_str_eq(expected, #prop, expected->prop, found->prop)) return 0 + +#define MESSAGE_CHECK_NUM_EQ(expected, found, prop) \ + if (!check_num_eq(expected, #prop, expected->prop, found->prop)) return 0 + +#define MESSAGE_CHECK_URL_EQ(u, expected, found, prop, fn) \ +do { \ + char ubuf[256]; \ + \ + if ((u)->field_set & (1 << (fn))) { \ + memcpy(ubuf, (found)->request_url + (u)->field_data[(fn)].off, \ + (u)->field_data[(fn)].len); \ + ubuf[(u)->field_data[(fn)].len] = '\0'; \ + } else { \ + ubuf[0] = '\0'; \ + } \ + \ + check_str_eq(expected, #prop, expected->prop, ubuf); \ +} while(0) + +int +message_eq (int index, const struct message *expected) +{ + int i; + struct message *m = &messages[index]; + + MESSAGE_CHECK_NUM_EQ(expected, m, http_major); + MESSAGE_CHECK_NUM_EQ(expected, m, http_minor); + + if (expected->type == HTTP_REQUEST) { + MESSAGE_CHECK_NUM_EQ(expected, m, method); + } else { + MESSAGE_CHECK_NUM_EQ(expected, m, status_code); + } + + MESSAGE_CHECK_NUM_EQ(expected, m, should_keep_alive); + MESSAGE_CHECK_NUM_EQ(expected, m, message_complete_on_eof); + + assert(m->message_begin_cb_called); + assert(m->headers_complete_cb_called); + assert(m->message_complete_cb_called); + + + MESSAGE_CHECK_STR_EQ(expected, m, request_url); + + /* Check URL components; we can't do this w/ CONNECT since it doesn't + * send us a well-formed URL. + */ + if (*m->request_url && m->method != HTTP_CONNECT) { + struct http_parser_url u; + + if (http_parser_parse_url(m->request_url, strlen(m->request_url), 0, &u)) { + fprintf(stderr, "\n\n*** failed to parse URL %s ***\n\n", + m->request_url); + abort(); + } + + if (expected->host) { + MESSAGE_CHECK_URL_EQ(&u, expected, m, host, UF_HOST); + } + + if (expected->userinfo) { + MESSAGE_CHECK_URL_EQ(&u, expected, m, userinfo, UF_USERINFO); + } + + m->port = (u.field_set & (1 << UF_PORT)) ? + u.port : 0; + + MESSAGE_CHECK_URL_EQ(&u, expected, m, query_string, UF_QUERY); + MESSAGE_CHECK_URL_EQ(&u, expected, m, fragment, UF_FRAGMENT); + MESSAGE_CHECK_URL_EQ(&u, expected, m, request_path, UF_PATH); + MESSAGE_CHECK_NUM_EQ(expected, m, port); + } + + if (expected->body_size) { + MESSAGE_CHECK_NUM_EQ(expected, m, body_size); + } else { + MESSAGE_CHECK_STR_EQ(expected, m, body); + } + + MESSAGE_CHECK_NUM_EQ(expected, m, num_headers); + + int r; + for (i = 0; i < m->num_headers; i++) { + r = check_str_eq(expected, "header field", expected->headers[i][0], m->headers[i][0]); + if (!r) return 0; + r = check_str_eq(expected, "header value", expected->headers[i][1], m->headers[i][1]); + if (!r) return 0; + } + + MESSAGE_CHECK_STR_EQ(expected, m, upgrade); + + return 1; +} + +/* Given a sequence of varargs messages, return the number of them that the + * parser should successfully parse, taking into account that upgraded + * messages prevent all subsequent messages from being parsed. + */ +size_t +count_parsed_messages(const size_t nmsgs, ...) { + size_t i; + va_list ap; + + va_start(ap, nmsgs); + + for (i = 0; i < nmsgs; i++) { + struct message *m = va_arg(ap, struct message *); + + if (m->upgrade) { + va_end(ap); + return i + 1; + } + } + + va_end(ap); + return nmsgs; +} + +/* Given a sequence of bytes and the number of these that we were able to + * parse, verify that upgrade bodies are correct. + */ +void +upgrade_message_fix(char *body, const size_t nread, const size_t nmsgs, ...) { + va_list ap; + size_t i; + size_t off = 0; + + va_start(ap, nmsgs); + + for (i = 0; i < nmsgs; i++) { + struct message *m = va_arg(ap, struct message *); + + off += strlen(m->raw); + + if (m->upgrade) { + off -= strlen(m->upgrade); + + /* Check the portion of the response after its specified upgrade */ + if (!check_str_eq(m, "upgrade", body + off, body + nread)) { + abort(); + } + + /* Fix up the response so that message_eq() will verify the beginning + * of the upgrade */ + *(body + nread + strlen(m->upgrade)) = '\0'; + messages[num_messages -1 ].upgrade = body + nread; + + va_end(ap); + return; + } + } + + va_end(ap); + printf("\n\n*** Error: expected a message with upgrade ***\n"); + + abort(); +} + +static void +print_error (const char *raw, size_t error_location) +{ + fprintf(stderr, "\n*** %s ***\n\n", + http_errno_description(HTTP_PARSER_ERRNO(parser))); + + int this_line = 0, char_len = 0; + size_t i, j, len = strlen(raw), error_location_line = 0; + for (i = 0; i < len; i++) { + if (i == error_location) this_line = 1; + switch (raw[i]) { + case '\r': + char_len = 2; + fprintf(stderr, "\\r"); + break; + + case '\n': + char_len = 2; + fprintf(stderr, "\\n\n"); + + if (this_line) goto print; + + error_location_line = 0; + continue; + + default: + char_len = 1; + fputc(raw[i], stderr); + break; + } + if (!this_line) error_location_line += char_len; + } + + fprintf(stderr, "[eof]\n"); + + print: + for (j = 0; j < error_location_line; j++) { + fputc(' ', stderr); + } + fprintf(stderr, "^\n\nerror location: %u\n", (unsigned int)error_location); +} + +void +test_preserve_data (void) +{ + char my_data[] = "application-specific data"; + http_parser parser; + parser.data = my_data; + http_parser_init(&parser, HTTP_REQUEST); + if (parser.data != my_data) { + printf("\n*** parser.data not preserved accross http_parser_init ***\n\n"); + abort(); + } +} + +struct url_test { + const char *name; + const char *url; + int is_connect; + struct http_parser_url u; + int rv; +}; + +const struct url_test url_tests[] = +{ {.name="proxy request" + ,.url="http://hostname/" + ,.is_connect=0 + ,.u= + {.field_set=(1 << UF_SCHEMA) | (1 << UF_HOST) | (1 << UF_PATH) + ,.port=0 + ,.field_data= + {{ 0, 4 } /* UF_SCHEMA */ + ,{ 7, 8 } /* UF_HOST */ + ,{ 0, 0 } /* UF_PORT */ + ,{ 15, 1 } /* UF_PATH */ + ,{ 0, 0 } /* UF_QUERY */ + ,{ 0, 0 } /* UF_FRAGMENT */ + ,{ 0, 0 } /* UF_USERINFO */ + } + } + ,.rv=0 + } + +, {.name="proxy request with port" + ,.url="http://hostname:444/" + ,.is_connect=0 + ,.u= + {.field_set=(1 << UF_SCHEMA) | (1 << UF_HOST) | (1 << UF_PORT) | (1 << UF_PATH) + ,.port=444 + ,.field_data= + {{ 0, 4 } /* UF_SCHEMA */ + ,{ 7, 8 } /* UF_HOST */ + ,{ 16, 3 } /* UF_PORT */ + ,{ 19, 1 } /* UF_PATH */ + ,{ 0, 0 } /* UF_QUERY */ + ,{ 0, 0 } /* UF_FRAGMENT */ + ,{ 0, 0 } /* UF_USERINFO */ + } + } + ,.rv=0 + } + +, {.name="CONNECT request" + ,.url="hostname:443" + ,.is_connect=1 + ,.u= + {.field_set=(1 << UF_HOST) | (1 << UF_PORT) + ,.port=443 + ,.field_data= + {{ 0, 0 } /* UF_SCHEMA */ + ,{ 0, 8 } /* UF_HOST */ + ,{ 9, 3 } /* UF_PORT */ + ,{ 0, 0 } /* UF_PATH */ + ,{ 0, 0 } /* UF_QUERY */ + ,{ 0, 0 } /* UF_FRAGMENT */ + ,{ 0, 0 } /* UF_USERINFO */ + } + } + ,.rv=0 + } + +, {.name="CONNECT request but not connect" + ,.url="hostname:443" + ,.is_connect=0 + ,.rv=1 + } + +, {.name="proxy ipv6 request" + ,.url="http://[1:2::3:4]/" + ,.is_connect=0 + ,.u= + {.field_set=(1 << UF_SCHEMA) | (1 << UF_HOST) | (1 << UF_PATH) + ,.port=0 + ,.field_data= + {{ 0, 4 } /* UF_SCHEMA */ + ,{ 8, 8 } /* UF_HOST */ + ,{ 0, 0 } /* UF_PORT */ + ,{ 17, 1 } /* UF_PATH */ + ,{ 0, 0 } /* UF_QUERY */ + ,{ 0, 0 } /* UF_FRAGMENT */ + ,{ 0, 0 } /* UF_USERINFO */ + } + } + ,.rv=0 + } + +, {.name="proxy ipv6 request with port" + ,.url="http://[1:2::3:4]:67/" + ,.is_connect=0 + ,.u= + {.field_set=(1 << UF_SCHEMA) | (1 << UF_HOST) | (1 << UF_PORT) | (1 << UF_PATH) + ,.port=67 + ,.field_data= + {{ 0, 4 } /* UF_SCHEMA */ + ,{ 8, 8 } /* UF_HOST */ + ,{ 18, 2 } /* UF_PORT */ + ,{ 20, 1 } /* UF_PATH */ + ,{ 0, 0 } /* UF_QUERY */ + ,{ 0, 0 } /* UF_FRAGMENT */ + ,{ 0, 0 } /* UF_USERINFO */ + } + } + ,.rv=0 + } + +, {.name="CONNECT ipv6 address" + ,.url="[1:2::3:4]:443" + ,.is_connect=1 + ,.u= + {.field_set=(1 << UF_HOST) | (1 << UF_PORT) + ,.port=443 + ,.field_data= + {{ 0, 0 } /* UF_SCHEMA */ + ,{ 1, 8 } /* UF_HOST */ + ,{ 11, 3 } /* UF_PORT */ + ,{ 0, 0 } /* UF_PATH */ + ,{ 0, 0 } /* UF_QUERY */ + ,{ 0, 0 } /* UF_FRAGMENT */ + ,{ 0, 0 } /* UF_USERINFO */ + } + } + ,.rv=0 + } + +, {.name="ipv4 in ipv6 address" + ,.url="http://[2001:0000:0000:0000:0000:0000:1.9.1.1]/" + ,.is_connect=0 + ,.u= + {.field_set=(1 << UF_SCHEMA) | (1 << UF_HOST) | (1 << UF_PATH) + ,.port=0 + ,.field_data= + {{ 0, 4 } /* UF_SCHEMA */ + ,{ 8, 37 } /* UF_HOST */ + ,{ 0, 0 } /* UF_PORT */ + ,{ 46, 1 } /* UF_PATH */ + ,{ 0, 0 } /* UF_QUERY */ + ,{ 0, 0 } /* UF_FRAGMENT */ + ,{ 0, 0 } /* UF_USERINFO */ + } + } + ,.rv=0 + } + +, {.name="extra ? in query string" + ,.url="http://a.tbcdn.cn/p/fp/2010c/??fp-header-min.css,fp-base-min.css," + "fp-channel-min.css,fp-product-min.css,fp-mall-min.css,fp-category-min.css," + "fp-sub-min.css,fp-gdp4p-min.css,fp-css3-min.css,fp-misc-min.css?t=20101022.css" + ,.is_connect=0 + ,.u= + {.field_set=(1<field_set, u->port); + for (i = 0; i < UF_MAX; i++) { + if ((u->field_set & (1 << i)) == 0) { + printf("\tfield_data[%u]: unset\n", i); + continue; + } + + printf("\tfield_data[%u]: off: %u len: %u part: \"%.*s\n\"", + i, + u->field_data[i].off, + u->field_data[i].len, + u->field_data[i].len, + url + u->field_data[i].off); + } +} + +void +test_parse_url (void) +{ + struct http_parser_url u; + const struct url_test *test; + unsigned int i; + int rv; + + for (i = 0; i < (sizeof(url_tests) / sizeof(url_tests[0])); i++) { + test = &url_tests[i]; + memset(&u, 0, sizeof(u)); + + rv = http_parser_parse_url(test->url, + strlen(test->url), + test->is_connect, + &u); + + if (test->rv == 0) { + if (rv != 0) { + printf("\n*** http_parser_parse_url(\"%s\") \"%s\" test failed, " + "unexpected rv %d ***\n\n", test->url, test->name, rv); + abort(); + } + + if (memcmp(&u, &test->u, sizeof(u)) != 0) { + printf("\n*** http_parser_parse_url(\"%s\") \"%s\" failed ***\n", + test->url, test->name); + + printf("target http_parser_url:\n"); + dump_url(test->url, &test->u); + printf("result http_parser_url:\n"); + dump_url(test->url, &u); + + abort(); + } + } else { + /* test->rv != 0 */ + if (rv == 0) { + printf("\n*** http_parser_parse_url(\"%s\") \"%s\" test failed, " + "unexpected rv %d ***\n\n", test->url, test->name, rv); + abort(); + } + } + } +} + +void +test_method_str (void) +{ + assert(0 == strcmp("GET", http_method_str(HTTP_GET))); + assert(0 == strcmp("", http_method_str(1337))); +} + +void +test_message (const struct message *message) +{ + size_t raw_len = strlen(message->raw); + size_t msg1len; + for (msg1len = 0; msg1len < raw_len; msg1len++) { + parser_init(message->type); + + size_t read; + const char *msg1 = message->raw; + const char *msg2 = msg1 + msg1len; + size_t msg2len = raw_len - msg1len; + + if (msg1len) { + read = parse(msg1, msg1len); + + if (message->upgrade && parser->upgrade) { + messages[num_messages - 1].upgrade = msg1 + read; + goto test; + } + + if (read != msg1len) { + print_error(msg1, read); + abort(); + } + } + + + read = parse(msg2, msg2len); + + if (message->upgrade && parser->upgrade) { + messages[num_messages - 1].upgrade = msg2 + read; + goto test; + } + + if (read != msg2len) { + print_error(msg2, read); + abort(); + } + + read = parse(NULL, 0); + + if (read != 0) { + print_error(message->raw, read); + abort(); + } + + test: + + if (num_messages != 1) { + printf("\n*** num_messages != 1 after testing '%s' ***\n\n", message->name); + abort(); + } + + if(!message_eq(0, message)) abort(); + + parser_free(); + } +} + +void +test_message_count_body (const struct message *message) +{ + parser_init(message->type); + + size_t read; + size_t l = strlen(message->raw); + size_t i, toread; + size_t chunk = 4024; + + for (i = 0; i < l; i+= chunk) { + toread = MIN(l-i, chunk); + read = parse_count_body(message->raw + i, toread); + if (read != toread) { + print_error(message->raw, read); + abort(); + } + } + + + read = parse_count_body(NULL, 0); + if (read != 0) { + print_error(message->raw, read); + abort(); + } + + if (num_messages != 1) { + printf("\n*** num_messages != 1 after testing '%s' ***\n\n", message->name); + abort(); + } + + if(!message_eq(0, message)) abort(); + + parser_free(); +} + +void +test_simple (const char *buf, enum http_errno err_expected) +{ + parser_init(HTTP_REQUEST); + + size_t parsed; + int pass; + enum http_errno err; + + parsed = parse(buf, strlen(buf)); + pass = (parsed == strlen(buf)); + err = HTTP_PARSER_ERRNO(parser); + parsed = parse(NULL, 0); + pass &= (parsed == 0); + + parser_free(); + + /* In strict mode, allow us to pass with an unexpected HPE_STRICT as + * long as the caller isn't expecting success. + */ +#if HTTP_PARSER_STRICT + if (err_expected != err && err_expected != HPE_OK && err != HPE_STRICT) { +#else + if (err_expected != err) { +#endif + fprintf(stderr, "\n*** test_simple expected %s, but saw %s ***\n\n%s\n", + http_errno_name(err_expected), http_errno_name(err), buf); + abort(); + } +} + +void +test_header_overflow_error (int req) +{ + http_parser parser; + http_parser_init(&parser, req ? HTTP_REQUEST : HTTP_RESPONSE); + size_t parsed; + const char *buf; + buf = req ? "GET / HTTP/1.1\r\n" : "HTTP/1.0 200 OK\r\n"; + parsed = http_parser_execute(&parser, &settings_null, buf, strlen(buf)); + assert(parsed == strlen(buf)); + + buf = "header-key: header-value\r\n"; + size_t buflen = strlen(buf); + + int i; + for (i = 0; i < 10000; i++) { + parsed = http_parser_execute(&parser, &settings_null, buf, buflen); + if (parsed != buflen) { + //fprintf(stderr, "error found on iter %d\n", i); + assert(HTTP_PARSER_ERRNO(&parser) == HPE_HEADER_OVERFLOW); + return; + } + } + + fprintf(stderr, "\n*** Error expected but none in header overflow test ***\n"); + abort(); +} + +static void +test_content_length_overflow (const char *buf, size_t buflen, int expect_ok) +{ + http_parser parser; + http_parser_init(&parser, HTTP_RESPONSE); + http_parser_execute(&parser, &settings_null, buf, buflen); + + if (expect_ok) + assert(HTTP_PARSER_ERRNO(&parser) == HPE_OK); + else + assert(HTTP_PARSER_ERRNO(&parser) == HPE_INVALID_CONTENT_LENGTH); +} + +void +test_header_content_length_overflow_error (void) +{ +#define X(size) \ + "HTTP/1.1 200 OK\r\n" \ + "Content-Length: " #size "\r\n" \ + "\r\n" + const char a[] = X(18446744073709551614); /* 2^64-2 */ + const char b[] = X(18446744073709551615); /* 2^64-1 */ + const char c[] = X(18446744073709551616); /* 2^64 */ +#undef X + test_content_length_overflow(a, sizeof(a) - 1, 1); /* expect ok */ + test_content_length_overflow(b, sizeof(b) - 1, 0); /* expect failure */ + test_content_length_overflow(c, sizeof(c) - 1, 0); /* expect failure */ +} + +void +test_chunk_content_length_overflow_error (void) +{ +#define X(size) \ + "HTTP/1.1 200 OK\r\n" \ + "Transfer-Encoding: chunked\r\n" \ + "\r\n" \ + #size "\r\n" \ + "..." + const char a[] = X(FFFFFFFFFFFFFFFE); /* 2^64-2 */ + const char b[] = X(FFFFFFFFFFFFFFFF); /* 2^64-1 */ + const char c[] = X(10000000000000000); /* 2^64 */ +#undef X + test_content_length_overflow(a, sizeof(a) - 1, 1); /* expect ok */ + test_content_length_overflow(b, sizeof(b) - 1, 0); /* expect failure */ + test_content_length_overflow(c, sizeof(c) - 1, 0); /* expect failure */ +} + +void +test_no_overflow_long_body (int req, size_t length) +{ + http_parser parser; + http_parser_init(&parser, req ? HTTP_REQUEST : HTTP_RESPONSE); + size_t parsed; + size_t i; + char buf1[3000]; + size_t buf1len = sprintf(buf1, "%s\r\nConnection: Keep-Alive\r\nContent-Length: %lu\r\n\r\n", + req ? "POST / HTTP/1.0" : "HTTP/1.0 200 OK", (unsigned long)length); + parsed = http_parser_execute(&parser, &settings_null, buf1, buf1len); + if (parsed != buf1len) + goto err; + + for (i = 0; i < length; i++) { + char foo = 'a'; + parsed = http_parser_execute(&parser, &settings_null, &foo, 1); + if (parsed != 1) + goto err; + } + + parsed = http_parser_execute(&parser, &settings_null, buf1, buf1len); + if (parsed != buf1len) goto err; + return; + + err: + fprintf(stderr, + "\n*** error in test_no_overflow_long_body %s of length %lu ***\n", + req ? "REQUEST" : "RESPONSE", + (unsigned long)length); + abort(); +} + +void +test_multiple3 (const struct message *r1, const struct message *r2, const struct message *r3) +{ + int message_count = count_parsed_messages(3, r1, r2, r3); + + char total[ strlen(r1->raw) + + strlen(r2->raw) + + strlen(r3->raw) + + 1 + ]; + total[0] = '\0'; + + strcat(total, r1->raw); + strcat(total, r2->raw); + strcat(total, r3->raw); + + parser_init(r1->type); + + size_t read; + + read = parse(total, strlen(total)); + + if (parser->upgrade) { + upgrade_message_fix(total, read, 3, r1, r2, r3); + goto test; + } + + if (read != strlen(total)) { + print_error(total, read); + abort(); + } + + read = parse(NULL, 0); + + if (read != 0) { + print_error(total, read); + abort(); + } + +test: + + if (message_count != num_messages) { + fprintf(stderr, "\n\n*** Parser didn't see 3 messages only %d *** \n", num_messages); + abort(); + } + + if (!message_eq(0, r1)) abort(); + if (message_count > 1 && !message_eq(1, r2)) abort(); + if (message_count > 2 && !message_eq(2, r3)) abort(); + + parser_free(); +} + +/* SCAN through every possible breaking to make sure the + * parser can handle getting the content in any chunks that + * might come from the socket + */ +void +test_scan (const struct message *r1, const struct message *r2, const struct message *r3) +{ + char total[80*1024] = "\0"; + char buf1[80*1024] = "\0"; + char buf2[80*1024] = "\0"; + char buf3[80*1024] = "\0"; + + strcat(total, r1->raw); + strcat(total, r2->raw); + strcat(total, r3->raw); + + size_t read; + + int total_len = strlen(total); + + int total_ops = 2 * (total_len - 1) * (total_len - 2) / 2; + int ops = 0 ; + + size_t buf1_len, buf2_len, buf3_len; + int message_count = count_parsed_messages(3, r1, r2, r3); + + int i,j,type_both; + for (type_both = 0; type_both < 2; type_both ++ ) { + for (j = 2; j < total_len; j ++ ) { + for (i = 1; i < j; i ++ ) { + + if (ops % 1000 == 0) { + printf("\b\b\b\b%3.0f%%", 100 * (float)ops /(float)total_ops); + fflush(stdout); + } + ops += 1; + + parser_init(type_both ? HTTP_BOTH : r1->type); + + buf1_len = i; + strlncpy(buf1, sizeof(buf1), total, buf1_len); + buf1[buf1_len] = 0; + + buf2_len = j - i; + strlncpy(buf2, sizeof(buf1), total+i, buf2_len); + buf2[buf2_len] = 0; + + buf3_len = total_len - j; + strlncpy(buf3, sizeof(buf1), total+j, buf3_len); + buf3[buf3_len] = 0; + + read = parse(buf1, buf1_len); + + if (parser->upgrade) goto test; + + if (read != buf1_len) { + print_error(buf1, read); + goto error; + } + + read += parse(buf2, buf2_len); + + if (parser->upgrade) goto test; + + if (read != buf1_len + buf2_len) { + print_error(buf2, read); + goto error; + } + + read += parse(buf3, buf3_len); + + if (parser->upgrade) goto test; + + if (read != buf1_len + buf2_len + buf3_len) { + print_error(buf3, read); + goto error; + } + + parse(NULL, 0); + +test: + if (parser->upgrade) { + upgrade_message_fix(total, read, 3, r1, r2, r3); + } + + if (message_count != num_messages) { + fprintf(stderr, "\n\nParser didn't see %d messages only %d\n", + message_count, num_messages); + goto error; + } + + if (!message_eq(0, r1)) { + fprintf(stderr, "\n\nError matching messages[0] in test_scan.\n"); + goto error; + } + + if (message_count > 1 && !message_eq(1, r2)) { + fprintf(stderr, "\n\nError matching messages[1] in test_scan.\n"); + goto error; + } + + if (message_count > 2 && !message_eq(2, r3)) { + fprintf(stderr, "\n\nError matching messages[2] in test_scan.\n"); + goto error; + } + + parser_free(); + } + } + } + puts("\b\b\b\b100%"); + return; + + error: + fprintf(stderr, "i=%d j=%d\n", i, j); + fprintf(stderr, "buf1 (%u) %s\n\n", (unsigned int)buf1_len, buf1); + fprintf(stderr, "buf2 (%u) %s\n\n", (unsigned int)buf2_len , buf2); + fprintf(stderr, "buf3 (%u) %s\n", (unsigned int)buf3_len, buf3); + abort(); +} + +// user required to free the result +// string terminated by \0 +char * +create_large_chunked_message (int body_size_in_kb, const char* headers) +{ + int i; + size_t wrote = 0; + size_t headers_len = strlen(headers); + size_t bufsize = headers_len + (5+1024+2)*body_size_in_kb + 6; + char * buf = malloc(bufsize); + + memcpy(buf, headers, headers_len); + wrote += headers_len; + + for (i = 0; i < body_size_in_kb; i++) { + // write 1kb chunk into the body. + memcpy(buf + wrote, "400\r\n", 5); + wrote += 5; + memset(buf + wrote, 'C', 1024); + wrote += 1024; + strcpy(buf + wrote, "\r\n"); + wrote += 2; + } + + memcpy(buf + wrote, "0\r\n\r\n", 6); + wrote += 6; + assert(wrote == bufsize); + + return buf; +} + +void +test_status_complete (void) +{ + parser_init(HTTP_RESPONSE); + parser->data = 0; + http_parser_settings settings = settings_null; + settings.on_status_complete = status_complete_cb; + + char *response = "don't mind me, just a simple response"; + http_parser_execute(parser, &settings, response, strlen(response)); + assert(parser->data == (void*)0); // the status_complete callback was never called + assert(parser->http_errno == HPE_INVALID_CONSTANT); // the errno for an invalid status line +} + +/* Verify that we can pause parsing at any of the bytes in the + * message and still get the result that we're expecting. */ +void +test_message_pause (const struct message *msg) +{ + char *buf = (char*) msg->raw; + size_t buflen = strlen(msg->raw); + size_t nread; + + parser_init(msg->type); + + do { + nread = parse_pause(buf, buflen); + + // We can only set the upgrade buffer once we've gotten our message + // completion callback. + if (messages[0].message_complete_cb_called && + msg->upgrade && + parser->upgrade) { + messages[0].upgrade = buf + nread; + goto test; + } + + if (nread < buflen) { + + // Not much do to if we failed a strict-mode check + if (HTTP_PARSER_ERRNO(parser) == HPE_STRICT) { + parser_free(); + return; + } + + assert (HTTP_PARSER_ERRNO(parser) == HPE_PAUSED); + } + + buf += nread; + buflen -= nread; + http_parser_pause(parser, 0); + } while (buflen > 0); + + nread = parse_pause(NULL, 0); + assert (nread == 0); + +test: + if (num_messages != 1) { + printf("\n*** num_messages != 1 after testing '%s' ***\n\n", msg->name); + abort(); + } + + if(!message_eq(0, msg)) abort(); + + parser_free(); +} + +int +main (void) +{ + parser = NULL; + int i, j, k; + int request_count; + int response_count; + + printf("sizeof(http_parser) = %u\n", (unsigned int)sizeof(http_parser)); + + for (request_count = 0; requests[request_count].name; request_count++); + for (response_count = 0; responses[response_count].name; response_count++); + + //// API + test_preserve_data(); + test_parse_url(); + test_method_str(); + + //// OVERFLOW CONDITIONS + + test_header_overflow_error(HTTP_REQUEST); + test_no_overflow_long_body(HTTP_REQUEST, 1000); + test_no_overflow_long_body(HTTP_REQUEST, 100000); + + test_header_overflow_error(HTTP_RESPONSE); + test_no_overflow_long_body(HTTP_RESPONSE, 1000); + test_no_overflow_long_body(HTTP_RESPONSE, 100000); + + test_header_content_length_overflow_error(); + test_chunk_content_length_overflow_error(); + + //// RESPONSES + + for (i = 0; i < response_count; i++) { + test_message(&responses[i]); + } + + for (i = 0; i < response_count; i++) { + test_message_pause(&responses[i]); + } + + for (i = 0; i < response_count; i++) { + if (!responses[i].should_keep_alive) continue; + for (j = 0; j < response_count; j++) { + if (!responses[j].should_keep_alive) continue; + for (k = 0; k < response_count; k++) { + test_multiple3(&responses[i], &responses[j], &responses[k]); + } + } + } + + test_message_count_body(&responses[NO_HEADERS_NO_BODY_404]); + test_message_count_body(&responses[TRAILING_SPACE_ON_CHUNKED_BODY]); + + // test very large chunked response + { + char * msg = create_large_chunked_message(31337, + "HTTP/1.0 200 OK\r\n" + "Transfer-Encoding: chunked\r\n" + "Content-Type: text/plain\r\n" + "\r\n"); + struct message large_chunked = + {.name= "large chunked" + ,.type= HTTP_RESPONSE + ,.raw= msg + ,.should_keep_alive= FALSE + ,.message_complete_on_eof= FALSE + ,.http_major= 1 + ,.http_minor= 0 + ,.status_code= 200 + ,.num_headers= 2 + ,.headers= + { { "Transfer-Encoding", "chunked" } + , { "Content-Type", "text/plain" } + } + ,.body_size= 31337*1024 + }; + test_message_count_body(&large_chunked); + free(msg); + } + + + + printf("response scan 1/2 "); + test_scan( &responses[TRAILING_SPACE_ON_CHUNKED_BODY] + , &responses[NO_BODY_HTTP10_KA_204] + , &responses[NO_REASON_PHRASE] + ); + + printf("response scan 2/2 "); + test_scan( &responses[BONJOUR_MADAME_FR] + , &responses[UNDERSTORE_HEADER_KEY] + , &responses[NO_CARRIAGE_RET] + ); + + puts("responses okay"); + + + /// REQUESTS + + test_simple("hello world", HPE_INVALID_METHOD); + test_simple("GET / HTP/1.1\r\n\r\n", HPE_INVALID_VERSION); + + + test_simple("ASDF / HTTP/1.1\r\n\r\n", HPE_INVALID_METHOD); + test_simple("PROPPATCHA / HTTP/1.1\r\n\r\n", HPE_INVALID_METHOD); + test_simple("GETA / HTTP/1.1\r\n\r\n", HPE_INVALID_METHOD); + + // Well-formed but incomplete + test_simple("GET / HTTP/1.1\r\n" + "Content-Type: text/plain\r\n" + "Content-Length: 6\r\n" + "\r\n" + "fooba", + HPE_OK); + + static const char *all_methods[] = { + "DELETE", + "GET", + "HEAD", + "POST", + "PUT", + //"CONNECT", //CONNECT can't be tested like other methods, it's a tunnel + "OPTIONS", + "TRACE", + "COPY", + "LOCK", + "MKCOL", + "MOVE", + "PROPFIND", + "PROPPATCH", + "UNLOCK", + "REPORT", + "MKACTIVITY", + "CHECKOUT", + "MERGE", + "M-SEARCH", + "NOTIFY", + "SUBSCRIBE", + "UNSUBSCRIBE", + "PATCH", + 0 }; + const char **this_method; + for (this_method = all_methods; *this_method; this_method++) { + char buf[200]; + sprintf(buf, "%s / HTTP/1.1\r\n\r\n", *this_method); + test_simple(buf, HPE_OK); + } + + static const char *bad_methods[] = { + "C******", + "M****", + 0 }; + for (this_method = bad_methods; *this_method; this_method++) { + char buf[200]; + sprintf(buf, "%s / HTTP/1.1\r\n\r\n", *this_method); + test_simple(buf, HPE_UNKNOWN); + } + + const char *dumbfuck2 = + "GET / HTTP/1.1\r\n" + "X-SSL-Bullshit: -----BEGIN CERTIFICATE-----\r\n" + "\tMIIFbTCCBFWgAwIBAgICH4cwDQYJKoZIhvcNAQEFBQAwcDELMAkGA1UEBhMCVUsx\r\n" + "\tETAPBgNVBAoTCGVTY2llbmNlMRIwEAYDVQQLEwlBdXRob3JpdHkxCzAJBgNVBAMT\r\n" + "\tAkNBMS0wKwYJKoZIhvcNAQkBFh5jYS1vcGVyYXRvckBncmlkLXN1cHBvcnQuYWMu\r\n" + "\tdWswHhcNMDYwNzI3MTQxMzI4WhcNMDcwNzI3MTQxMzI4WjBbMQswCQYDVQQGEwJV\r\n" + "\tSzERMA8GA1UEChMIZVNjaWVuY2UxEzARBgNVBAsTCk1hbmNoZXN0ZXIxCzAJBgNV\r\n" + "\tBAcTmrsogriqMWLAk1DMRcwFQYDVQQDEw5taWNoYWVsIHBhcmQYJKoZIhvcNAQEB\r\n" + "\tBQADggEPADCCAQoCggEBANPEQBgl1IaKdSS1TbhF3hEXSl72G9J+WC/1R64fAcEF\r\n" + "\tW51rEyFYiIeZGx/BVzwXbeBoNUK41OK65sxGuflMo5gLflbwJtHBRIEKAfVVp3YR\r\n" + "\tgW7cMA/s/XKgL1GEC7rQw8lIZT8RApukCGqOVHSi/F1SiFlPDxuDfmdiNzL31+sL\r\n" + "\t0iwHDdNkGjy5pyBSB8Y79dsSJtCW/iaLB0/n8Sj7HgvvZJ7x0fr+RQjYOUUfrePP\r\n" + "\tu2MSpFyf+9BbC/aXgaZuiCvSR+8Snv3xApQY+fULK/xY8h8Ua51iXoQ5jrgu2SqR\r\n" + "\twgA7BUi3G8LFzMBl8FRCDYGUDy7M6QaHXx1ZWIPWNKsCAwEAAaOCAiQwggIgMAwG\r\n" + "\tA1UdEwEB/wQCMAAwEQYJYIZIAYb4QgHTTPAQDAgWgMA4GA1UdDwEB/wQEAwID6DAs\r\n" + "\tBglghkgBhvhCAQ0EHxYdVUsgZS1TY2llbmNlIFVzZXIgQ2VydGlmaWNhdGUwHQYD\r\n" + "\tVR0OBBYEFDTt/sf9PeMaZDHkUIldrDYMNTBZMIGaBgNVHSMEgZIwgY+AFAI4qxGj\r\n" + "\tloCLDdMVKwiljjDastqooXSkcjBwMQswCQYDVQQGEwJVSzERMA8GA1UEChMIZVNj\r\n" + "\taWVuY2UxEjAQBgNVBAsTCUF1dGhvcml0eTELMAkGA1UEAxMCQ0ExLTArBgkqhkiG\r\n" + "\t9w0BCQEWHmNhLW9wZXJhdG9yQGdyaWQtc3VwcG9ydC5hYy51a4IBADApBgNVHRIE\r\n" + "\tIjAggR5jYS1vcGVyYXRvckBncmlkLXN1cHBvcnQuYWMudWswGQYDVR0gBBIwEDAO\r\n" + "\tBgwrBgEEAdkvAQEBAQYwPQYJYIZIAYb4QgEEBDAWLmh0dHA6Ly9jYS5ncmlkLXN1\r\n" + "\tcHBvcnQuYWMudmT4sopwqlBWsvcHViL2NybC9jYWNybC5jcmwwPQYJYIZIAYb4QgEDBDAWLmh0\r\n" + "\tdHA6Ly9jYS5ncmlkLXN1cHBvcnQuYWMudWsvcHViL2NybC9jYWNybC5jcmwwPwYD\r\n" + "\tVR0fBDgwNjA0oDKgMIYuaHR0cDovL2NhLmdyaWQt5hYy51ay9wdWIv\r\n" + "\tY3JsL2NhY3JsLmNybDANBgkqhkiG9w0BAQUFAAOCAQEAS/U4iiooBENGW/Hwmmd3\r\n" + "\tXCy6Zrt08YjKCzGNjorT98g8uGsqYjSxv/hmi0qlnlHs+k/3Iobc3LjS5AMYr5L8\r\n" + "\tUO7OSkgFFlLHQyC9JzPfmLCAugvzEbyv4Olnsr8hbxF1MbKZoQxUZtMVu29wjfXk\r\n" + "\thTeApBv7eaKCWpSp7MCbvgzm74izKhu3vlDk9w6qVrxePfGgpKPqfHiOoGhFnbTK\r\n" + "\twTC6o2xq5y0qZ03JonF7OJspEd3I5zKY3E+ov7/ZhW6DqT8UFvsAdjvQbXyhV8Eu\r\n" + "\tYhixw1aKEPzNjNowuIseVogKOLXxWI5vAi5HgXdS0/ES5gDGsABo4fqovUKlgop3\r\n" + "\tRA==\r\n" + "\t-----END CERTIFICATE-----\r\n" + "\r\n"; + test_simple(dumbfuck2, HPE_OK); + +#if 0 + // NOTE(Wed Nov 18 11:57:27 CET 2009) this seems okay. we just read body + // until EOF. + // + // no content-length + // error if there is a body without content length + const char *bad_get_no_headers_no_body = "GET /bad_get_no_headers_no_body/world HTTP/1.1\r\n" + "Accept: */*\r\n" + "\r\n" + "HELLO"; + test_simple(bad_get_no_headers_no_body, 0); +#endif + /* TODO sending junk and large headers gets rejected */ + + + /* check to make sure our predefined requests are okay */ + for (i = 0; requests[i].name; i++) { + test_message(&requests[i]); + } + + for (i = 0; i < request_count; i++) { + test_message_pause(&requests[i]); + } + + for (i = 0; i < request_count; i++) { + if (!requests[i].should_keep_alive) continue; + for (j = 0; j < request_count; j++) { + if (!requests[j].should_keep_alive) continue; + for (k = 0; k < request_count; k++) { + test_multiple3(&requests[i], &requests[j], &requests[k]); + } + } + } + + printf("request scan 1/4 "); + test_scan( &requests[GET_NO_HEADERS_NO_BODY] + , &requests[GET_ONE_HEADER_NO_BODY] + , &requests[GET_NO_HEADERS_NO_BODY] + ); + + printf("request scan 2/4 "); + test_scan( &requests[POST_CHUNKED_ALL_YOUR_BASE] + , &requests[POST_IDENTITY_BODY_WORLD] + , &requests[GET_FUNKY_CONTENT_LENGTH] + ); + + printf("request scan 3/4 "); + test_scan( &requests[TWO_CHUNKS_MULT_ZERO_END] + , &requests[CHUNKED_W_TRAILING_HEADERS] + , &requests[CHUNKED_W_BULLSHIT_AFTER_LENGTH] + ); + + printf("request scan 4/4 "); + test_scan( &requests[QUERY_URL_WITH_QUESTION_MARK_GET] + , &requests[PREFIX_NEWLINE_GET ] + , &requests[CONNECT_REQUEST] + ); + + test_status_complete(); + + puts("requests okay"); + + return 0; +} diff --git a/node/Http.cpp b/node/Http.cpp new file mode 100644 index 00000000..86ae2d25 --- /dev/null +++ b/node/Http.cpp @@ -0,0 +1,323 @@ +/* + * ZeroTier One - Global Peer to Peer Ethernet + * Copyright (C) 2012-2013 ZeroTier Networks LLC + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation, either version 3 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see . + * + * -- + * + * ZeroTier may be used and distributed under the terms of the GPLv3, which + * are available at: http://www.gnu.org/licenses/gpl-3.0.html + * + * If you would like to embed ZeroTier into a commercial application or + * redistribute it in a modified binary form, please contact ZeroTier Networks + * LLC. Start here: http://www.zerotier.com/ + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#ifndef _WIN32 +#include +#endif + +#include "Http.hpp" +#include "Utils.hpp" +#include "InetAddress.hpp" + +static http_parser_settings _http_parser_settings; + +namespace ZeroTier { + +static bool _sendAll(int fd,const void *buf,unsigned int len) +{ + for(;;) { + int n = (int)::send(fd,buf,len,0); + if ((n < 0)&&(errno == EINTR)) + continue; + return (n == (int)len); + } +} + +const std::map Http::EMPTY_HEADERS; + +Http::Request::Request( + Http::Method m, + const std::string &url, + const std::map &rh, + const std::string &rb, + bool (*handler)(Request *,void *,const std::string &,int,const std::map &,const std::string &), + void *arg) : + _url(url), + _requestHeaders(rh), + _handler(handler), + _arg(arg), + _method(m), + _fd(0) +{ + _http_parser_settings.on_message_begin = &Http::Request::_http_on_message_begin; + _http_parser_settings.on_url = &Http::Request::_http_on_url; + _http_parser_settings.on_status_complete = &Http::Request::_http_on_status_complete; + _http_parser_settings.on_header_field = &Http::Request::_http_on_header_field; + _http_parser_settings.on_header_value = &Http::Request::_http_on_header_value; + _http_parser_settings.on_headers_complete = &Http::Request::_http_on_headers_complete; + _http_parser_settings.on_body = &Http::Request::_http_on_body; + _http_parser_settings.on_message_complete = &Http::Request::_http_on_message_complete; + + start(); +} + +Http::Request::~Request() +{ + if (_fd > 0) + ::close(_fd); + join(); +} + +void Http::Request::main() + throw() +{ + char buf[131072]; + + try { + http_parser_init(&_parser,HTTP_RESPONSE); + _parser.data = this; + + http_parser_url urlParsed; + if (http_parser_parse_url(_url.c_str(),_url.length(),0,&urlParsed)) { + suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"URL parse error"); + return; + } + if (!(urlParsed.field_set & (1 << UF_SCHEMA))) { + suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"URL specifies no schema"); + return; + } + + std::string schema(_url.substr(urlParsed.field_data[UF_SCHEMA].off,urlParsed.field_data[UF_SCHEMA].len)); + + if (schema == "file") { + const std::string filePath(_url.substr(urlParsed.field_data[UF_PATH].off,urlParsed.field_data[UF_PATH].len)); + + uint64_t lm = Utils::getLastModified(filePath.c_str()); + if (lm) { + const std::map::const_iterator ifModSince(_requestHeaders.find("If-Modified-Since")); + if ((ifModSince != _requestHeaders.end())&&(ifModSince->second.length())) { + uint64_t t64 = Utils::fromRfc1123(ifModSince->second); + if ((t64)&&(lm > t64)) { + suicidalThread = !_handler(this,_arg,_url,304,_responseHeaders,""); + return; + } + } + + if (Utils::readFile(filePath.c_str(),_responseBody)) { + _responseHeaders["Last-Modified"] = Utils::toRfc1123(lm); + suicidalThread = !_handler(this,_arg,_url,200,_responseHeaders,_responseBody); + return; + } + } + + suicidalThread = !_handler(this,_arg,_url,404,_responseHeaders,"file not found or not readable"); + return; + } else if (schema == "http") { + if (!(urlParsed.field_set & (1 << UF_HOST))) { + suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"URL contains no host"); + return; + } + std::string host(_url.substr(urlParsed.field_data[UF_HOST].off,urlParsed.field_data[UF_HOST].len)); + + std::list v4,v6; + { + struct addrinfo *res = (struct addrinfo *)0; + if (!getaddrinfo(host.c_str(),(const char *)0,(const struct addrinfo *)0,&res)) { + struct addrinfo *p = res; + do { + if (p->ai_family == AF_INET) + v4.push_back(InetAddress(p->ai_addr)); + else if (p->ai_family == AF_INET6) + v6.push_back(InetAddress(p->ai_addr)); + } while ((p = p->ai_next)); + freeaddrinfo(res); + } + } + + std::list *addrList; + if (v4.empty()&&v6.empty()) { + suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"could not find address for host in URL"); + return; + } else if (v4.empty()) { + addrList = &v6; + } else { + addrList = &v4; + } + InetAddress *addr; + { + addrList->sort(); + addrList->unique(); + unsigned int i = 0,k = 0; + k = rand() % addrList->size(); + std::list::iterator a(addrList->begin()); + while (i++ != k) ++a; + addr = &(*a); + } + + int remotePort = ((urlParsed.field_set & (1 << UF_PORT))&&(urlParsed.port)) ? (int)urlParsed.port : (int)80; + if ((remotePort <= 0)||(remotePort > 0xffff)) { + suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"URL port out of range"); + return; + } + addr->setPort(remotePort); + + _fd = socket(addr->isV6() ? AF_INET6 : AF_INET,SOCK_STREAM,0); + if (_fd <= 0) { + suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"could not open socket"); + return; + } + + for(;;) { + if (connect(_fd,addr->saddr(),addr->saddrLen())) { + if (errno == EINTR) + continue; + ::close(_fd); _fd = 0; + suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"connection failed to remote host"); + return; + } else break; + } + + const char *mstr = "GET"; + switch(_method) { + case HTTP_METHOD_HEAD: mstr = "HEAD"; break; + default: break; + } + int mlen = (int)snprintf(buf,sizeof(buf),"%s %s HTTP/1.1\r\nAccept-Encoding: \r\nHost: %s\r\n",mstr,_url.substr(urlParsed.field_data[UF_PATH].off,urlParsed.field_data[UF_PATH].len).c_str(),host.c_str()); + if (mlen >= (int)sizeof(buf)) { + ::close(_fd); _fd = 0; + suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"URL too long"); + return; + } + if (!_sendAll(_fd,buf,mlen)) { + ::close(_fd); _fd = 0; + suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"write error"); + return; + } + + for(std::map::const_iterator rh(_requestHeaders.begin());rh!=_requestHeaders.end();++rh) { + mlen = (int)snprintf(buf,sizeof(buf),"%s: %s\r\n",rh->first.c_str(),rh->second.c_str()); + if (mlen >= (int)sizeof(buf)) { + ::close(_fd); _fd = 0; + suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"header too long"); + return; + } + if (!_sendAll(_fd,buf,mlen)) { + ::close(_fd); _fd = 0; + suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"write error"); + return; + } + } + + if (!_sendAll(_fd,"\r\n",2)) { + ::close(_fd); _fd = 0; + suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"write error"); + return; + } + + _responseStatusCode = 0; + _messageComplete = false; + for(;;) { + mlen = (int)::recv(_fd,buf,sizeof(buf),0); + if (mlen < 0) { + if (errno != EINTR) + break; + else continue; + } + if (((int)http_parser_execute(&_parser,&_http_parser_settings,buf,mlen) != mlen)||(_parser.upgrade)) { + ::close(_fd); _fd = 0; + suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"invalid HTTP response from server"); + return; + } + if (_messageComplete) { + ::close(_fd); _fd = 0; + suicidalThread = !_handler(this,_arg,_url,_responseStatusCode,_responseHeaders,_responseBody); + return; + } + } + + ::close(_fd); _fd = 0; + suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"empty HTTP response from server"); + return; + } else { + suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"only 'file' and 'http' methods are supported"); + return; + } + } catch ( ... ) { + suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"unexpected exception retrieving URL"); + return; + } +} + +int Http::Request::_http_on_message_begin(http_parser *parser) +{ + return 0; +} +int Http::Request::_http_on_url(http_parser *parser,const char *data,size_t length) +{ + return 0; +} +int Http::Request::_http_on_status_complete(http_parser *parser) +{ + Http::Request *r = (Http::Request *)parser->data; + r->_responseStatusCode = parser->status_code; + return 0; +} +int Http::Request::_http_on_header_field(http_parser *parser,const char *data,size_t length) +{ + Http::Request *r = (Http::Request *)parser->data; + if ((r->_currentHeaderField.length())&&(r->_responseHeaders.find(r->_currentHeaderField) != r->_responseHeaders.end())) + r->_currentHeaderField.assign(""); + r->_currentHeaderField.append(data,length); + return 0; +} +int Http::Request::_http_on_header_value(http_parser *parser,const char *data,size_t length) +{ + Http::Request *r = (Http::Request *)parser->data; + if (r->_currentHeaderField.length()) + r->_responseHeaders[r->_currentHeaderField].append(data,length); + return 0; +} +int Http::Request::_http_on_headers_complete(http_parser *parser) +{ + Http::Request *r = (Http::Request *)parser->data; + return ((r->_method == Http::HTTP_METHOD_HEAD) ? 1 : 0); +} +int Http::Request::_http_on_body(http_parser *parser,const char *data,size_t length) +{ + Http::Request *r = (Http::Request *)parser->data; + r->_responseBody.append(data,length); + return 0; +} +int Http::Request::_http_on_message_complete(http_parser *parser) +{ + Http::Request *r = (Http::Request *)parser->data; + r->_messageComplete = true; + return 0; +} + +} // namespace ZeroTier diff --git a/node/Http.hpp b/node/Http.hpp new file mode 100644 index 00000000..a099b15d --- /dev/null +++ b/node/Http.hpp @@ -0,0 +1,129 @@ +/* + * ZeroTier One - Global Peer to Peer Ethernet + * Copyright (C) 2012-2013 ZeroTier Networks LLC + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation, either version 3 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see . + * + * -- + * + * ZeroTier may be used and distributed under the terms of the GPLv3, which + * are available at: http://www.gnu.org/licenses/gpl-3.0.html + * + * If you would like to embed ZeroTier into a commercial application or + * redistribute it in a modified binary form, please contact ZeroTier Networks + * LLC. Start here: http://www.zerotier.com/ + */ + +#ifndef _ZT_HTTP_HPP +#define _ZT_HTTP_HPP + +#include +#include +#include +#include "Thread.hpp" + +#include "../ext/http-parser/http_parser.h" + +namespace ZeroTier { + +class Http +{ +public: + /** + * HTTP request methods + */ + enum Method + { + HTTP_METHOD_GET, + HTTP_METHOD_HEAD + }; + + /** + * An empty headers map for convenience + */ + static const std::map EMPTY_HEADERS; + + /** + * HTTP request + */ + class Request : protected Thread + { + public: + /** + * Create and issue an HTTP request + * + * The supplied handler is called when the request is + * complete or if an error occurs. A code of zero indicates + * that the server could not be reached, and a description + * of the error will be in 'body'. If the handler returns + * false the Request object deletes itself. Otherwise the + * object must be deleted by other code. + * + * @param m Request method + * @param url Destination URL + * @param rh Request headers + * @param rb Request body or empty string for none (currently unused) + * @param handler Request handler function + * @param arg First argument to request handler + */ + Request( + Http::Method m, + const std::string &url, + const std::map &rh, + const std::string &rb, + bool (*handler)(Request *,void *,const std::string &,int,const std::map &,const std::string &), + void *arg); + + /** + * Destruction cancels any in-progress request + */ + virtual ~Request(); + + protected: + virtual void main() + throw(); + + private: + // HTTP parser handlers + static int _http_on_message_begin(http_parser *parser); + static int _http_on_url(http_parser *parser,const char *data,size_t length); + static int _http_on_status_complete(http_parser *parser); + static int _http_on_header_field(http_parser *parser,const char *data,size_t length); + static int _http_on_header_value(http_parser *parser,const char *data,size_t length); + static int _http_on_headers_complete(http_parser *parser); + static int _http_on_body(http_parser *parser,const char *data,size_t length); + static int _http_on_message_complete(http_parser *parser); + + http_parser _parser; + std::string _url; + + std::map _requestHeaders; + std::map _responseHeaders; + + std::string _currentHeaderField; + std::string _responseBody; + + bool (*_handler)(Request *,void *,const std::string &,int,const std::map &,const std::string &); + void *_arg; + + Http::Method _method; + int _responseStatusCode; + bool _messageComplete; + volatile int _fd; + }; +}; + +} // namespace ZeroTier + +#endif diff --git a/objects.mk b/objects.mk index 6ad6f5ec..49312994 100644 --- a/objects.mk +++ b/objects.mk @@ -1,4 +1,5 @@ OBJS=\ + ext/http-parser/http_parser.o \ ext/kissdb/kissdb.o \ ext/lz4/lz4hc.o \ ext/lz4/lz4.o \ @@ -8,6 +9,7 @@ OBJS=\ node/EthernetTap.o \ node/Filter.o \ node/HMAC.o \ + node/Http.o \ node/Identity.o \ node/InetAddress.o \ node/Logger.o \ -- cgit v1.2.3 From 9df88a3933ae75ff2ec9bfa951a71fcc0164df74 Mon Sep 17 00:00:00 2001 From: Adam Ierymenko Date: Wed, 31 Jul 2013 10:05:00 -0400 Subject: Change mind again... dump Http. Launcher will do this and will use libcurl. Also fix some format string errors. --- AUTHORS.txt | 3 - ext/http-parser/AUTHORS | 41 - ext/http-parser/CONTRIBUTIONS | 4 - ext/http-parser/LICENSE-MIT | 23 - ext/http-parser/README.md | 180 -- ext/http-parser/contrib/parsertrace.c | 156 -- ext/http-parser/contrib/url_parser.c | 44 - ext/http-parser/http_parser.c | 2175 --------------------- ext/http-parser/http_parser.gyp | 111 -- ext/http-parser/http_parser.h | 304 --- ext/http-parser/test.c | 3425 --------------------------------- node/Address.hpp | 2 +- node/Http.cpp | 323 ---- node/Http.hpp | 129 -- node/Network.hpp | 12 +- objects.mk | 2 - 16 files changed, 7 insertions(+), 6927 deletions(-) delete mode 100644 ext/http-parser/AUTHORS delete mode 100644 ext/http-parser/CONTRIBUTIONS delete mode 100644 ext/http-parser/LICENSE-MIT delete mode 100644 ext/http-parser/README.md delete mode 100644 ext/http-parser/contrib/parsertrace.c delete mode 100644 ext/http-parser/contrib/url_parser.c delete mode 100644 ext/http-parser/http_parser.c delete mode 100644 ext/http-parser/http_parser.gyp delete mode 100644 ext/http-parser/http_parser.h delete mode 100644 ext/http-parser/test.c delete mode 100644 node/Http.cpp delete mode 100644 node/Http.hpp (limited to 'ext') diff --git a/AUTHORS.txt b/AUTHORS.txt index 41c65af9..f304f6af 100644 --- a/AUTHORS.txt +++ b/AUTHORS.txt @@ -16,9 +16,6 @@ ZeroTier One includes or links with the following third party software: * LZ4 compression algorithm by Yann Collet (BSD license) http://code.google.com/p/lz4/ - * http-parser, a simple C http parser library (MIT license) - https://github.com/joyent/http-parser - * OpenSSL libcrypto (BSD-style OpenSSL license) http://www.openssl.org/ diff --git a/ext/http-parser/AUTHORS b/ext/http-parser/AUTHORS deleted file mode 100644 index 92ee45ca..00000000 --- a/ext/http-parser/AUTHORS +++ /dev/null @@ -1,41 +0,0 @@ -# Authors ordered by first contribution. -Ryan Dahl -Jeremy Hinegardner -Sergey Shepelev -Joe Damato -tomika -Phoenix Sol -Cliff Frey -Ewen Cheslack-Postava -Santiago Gala -Tim Becker -Jeff Terrace -Ben Noordhuis -Nathan Rajlich -Mark Nottingham -Aman Gupta -Tim Becker -Sean Cunningham -Peter Griess -Salman Haq -Cliff Frey -Jon Kolb -Fouad Mardini -Paul Querna -Felix Geisendörfer -koichik -Andre Caron -Ivo Raisr -James McLaughlin -David Gwynne -Thomas LE ROUX -Randy Rizun -Andre Louis Caron -Simon Zimmermann -Erik Dubbelboer -Martell Malone -Bertrand Paquet -BogDan Vatra -Peter Faiman -Corey Richardson -Tóth Tamás diff --git a/ext/http-parser/CONTRIBUTIONS b/ext/http-parser/CONTRIBUTIONS deleted file mode 100644 index 11ba31e4..00000000 --- a/ext/http-parser/CONTRIBUTIONS +++ /dev/null @@ -1,4 +0,0 @@ -Contributors must agree to the Contributor License Agreement before patches -can be accepted. - -http://spreadsheets2.google.com/viewform?hl=en&formkey=dDJXOGUwbzlYaWM4cHN1MERwQS1CSnc6MQ diff --git a/ext/http-parser/LICENSE-MIT b/ext/http-parser/LICENSE-MIT deleted file mode 100644 index 58010b38..00000000 --- a/ext/http-parser/LICENSE-MIT +++ /dev/null @@ -1,23 +0,0 @@ -http_parser.c is based on src/http/ngx_http_parse.c from NGINX copyright -Igor Sysoev. - -Additional changes are licensed under the same terms as NGINX and -copyright Joyent, Inc. and other Node contributors. All rights reserved. - -Permission is hereby granted, free of charge, to any person obtaining a copy -of this software and associated documentation files (the "Software"), to -deal in the Software without restriction, including without limitation the -rights to use, copy, modify, merge, publish, distribute, sublicense, and/or -sell copies of the Software, and to permit persons to whom the Software is -furnished to do so, subject to the following conditions: - -The above copyright notice and this permission notice shall be included in -all copies or substantial portions of the Software. - -THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR -IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, -FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE -AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER -LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING -FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS -IN THE SOFTWARE. diff --git a/ext/http-parser/README.md b/ext/http-parser/README.md deleted file mode 100644 index b63418af..00000000 --- a/ext/http-parser/README.md +++ /dev/null @@ -1,180 +0,0 @@ -HTTP Parser -=========== - -[![Build Status](https://travis-ci.org/joyent/http-parser.png?branch=master)](https://travis-ci.org/joyent/http-parser) - -This is a parser for HTTP messages written in C. It parses both requests and -responses. The parser is designed to be used in performance HTTP -applications. It does not make any syscalls nor allocations, it does not -buffer data, it can be interrupted at anytime. Depending on your -architecture, it only requires about 40 bytes of data per message -stream (in a web server that is per connection). - -Features: - - * No dependencies - * Handles persistent streams (keep-alive). - * Decodes chunked encoding. - * Upgrade support - * Defends against buffer overflow attacks. - -The parser extracts the following information from HTTP messages: - - * Header fields and values - * Content-Length - * Request method - * Response status code - * Transfer-Encoding - * HTTP version - * Request URL - * Message body - - -Usage ------ - -One `http_parser` object is used per TCP connection. Initialize the struct -using `http_parser_init()` and set the callbacks. That might look something -like this for a request parser: - - http_parser_settings settings; - settings.on_url = my_url_callback; - settings.on_header_field = my_header_field_callback; - /* ... */ - - http_parser *parser = malloc(sizeof(http_parser)); - http_parser_init(parser, HTTP_REQUEST); - parser->data = my_socket; - -When data is received on the socket execute the parser and check for errors. - - size_t len = 80*1024, nparsed; - char buf[len]; - ssize_t recved; - - recved = recv(fd, buf, len, 0); - - if (recved < 0) { - /* Handle error. */ - } - - /* Start up / continue the parser. - * Note we pass recved==0 to signal that EOF has been recieved. - */ - nparsed = http_parser_execute(parser, &settings, buf, recved); - - if (parser->upgrade) { - /* handle new protocol */ - } else if (nparsed != recved) { - /* Handle error. Usually just close the connection. */ - } - -HTTP needs to know where the end of the stream is. For example, sometimes -servers send responses without Content-Length and expect the client to -consume input (for the body) until EOF. To tell http_parser about EOF, give -`0` as the forth parameter to `http_parser_execute()`. Callbacks and errors -can still be encountered during an EOF, so one must still be prepared -to receive them. - -Scalar valued message information such as `status_code`, `method`, and the -HTTP version are stored in the parser structure. This data is only -temporally stored in `http_parser` and gets reset on each new message. If -this information is needed later, copy it out of the structure during the -`headers_complete` callback. - -The parser decodes the transfer-encoding for both requests and responses -transparently. That is, a chunked encoding is decoded before being sent to -the on_body callback. - - -The Special Problem of Upgrade ------------------------------- - -HTTP supports upgrading the connection to a different protocol. An -increasingly common example of this is the Web Socket protocol which sends -a request like - - GET /demo HTTP/1.1 - Upgrade: WebSocket - Connection: Upgrade - Host: example.com - Origin: http://example.com - WebSocket-Protocol: sample - -followed by non-HTTP data. - -(See http://tools.ietf.org/html/draft-hixie-thewebsocketprotocol-75 for more -information the Web Socket protocol.) - -To support this, the parser will treat this as a normal HTTP message without a -body. Issuing both on_headers_complete and on_message_complete callbacks. However -http_parser_execute() will stop parsing at the end of the headers and return. - -The user is expected to check if `parser->upgrade` has been set to 1 after -`http_parser_execute()` returns. Non-HTTP data begins at the buffer supplied -offset by the return value of `http_parser_execute()`. - - -Callbacks ---------- - -During the `http_parser_execute()` call, the callbacks set in -`http_parser_settings` will be executed. The parser maintains state and -never looks behind, so buffering the data is not necessary. If you need to -save certain data for later usage, you can do that from the callbacks. - -There are two types of callbacks: - -* notification `typedef int (*http_cb) (http_parser*);` - Callbacks: on_message_begin, on_headers_complete, on_message_complete. -* data `typedef int (*http_data_cb) (http_parser*, const char *at, size_t length);` - Callbacks: (requests only) on_uri, - (common) on_header_field, on_header_value, on_body; - -Callbacks must return 0 on success. Returning a non-zero value indicates -error to the parser, making it exit immediately. - -In case you parse HTTP message in chunks (i.e. `read()` request line -from socket, parse, read half headers, parse, etc) your data callbacks -may be called more than once. Http-parser guarantees that data pointer is only -valid for the lifetime of callback. You can also `read()` into a heap allocated -buffer to avoid copying memory around if this fits your application. - -Reading headers may be a tricky task if you read/parse headers partially. -Basically, you need to remember whether last header callback was field or value -and apply following logic: - - (on_header_field and on_header_value shortened to on_h_*) - ------------------------ ------------ -------------------------------------------- - | State (prev. callback) | Callback | Description/action | - ------------------------ ------------ -------------------------------------------- - | nothing (first call) | on_h_field | Allocate new buffer and copy callback data | - | | | into it | - ------------------------ ------------ -------------------------------------------- - | value | on_h_field | New header started. | - | | | Copy current name,value buffers to headers | - | | | list and allocate new buffer for new name | - ------------------------ ------------ -------------------------------------------- - | field | on_h_field | Previous name continues. Reallocate name | - | | | buffer and append callback data to it | - ------------------------ ------------ -------------------------------------------- - | field | on_h_value | Value for current header started. Allocate | - | | | new buffer and copy callback data to it | - ------------------------ ------------ -------------------------------------------- - | value | on_h_value | Value continues. Reallocate value buffer | - | | | and append callback data to it | - ------------------------ ------------ -------------------------------------------- - - -Parsing URLs ------------- - -A simplistic zero-copy URL parser is provided as `http_parser_parse_url()`. -Users of this library may wish to use it to parse URLs constructed from -consecutive `on_url` callbacks. - -See examples of reading in headers: - -* [partial example](http://gist.github.com/155877) in C -* [from http-parser tests](http://github.com/joyent/http-parser/blob/37a0ff8/test.c#L403) in C -* [from Node library](http://github.com/joyent/node/blob/842eaf4/src/http.js#L284) in Javascript diff --git a/ext/http-parser/contrib/parsertrace.c b/ext/http-parser/contrib/parsertrace.c deleted file mode 100644 index c9bc71ec..00000000 --- a/ext/http-parser/contrib/parsertrace.c +++ /dev/null @@ -1,156 +0,0 @@ -/* Based on src/http/ngx_http_parse.c from NGINX copyright Igor Sysoev - * - * Additional changes are licensed under the same terms as NGINX and - * copyright Joyent, Inc. and other Node contributors. All rights reserved. - * - * Permission is hereby granted, free of charge, to any person obtaining a copy - * of this software and associated documentation files (the "Software"), to - * deal in the Software without restriction, including without limitation the - * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or - * sell copies of the Software, and to permit persons to whom the Software is - * furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE - * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER - * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING - * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS - * IN THE SOFTWARE. - */ - -/* Dump what the parser finds to stdout as it happen */ - -#include "http_parser.h" -#include -#include -#include - -int on_message_begin(http_parser* _) { - (void)_; - printf("\n***MESSAGE BEGIN***\n\n"); - return 0; -} - -int on_headers_complete(http_parser* _) { - (void)_; - printf("\n***HEADERS COMPLETE***\n\n"); - return 0; -} - -int on_message_complete(http_parser* _) { - (void)_; - printf("\n***MESSAGE COMPLETE***\n\n"); - return 0; -} - -int on_url(http_parser* _, const char* at, size_t length) { - (void)_; - printf("Url: %.*s\n", (int)length, at); - return 0; -} - -int on_header_field(http_parser* _, const char* at, size_t length) { - (void)_; - printf("Header field: %.*s\n", (int)length, at); - return 0; -} - -int on_header_value(http_parser* _, const char* at, size_t length) { - (void)_; - printf("Header value: %.*s\n", (int)length, at); - return 0; -} - -int on_body(http_parser* _, const char* at, size_t length) { - (void)_; - printf("Body: %.*s\n", (int)length, at); - return 0; -} - -void usage(const char* name) { - fprintf(stderr, - "Usage: %s $type $filename\n" - " type: -x, where x is one of {r,b,q}\n" - " parses file as a Response, reQuest, or Both\n", - name); - exit(EXIT_FAILURE); -} - -int main(int argc, char* argv[]) { - enum http_parser_type file_type; - - if (argc != 3) { - usage(argv[0]); - } - - char* type = argv[1]; - if (type[0] != '-') { - usage(argv[0]); - } - - switch (type[1]) { - /* in the case of "-", type[1] will be NUL */ - case 'r': - file_type = HTTP_RESPONSE; - break; - case 'q': - file_type = HTTP_REQUEST; - break; - case 'b': - file_type = HTTP_BOTH; - break; - default: - usage(argv[0]); - } - - char* filename = argv[2]; - FILE* file = fopen(filename, "r"); - if (file == NULL) { - perror("fopen"); - return EXIT_FAILURE; - } - - fseek(file, 0, SEEK_END); - long file_length = ftell(file); - if (file_length == -1) { - perror("ftell"); - return EXIT_FAILURE; - } - fseek(file, 0, SEEK_SET); - - char* data = malloc(file_length); - if (fread(data, 1, file_length, file) != (size_t)file_length) { - fprintf(stderr, "couldn't read entire file\n"); - free(data); - return EXIT_FAILURE; - } - - http_parser_settings settings; - memset(&settings, 0, sizeof(settings)); - settings.on_message_begin = on_message_begin; - settings.on_url = on_url; - settings.on_header_field = on_header_field; - settings.on_header_value = on_header_value; - settings.on_headers_complete = on_headers_complete; - settings.on_body = on_body; - settings.on_message_complete = on_message_complete; - - http_parser parser; - http_parser_init(&parser, file_type); - size_t nparsed = http_parser_execute(&parser, &settings, data, file_length); - free(data); - - if (nparsed != (size_t)file_length) { - fprintf(stderr, - "Error: %s (%s)\n", - http_errno_description(HTTP_PARSER_ERRNO(&parser)), - http_errno_name(HTTP_PARSER_ERRNO(&parser))); - return EXIT_FAILURE; - } - - return EXIT_SUCCESS; -} diff --git a/ext/http-parser/contrib/url_parser.c b/ext/http-parser/contrib/url_parser.c deleted file mode 100644 index b1f9c979..00000000 --- a/ext/http-parser/contrib/url_parser.c +++ /dev/null @@ -1,44 +0,0 @@ -#include "http_parser.h" -#include -#include - -void -dump_url (const char *url, const struct http_parser_url *u) -{ - unsigned int i; - - printf("\tfield_set: 0x%x, port: %u\n", u->field_set, u->port); - for (i = 0; i < UF_MAX; i++) { - if ((u->field_set & (1 << i)) == 0) { - printf("\tfield_data[%u]: unset\n", i); - continue; - } - - printf("\tfield_data[%u]: off: %u len: %u part: \"%.*s\n", - i, - u->field_data[i].off, - u->field_data[i].len, - u->field_data[i].len, - url + u->field_data[i].off); - } -} - -int main(int argc, char ** argv) { - if (argc != 3) { - printf("Syntax : %s connect|get url\n", argv[0]); - return 1; - } - struct http_parser_url u; - int len = strlen(argv[2]); - int connect = strcmp("connect", argv[1]) == 0 ? 1 : 0; - printf("Parsing %s, connect %d\n", argv[2], connect); - - int result = http_parser_parse_url(argv[2], len, connect, &u); - if (result != 0) { - printf("Parse error : %d\n", result); - return result; - } - printf("Parse ok, result : \n"); - dump_url(argv[2], &u); - return 0; -} \ No newline at end of file diff --git a/ext/http-parser/http_parser.c b/ext/http-parser/http_parser.c deleted file mode 100644 index ed3a9232..00000000 --- a/ext/http-parser/http_parser.c +++ /dev/null @@ -1,2175 +0,0 @@ -/* Based on src/http/ngx_http_parse.c from NGINX copyright Igor Sysoev - * - * Additional changes are licensed under the same terms as NGINX and - * copyright Joyent, Inc. and other Node contributors. All rights reserved. - * - * Permission is hereby granted, free of charge, to any person obtaining a copy - * of this software and associated documentation files (the "Software"), to - * deal in the Software without restriction, including without limitation the - * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or - * sell copies of the Software, and to permit persons to whom the Software is - * furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE - * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER - * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING - * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS - * IN THE SOFTWARE. - */ -#include "http_parser.h" -#include -#include -#include -#include -#include -#include - -#ifndef ULLONG_MAX -# define ULLONG_MAX ((uint64_t) -1) /* 2^64-1 */ -#endif - -#ifndef MIN -# define MIN(a,b) ((a) < (b) ? (a) : (b)) -#endif - -#ifndef ARRAY_SIZE -# define ARRAY_SIZE(a) (sizeof(a) / sizeof((a)[0])) -#endif - -#ifndef BIT_AT -# define BIT_AT(a, i) \ - (!!((unsigned int) (a)[(unsigned int) (i) >> 3] & \ - (1 << ((unsigned int) (i) & 7)))) -#endif - -#ifndef ELEM_AT -# define ELEM_AT(a, i, v) ((unsigned int) (i) < ARRAY_SIZE(a) ? (a)[(i)] : (v)) -#endif - -#define SET_ERRNO(e) \ -do { \ - parser->http_errno = (e); \ -} while(0) - - -/* Run the notify callback FOR, returning ER if it fails */ -#define CALLBACK_NOTIFY_(FOR, ER) \ -do { \ - assert(HTTP_PARSER_ERRNO(parser) == HPE_OK); \ - \ - if (settings->on_##FOR) { \ - if (0 != settings->on_##FOR(parser)) { \ - SET_ERRNO(HPE_CB_##FOR); \ - } \ - \ - /* We either errored above or got paused; get out */ \ - if (HTTP_PARSER_ERRNO(parser) != HPE_OK) { \ - return (ER); \ - } \ - } \ -} while (0) - -/* Run the notify callback FOR and consume the current byte */ -#define CALLBACK_NOTIFY(FOR) CALLBACK_NOTIFY_(FOR, p - data + 1) - -/* Run the notify callback FOR and don't consume the current byte */ -#define CALLBACK_NOTIFY_NOADVANCE(FOR) CALLBACK_NOTIFY_(FOR, p - data) - -/* Run data callback FOR with LEN bytes, returning ER if it fails */ -#define CALLBACK_DATA_(FOR, LEN, ER) \ -do { \ - assert(HTTP_PARSER_ERRNO(parser) == HPE_OK); \ - \ - if (FOR##_mark) { \ - if (settings->on_##FOR) { \ - if (0 != settings->on_##FOR(parser, FOR##_mark, (LEN))) { \ - SET_ERRNO(HPE_CB_##FOR); \ - } \ - \ - /* We either errored above or got paused; get out */ \ - if (HTTP_PARSER_ERRNO(parser) != HPE_OK) { \ - return (ER); \ - } \ - } \ - FOR##_mark = NULL; \ - } \ -} while (0) - -/* Run the data callback FOR and consume the current byte */ -#define CALLBACK_DATA(FOR) \ - CALLBACK_DATA_(FOR, p - FOR##_mark, p - data + 1) - -/* Run the data callback FOR and don't consume the current byte */ -#define CALLBACK_DATA_NOADVANCE(FOR) \ - CALLBACK_DATA_(FOR, p - FOR##_mark, p - data) - -/* Set the mark FOR; non-destructive if mark is already set */ -#define MARK(FOR) \ -do { \ - if (!FOR##_mark) { \ - FOR##_mark = p; \ - } \ -} while (0) - - -#define PROXY_CONNECTION "proxy-connection" -#define CONNECTION "connection" -#define CONTENT_LENGTH "content-length" -#define TRANSFER_ENCODING "transfer-encoding" -#define UPGRADE "upgrade" -#define CHUNKED "chunked" -#define KEEP_ALIVE "keep-alive" -#define CLOSE "close" - - -static const char *method_strings[] = - { -#define XX(num, name, string) #string, - HTTP_METHOD_MAP(XX) -#undef XX - }; - - -/* Tokens as defined by rfc 2616. Also lowercases them. - * token = 1* - * separators = "(" | ")" | "<" | ">" | "@" - * | "," | ";" | ":" | "\" | <"> - * | "/" | "[" | "]" | "?" | "=" - * | "{" | "}" | SP | HT - */ -static const char tokens[256] = { -/* 0 nul 1 soh 2 stx 3 etx 4 eot 5 enq 6 ack 7 bel */ - 0, 0, 0, 0, 0, 0, 0, 0, -/* 8 bs 9 ht 10 nl 11 vt 12 np 13 cr 14 so 15 si */ - 0, 0, 0, 0, 0, 0, 0, 0, -/* 16 dle 17 dc1 18 dc2 19 dc3 20 dc4 21 nak 22 syn 23 etb */ - 0, 0, 0, 0, 0, 0, 0, 0, -/* 24 can 25 em 26 sub 27 esc 28 fs 29 gs 30 rs 31 us */ - 0, 0, 0, 0, 0, 0, 0, 0, -/* 32 sp 33 ! 34 " 35 # 36 $ 37 % 38 & 39 ' */ - 0, '!', 0, '#', '$', '%', '&', '\'', -/* 40 ( 41 ) 42 * 43 + 44 , 45 - 46 . 47 / */ - 0, 0, '*', '+', 0, '-', '.', 0, -/* 48 0 49 1 50 2 51 3 52 4 53 5 54 6 55 7 */ - '0', '1', '2', '3', '4', '5', '6', '7', -/* 56 8 57 9 58 : 59 ; 60 < 61 = 62 > 63 ? */ - '8', '9', 0, 0, 0, 0, 0, 0, -/* 64 @ 65 A 66 B 67 C 68 D 69 E 70 F 71 G */ - 0, 'a', 'b', 'c', 'd', 'e', 'f', 'g', -/* 72 H 73 I 74 J 75 K 76 L 77 M 78 N 79 O */ - 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', -/* 80 P 81 Q 82 R 83 S 84 T 85 U 86 V 87 W */ - 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', -/* 88 X 89 Y 90 Z 91 [ 92 \ 93 ] 94 ^ 95 _ */ - 'x', 'y', 'z', 0, 0, 0, '^', '_', -/* 96 ` 97 a 98 b 99 c 100 d 101 e 102 f 103 g */ - '`', 'a', 'b', 'c', 'd', 'e', 'f', 'g', -/* 104 h 105 i 106 j 107 k 108 l 109 m 110 n 111 o */ - 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', -/* 112 p 113 q 114 r 115 s 116 t 117 u 118 v 119 w */ - 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', -/* 120 x 121 y 122 z 123 { 124 | 125 } 126 ~ 127 del */ - 'x', 'y', 'z', 0, '|', 0, '~', 0 }; - - -static const int8_t unhex[256] = - {-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 - ,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 - ,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 - , 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,-1,-1,-1,-1,-1,-1 - ,-1,10,11,12,13,14,15,-1,-1,-1,-1,-1,-1,-1,-1,-1 - ,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 - ,-1,10,11,12,13,14,15,-1,-1,-1,-1,-1,-1,-1,-1,-1 - ,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 - }; - - -#if HTTP_PARSER_STRICT -# define T(v) 0 -#else -# define T(v) v -#endif - - -static const uint8_t normal_url_char[32] = { -/* 0 nul 1 soh 2 stx 3 etx 4 eot 5 enq 6 ack 7 bel */ - 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0, -/* 8 bs 9 ht 10 nl 11 vt 12 np 13 cr 14 so 15 si */ - 0 | T(2) | 0 | 0 | T(16) | 0 | 0 | 0, -/* 16 dle 17 dc1 18 dc2 19 dc3 20 dc4 21 nak 22 syn 23 etb */ - 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0, -/* 24 can 25 em 26 sub 27 esc 28 fs 29 gs 30 rs 31 us */ - 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0, -/* 32 sp 33 ! 34 " 35 # 36 $ 37 % 38 & 39 ' */ - 0 | 2 | 4 | 0 | 16 | 32 | 64 | 128, -/* 40 ( 41 ) 42 * 43 + 44 , 45 - 46 . 47 / */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, -/* 48 0 49 1 50 2 51 3 52 4 53 5 54 6 55 7 */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, -/* 56 8 57 9 58 : 59 ; 60 < 61 = 62 > 63 ? */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 0, -/* 64 @ 65 A 66 B 67 C 68 D 69 E 70 F 71 G */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, -/* 72 H 73 I 74 J 75 K 76 L 77 M 78 N 79 O */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, -/* 80 P 81 Q 82 R 83 S 84 T 85 U 86 V 87 W */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, -/* 88 X 89 Y 90 Z 91 [ 92 \ 93 ] 94 ^ 95 _ */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, -/* 96 ` 97 a 98 b 99 c 100 d 101 e 102 f 103 g */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, -/* 104 h 105 i 106 j 107 k 108 l 109 m 110 n 111 o */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, -/* 112 p 113 q 114 r 115 s 116 t 117 u 118 v 119 w */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128, -/* 120 x 121 y 122 z 123 { 124 | 125 } 126 ~ 127 del */ - 1 | 2 | 4 | 8 | 16 | 32 | 64 | 0, }; - -#undef T - -enum state - { s_dead = 1 /* important that this is > 0 */ - - , s_start_req_or_res - , s_res_or_resp_H - , s_start_res - , s_res_H - , s_res_HT - , s_res_HTT - , s_res_HTTP - , s_res_first_http_major - , s_res_http_major - , s_res_first_http_minor - , s_res_http_minor - , s_res_first_status_code - , s_res_status_code - , s_res_status - , s_res_line_almost_done - - , s_start_req - - , s_req_method - , s_req_spaces_before_url - , s_req_schema - , s_req_schema_slash - , s_req_schema_slash_slash - , s_req_server_start - , s_req_server - , s_req_server_with_at - , s_req_path - , s_req_query_string_start - , s_req_query_string - , s_req_fragment_start - , s_req_fragment - , s_req_http_start - , s_req_http_H - , s_req_http_HT - , s_req_http_HTT - , s_req_http_HTTP - , s_req_first_http_major - , s_req_http_major - , s_req_first_http_minor - , s_req_http_minor - , s_req_line_almost_done - - , s_header_field_start - , s_header_field - , s_header_value_start - , s_header_value - , s_header_value_lws - - , s_header_almost_done - - , s_chunk_size_start - , s_chunk_size - , s_chunk_parameters - , s_chunk_size_almost_done - - , s_headers_almost_done - , s_headers_done - - /* Important: 's_headers_done' must be the last 'header' state. All - * states beyond this must be 'body' states. It is used for overflow - * checking. See the PARSING_HEADER() macro. - */ - - , s_chunk_data - , s_chunk_data_almost_done - , s_chunk_data_done - - , s_body_identity - , s_body_identity_eof - - , s_message_done - }; - - -#define PARSING_HEADER(state) (state <= s_headers_done) - - -enum header_states - { h_general = 0 - , h_C - , h_CO - , h_CON - - , h_matching_connection - , h_matching_proxy_connection - , h_matching_content_length - , h_matching_transfer_encoding - , h_matching_upgrade - - , h_connection - , h_content_length - , h_transfer_encoding - , h_upgrade - - , h_matching_transfer_encoding_chunked - , h_matching_connection_keep_alive - , h_matching_connection_close - - , h_transfer_encoding_chunked - , h_connection_keep_alive - , h_connection_close - }; - -enum http_host_state - { - s_http_host_dead = 1 - , s_http_userinfo_start - , s_http_userinfo - , s_http_host_start - , s_http_host_v6_start - , s_http_host - , s_http_host_v6 - , s_http_host_v6_end - , s_http_host_port_start - , s_http_host_port -}; - -/* Macros for character classes; depends on strict-mode */ -#define CR '\r' -#define LF '\n' -#define LOWER(c) (unsigned char)(c | 0x20) -#define IS_ALPHA(c) (LOWER(c) >= 'a' && LOWER(c) <= 'z') -#define IS_NUM(c) ((c) >= '0' && (c) <= '9') -#define IS_ALPHANUM(c) (IS_ALPHA(c) || IS_NUM(c)) -#define IS_HEX(c) (IS_NUM(c) || (LOWER(c) >= 'a' && LOWER(c) <= 'f')) -#define IS_MARK(c) ((c) == '-' || (c) == '_' || (c) == '.' || \ - (c) == '!' || (c) == '~' || (c) == '*' || (c) == '\'' || (c) == '(' || \ - (c) == ')') -#define IS_USERINFO_CHAR(c) (IS_ALPHANUM(c) || IS_MARK(c) || (c) == '%' || \ - (c) == ';' || (c) == ':' || (c) == '&' || (c) == '=' || (c) == '+' || \ - (c) == '$' || (c) == ',') - -#if HTTP_PARSER_STRICT -#define TOKEN(c) (tokens[(unsigned char)c]) -#define IS_URL_CHAR(c) (BIT_AT(normal_url_char, (unsigned char)c)) -#define IS_HOST_CHAR(c) (IS_ALPHANUM(c) || (c) == '.' || (c) == '-') -#else -#define TOKEN(c) ((c == ' ') ? ' ' : tokens[(unsigned char)c]) -#define IS_URL_CHAR(c) \ - (BIT_AT(normal_url_char, (unsigned char)c) || ((c) & 0x80)) -#define IS_HOST_CHAR(c) \ - (IS_ALPHANUM(c) || (c) == '.' || (c) == '-' || (c) == '_') -#endif - - -#define start_state (parser->type == HTTP_REQUEST ? s_start_req : s_start_res) - - -#if HTTP_PARSER_STRICT -# define STRICT_CHECK(cond) \ -do { \ - if (cond) { \ - SET_ERRNO(HPE_STRICT); \ - goto error; \ - } \ -} while (0) -# define NEW_MESSAGE() (http_should_keep_alive(parser) ? start_state : s_dead) -#else -# define STRICT_CHECK(cond) -# define NEW_MESSAGE() start_state -#endif - - -/* Map errno values to strings for human-readable output */ -#define HTTP_STRERROR_GEN(n, s) { "HPE_" #n, s }, -static struct { - const char *name; - const char *description; -} http_strerror_tab[] = { - HTTP_ERRNO_MAP(HTTP_STRERROR_GEN) -}; -#undef HTTP_STRERROR_GEN - -int http_message_needs_eof(const http_parser *parser); - -/* Our URL parser. - * - * This is designed to be shared by http_parser_execute() for URL validation, - * hence it has a state transition + byte-for-byte interface. In addition, it - * is meant to be embedded in http_parser_parse_url(), which does the dirty - * work of turning state transitions URL components for its API. - * - * This function should only be invoked with non-space characters. It is - * assumed that the caller cares about (and can detect) the transition between - * URL and non-URL states by looking for these. - */ -static enum state -parse_url_char(enum state s, const char ch) -{ - if (ch == ' ' || ch == '\r' || ch == '\n') { - return s_dead; - } - -#if HTTP_PARSER_STRICT - if (ch == '\t' || ch == '\f') { - return s_dead; - } -#endif - - switch (s) { - case s_req_spaces_before_url: - /* Proxied requests are followed by scheme of an absolute URI (alpha). - * All methods except CONNECT are followed by '/' or '*'. - */ - - if (ch == '/' || ch == '*') { - return s_req_path; - } - - if (IS_ALPHA(ch)) { - return s_req_schema; - } - - break; - - case s_req_schema: - if (IS_ALPHA(ch)) { - return s; - } - - if (ch == ':') { - return s_req_schema_slash; - } - - break; - - case s_req_schema_slash: - if (ch == '/') { - return s_req_schema_slash_slash; - } - - break; - - case s_req_schema_slash_slash: - if (ch == '/') { - return s_req_server_start; - } - - break; - - case s_req_server_with_at: - if (ch == '@') { - return s_dead; - } - - /* FALLTHROUGH */ - case s_req_server_start: - case s_req_server: - if (ch == '/') { - return s_req_path; - } - - if (ch == '?') { - return s_req_query_string_start; - } - - if (ch == '@') { - return s_req_server_with_at; - } - - if (IS_USERINFO_CHAR(ch) || ch == '[' || ch == ']') { - return s_req_server; - } - - break; - - case s_req_path: - if (IS_URL_CHAR(ch)) { - return s; - } - - switch (ch) { - case '?': - return s_req_query_string_start; - - case '#': - return s_req_fragment_start; - } - - break; - - case s_req_query_string_start: - case s_req_query_string: - if (IS_URL_CHAR(ch)) { - return s_req_query_string; - } - - switch (ch) { - case '?': - /* allow extra '?' in query string */ - return s_req_query_string; - - case '#': - return s_req_fragment_start; - } - - break; - - case s_req_fragment_start: - if (IS_URL_CHAR(ch)) { - return s_req_fragment; - } - - switch (ch) { - case '?': - return s_req_fragment; - - case '#': - return s; - } - - break; - - case s_req_fragment: - if (IS_URL_CHAR(ch)) { - return s; - } - - switch (ch) { - case '?': - case '#': - return s; - } - - break; - - default: - break; - } - - /* We should never fall out of the switch above unless there's an error */ - return s_dead; -} - -size_t http_parser_execute (http_parser *parser, - const http_parser_settings *settings, - const char *data, - size_t len) -{ - char c, ch; - int8_t unhex_val; - const char *p = data; - const char *header_field_mark = 0; - const char *header_value_mark = 0; - const char *url_mark = 0; - const char *body_mark = 0; - - /* We're in an error state. Don't bother doing anything. */ - if (HTTP_PARSER_ERRNO(parser) != HPE_OK) { - return 0; - } - - if (len == 0) { - switch (parser->state) { - case s_body_identity_eof: - /* Use of CALLBACK_NOTIFY() here would erroneously return 1 byte read if - * we got paused. - */ - CALLBACK_NOTIFY_NOADVANCE(message_complete); - return 0; - - case s_dead: - case s_start_req_or_res: - case s_start_res: - case s_start_req: - return 0; - - default: - SET_ERRNO(HPE_INVALID_EOF_STATE); - return 1; - } - } - - - if (parser->state == s_header_field) - header_field_mark = data; - if (parser->state == s_header_value) - header_value_mark = data; - switch (parser->state) { - case s_req_path: - case s_req_schema: - case s_req_schema_slash: - case s_req_schema_slash_slash: - case s_req_server_start: - case s_req_server: - case s_req_server_with_at: - case s_req_query_string_start: - case s_req_query_string: - case s_req_fragment_start: - case s_req_fragment: - url_mark = data; - break; - } - - for (p=data; p != data + len; p++) { - ch = *p; - - if (PARSING_HEADER(parser->state)) { - ++parser->nread; - /* Buffer overflow attack */ - if (parser->nread > HTTP_MAX_HEADER_SIZE) { - SET_ERRNO(HPE_HEADER_OVERFLOW); - goto error; - } - } - - reexecute_byte: - switch (parser->state) { - - case s_dead: - /* this state is used after a 'Connection: close' message - * the parser will error out if it reads another message - */ - if (ch == CR || ch == LF) - break; - - SET_ERRNO(HPE_CLOSED_CONNECTION); - goto error; - - case s_start_req_or_res: - { - if (ch == CR || ch == LF) - break; - parser->flags = 0; - parser->content_length = ULLONG_MAX; - - if (ch == 'H') { - parser->state = s_res_or_resp_H; - - CALLBACK_NOTIFY(message_begin); - } else { - parser->type = HTTP_REQUEST; - parser->state = s_start_req; - goto reexecute_byte; - } - - break; - } - - case s_res_or_resp_H: - if (ch == 'T') { - parser->type = HTTP_RESPONSE; - parser->state = s_res_HT; - } else { - if (ch != 'E') { - SET_ERRNO(HPE_INVALID_CONSTANT); - goto error; - } - - parser->type = HTTP_REQUEST; - parser->method = HTTP_HEAD; - parser->index = 2; - parser->state = s_req_method; - } - break; - - case s_start_res: - { - parser->flags = 0; - parser->content_length = ULLONG_MAX; - - switch (ch) { - case 'H': - parser->state = s_res_H; - break; - - case CR: - case LF: - break; - - default: - SET_ERRNO(HPE_INVALID_CONSTANT); - goto error; - } - - CALLBACK_NOTIFY(message_begin); - break; - } - - case s_res_H: - STRICT_CHECK(ch != 'T'); - parser->state = s_res_HT; - break; - - case s_res_HT: - STRICT_CHECK(ch != 'T'); - parser->state = s_res_HTT; - break; - - case s_res_HTT: - STRICT_CHECK(ch != 'P'); - parser->state = s_res_HTTP; - break; - - case s_res_HTTP: - STRICT_CHECK(ch != '/'); - parser->state = s_res_first_http_major; - break; - - case s_res_first_http_major: - if (ch < '0' || ch > '9') { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - parser->http_major = ch - '0'; - parser->state = s_res_http_major; - break; - - /* major HTTP version or dot */ - case s_res_http_major: - { - if (ch == '.') { - parser->state = s_res_first_http_minor; - break; - } - - if (!IS_NUM(ch)) { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - parser->http_major *= 10; - parser->http_major += ch - '0'; - - if (parser->http_major > 999) { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - break; - } - - /* first digit of minor HTTP version */ - case s_res_first_http_minor: - if (!IS_NUM(ch)) { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - parser->http_minor = ch - '0'; - parser->state = s_res_http_minor; - break; - - /* minor HTTP version or end of request line */ - case s_res_http_minor: - { - if (ch == ' ') { - parser->state = s_res_first_status_code; - break; - } - - if (!IS_NUM(ch)) { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - parser->http_minor *= 10; - parser->http_minor += ch - '0'; - - if (parser->http_minor > 999) { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - break; - } - - case s_res_first_status_code: - { - if (!IS_NUM(ch)) { - if (ch == ' ') { - break; - } - - SET_ERRNO(HPE_INVALID_STATUS); - goto error; - } - parser->status_code = ch - '0'; - parser->state = s_res_status_code; - break; - } - - case s_res_status_code: - { - if (!IS_NUM(ch)) { - switch (ch) { - case ' ': - parser->state = s_res_status; - break; - case CR: - parser->state = s_res_line_almost_done; - break; - case LF: - parser->state = s_header_field_start; - break; - default: - SET_ERRNO(HPE_INVALID_STATUS); - goto error; - } - break; - } - - parser->status_code *= 10; - parser->status_code += ch - '0'; - - if (parser->status_code > 999) { - SET_ERRNO(HPE_INVALID_STATUS); - goto error; - } - - break; - } - - case s_res_status: - /* the human readable status. e.g. "NOT FOUND" - * we are not humans so just ignore this */ - if (ch == CR) { - parser->state = s_res_line_almost_done; - break; - } - - if (ch == LF) { - parser->state = s_header_field_start; - break; - } - break; - - case s_res_line_almost_done: - STRICT_CHECK(ch != LF); - parser->state = s_header_field_start; - CALLBACK_NOTIFY(status_complete); - break; - - case s_start_req: - { - if (ch == CR || ch == LF) - break; - parser->flags = 0; - parser->content_length = ULLONG_MAX; - - if (!IS_ALPHA(ch)) { - SET_ERRNO(HPE_INVALID_METHOD); - goto error; - } - - parser->method = (enum http_method) 0; - parser->index = 1; - switch (ch) { - case 'C': parser->method = HTTP_CONNECT; /* or COPY, CHECKOUT */ break; - case 'D': parser->method = HTTP_DELETE; break; - case 'G': parser->method = HTTP_GET; break; - case 'H': parser->method = HTTP_HEAD; break; - case 'L': parser->method = HTTP_LOCK; break; - case 'M': parser->method = HTTP_MKCOL; /* or MOVE, MKACTIVITY, MERGE, M-SEARCH */ break; - case 'N': parser->method = HTTP_NOTIFY; break; - case 'O': parser->method = HTTP_OPTIONS; break; - case 'P': parser->method = HTTP_POST; - /* or PROPFIND|PROPPATCH|PUT|PATCH|PURGE */ - break; - case 'R': parser->method = HTTP_REPORT; break; - case 'S': parser->method = HTTP_SUBSCRIBE; /* or SEARCH */ break; - case 'T': parser->method = HTTP_TRACE; break; - case 'U': parser->method = HTTP_UNLOCK; /* or UNSUBSCRIBE */ break; - default: - SET_ERRNO(HPE_INVALID_METHOD); - goto error; - } - parser->state = s_req_method; - - CALLBACK_NOTIFY(message_begin); - - break; - } - - case s_req_method: - { - const char *matcher; - if (ch == '\0') { - SET_ERRNO(HPE_INVALID_METHOD); - goto error; - } - - matcher = method_strings[parser->method]; - if (ch == ' ' && matcher[parser->index] == '\0') { - parser->state = s_req_spaces_before_url; - } else if (ch == matcher[parser->index]) { - ; /* nada */ - } else if (parser->method == HTTP_CONNECT) { - if (parser->index == 1 && ch == 'H') { - parser->method = HTTP_CHECKOUT; - } else if (parser->index == 2 && ch == 'P') { - parser->method = HTTP_COPY; - } else { - goto error; - } - } else if (parser->method == HTTP_MKCOL) { - if (parser->index == 1 && ch == 'O') { - parser->method = HTTP_MOVE; - } else if (parser->index == 1 && ch == 'E') { - parser->method = HTTP_MERGE; - } else if (parser->index == 1 && ch == '-') { - parser->method = HTTP_MSEARCH; - } else if (parser->index == 2 && ch == 'A') { - parser->method = HTTP_MKACTIVITY; - } else { - goto error; - } - } else if (parser->method == HTTP_SUBSCRIBE) { - if (parser->index == 1 && ch == 'E') { - parser->method = HTTP_SEARCH; - } else { - goto error; - } - } else if (parser->index == 1 && parser->method == HTTP_POST) { - if (ch == 'R') { - parser->method = HTTP_PROPFIND; /* or HTTP_PROPPATCH */ - } else if (ch == 'U') { - parser->method = HTTP_PUT; /* or HTTP_PURGE */ - } else if (ch == 'A') { - parser->method = HTTP_PATCH; - } else { - goto error; - } - } else if (parser->index == 2) { - if (parser->method == HTTP_PUT) { - if (ch == 'R') parser->method = HTTP_PURGE; - } else if (parser->method == HTTP_UNLOCK) { - if (ch == 'S') parser->method = HTTP_UNSUBSCRIBE; - } - } else if (parser->index == 4 && parser->method == HTTP_PROPFIND && ch == 'P') { - parser->method = HTTP_PROPPATCH; - } else { - SET_ERRNO(HPE_INVALID_METHOD); - goto error; - } - - ++parser->index; - break; - } - - case s_req_spaces_before_url: - { - if (ch == ' ') break; - - MARK(url); - if (parser->method == HTTP_CONNECT) { - parser->state = s_req_server_start; - } - - parser->state = parse_url_char((enum state)parser->state, ch); - if (parser->state == s_dead) { - SET_ERRNO(HPE_INVALID_URL); - goto error; - } - - break; - } - - case s_req_schema: - case s_req_schema_slash: - case s_req_schema_slash_slash: - case s_req_server_start: - { - switch (ch) { - /* No whitespace allowed here */ - case ' ': - case CR: - case LF: - SET_ERRNO(HPE_INVALID_URL); - goto error; - default: - parser->state = parse_url_char((enum state)parser->state, ch); - if (parser->state == s_dead) { - SET_ERRNO(HPE_INVALID_URL); - goto error; - } - } - - break; - } - - case s_req_server: - case s_req_server_with_at: - case s_req_path: - case s_req_query_string_start: - case s_req_query_string: - case s_req_fragment_start: - case s_req_fragment: - { - switch (ch) { - case ' ': - parser->state = s_req_http_start; - CALLBACK_DATA(url); - break; - case CR: - case LF: - parser->http_major = 0; - parser->http_minor = 9; - parser->state = (ch == CR) ? - s_req_line_almost_done : - s_header_field_start; - CALLBACK_DATA(url); - break; - default: - parser->state = parse_url_char((enum state)parser->state, ch); - if (parser->state == s_dead) { - SET_ERRNO(HPE_INVALID_URL); - goto error; - } - } - break; - } - - case s_req_http_start: - switch (ch) { - case 'H': - parser->state = s_req_http_H; - break; - case ' ': - break; - default: - SET_ERRNO(HPE_INVALID_CONSTANT); - goto error; - } - break; - - case s_req_http_H: - STRICT_CHECK(ch != 'T'); - parser->state = s_req_http_HT; - break; - - case s_req_http_HT: - STRICT_CHECK(ch != 'T'); - parser->state = s_req_http_HTT; - break; - - case s_req_http_HTT: - STRICT_CHECK(ch != 'P'); - parser->state = s_req_http_HTTP; - break; - - case s_req_http_HTTP: - STRICT_CHECK(ch != '/'); - parser->state = s_req_first_http_major; - break; - - /* first digit of major HTTP version */ - case s_req_first_http_major: - if (ch < '1' || ch > '9') { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - parser->http_major = ch - '0'; - parser->state = s_req_http_major; - break; - - /* major HTTP version or dot */ - case s_req_http_major: - { - if (ch == '.') { - parser->state = s_req_first_http_minor; - break; - } - - if (!IS_NUM(ch)) { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - parser->http_major *= 10; - parser->http_major += ch - '0'; - - if (parser->http_major > 999) { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - break; - } - - /* first digit of minor HTTP version */ - case s_req_first_http_minor: - if (!IS_NUM(ch)) { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - parser->http_minor = ch - '0'; - parser->state = s_req_http_minor; - break; - - /* minor HTTP version or end of request line */ - case s_req_http_minor: - { - if (ch == CR) { - parser->state = s_req_line_almost_done; - break; - } - - if (ch == LF) { - parser->state = s_header_field_start; - break; - } - - /* XXX allow spaces after digit? */ - - if (!IS_NUM(ch)) { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - parser->http_minor *= 10; - parser->http_minor += ch - '0'; - - if (parser->http_minor > 999) { - SET_ERRNO(HPE_INVALID_VERSION); - goto error; - } - - break; - } - - /* end of request line */ - case s_req_line_almost_done: - { - if (ch != LF) { - SET_ERRNO(HPE_LF_EXPECTED); - goto error; - } - - parser->state = s_header_field_start; - break; - } - - case s_header_field_start: - { - if (ch == CR) { - parser->state = s_headers_almost_done; - break; - } - - if (ch == LF) { - /* they might be just sending \n instead of \r\n so this would be - * the second \n to denote the end of headers*/ - parser->state = s_headers_almost_done; - goto reexecute_byte; - } - - c = TOKEN(ch); - - if (!c) { - SET_ERRNO(HPE_INVALID_HEADER_TOKEN); - goto error; - } - - MARK(header_field); - - parser->index = 0; - parser->state = s_header_field; - - switch (c) { - case 'c': - parser->header_state = h_C; - break; - - case 'p': - parser->header_state = h_matching_proxy_connection; - break; - - case 't': - parser->header_state = h_matching_transfer_encoding; - break; - - case 'u': - parser->header_state = h_matching_upgrade; - break; - - default: - parser->header_state = h_general; - break; - } - break; - } - - case s_header_field: - { - c = TOKEN(ch); - - if (c) { - switch (parser->header_state) { - case h_general: - break; - - case h_C: - parser->index++; - parser->header_state = (c == 'o' ? h_CO : h_general); - break; - - case h_CO: - parser->index++; - parser->header_state = (c == 'n' ? h_CON : h_general); - break; - - case h_CON: - parser->index++; - switch (c) { - case 'n': - parser->header_state = h_matching_connection; - break; - case 't': - parser->header_state = h_matching_content_length; - break; - default: - parser->header_state = h_general; - break; - } - break; - - /* connection */ - - case h_matching_connection: - parser->index++; - if (parser->index > sizeof(CONNECTION)-1 - || c != CONNECTION[parser->index]) { - parser->header_state = h_general; - } else if (parser->index == sizeof(CONNECTION)-2) { - parser->header_state = h_connection; - } - break; - - /* proxy-connection */ - - case h_matching_proxy_connection: - parser->index++; - if (parser->index > sizeof(PROXY_CONNECTION)-1 - || c != PROXY_CONNECTION[parser->index]) { - parser->header_state = h_general; - } else if (parser->index == sizeof(PROXY_CONNECTION)-2) { - parser->header_state = h_connection; - } - break; - - /* content-length */ - - case h_matching_content_length: - parser->index++; - if (parser->index > sizeof(CONTENT_LENGTH)-1 - || c != CONTENT_LENGTH[parser->index]) { - parser->header_state = h_general; - } else if (parser->index == sizeof(CONTENT_LENGTH)-2) { - parser->header_state = h_content_length; - } - break; - - /* transfer-encoding */ - - case h_matching_transfer_encoding: - parser->index++; - if (parser->index > sizeof(TRANSFER_ENCODING)-1 - || c != TRANSFER_ENCODING[parser->index]) { - parser->header_state = h_general; - } else if (parser->index == sizeof(TRANSFER_ENCODING)-2) { - parser->header_state = h_transfer_encoding; - } - break; - - /* upgrade */ - - case h_matching_upgrade: - parser->index++; - if (parser->index > sizeof(UPGRADE)-1 - || c != UPGRADE[parser->index]) { - parser->header_state = h_general; - } else if (parser->index == sizeof(UPGRADE)-2) { - parser->header_state = h_upgrade; - } - break; - - case h_connection: - case h_content_length: - case h_transfer_encoding: - case h_upgrade: - if (ch != ' ') parser->header_state = h_general; - break; - - default: - assert(0 && "Unknown header_state"); - break; - } - break; - } - - if (ch == ':') { - parser->state = s_header_value_start; - CALLBACK_DATA(header_field); - break; - } - - if (ch == CR) { - parser->state = s_header_almost_done; - CALLBACK_DATA(header_field); - break; - } - - if (ch == LF) { - parser->state = s_header_field_start; - CALLBACK_DATA(header_field); - break; - } - - SET_ERRNO(HPE_INVALID_HEADER_TOKEN); - goto error; - } - - case s_header_value_start: - { - if (ch == ' ' || ch == '\t') break; - - MARK(header_value); - - parser->state = s_header_value; - parser->index = 0; - - if (ch == CR) { - parser->header_state = h_general; - parser->state = s_header_almost_done; - CALLBACK_DATA(header_value); - break; - } - - if (ch == LF) { - parser->state = s_header_field_start; - CALLBACK_DATA(header_value); - break; - } - - c = LOWER(ch); - - switch (parser->header_state) { - case h_upgrade: - parser->flags |= F_UPGRADE; - parser->header_state = h_general; - break; - - case h_transfer_encoding: - /* looking for 'Transfer-Encoding: chunked' */ - if ('c' == c) { - parser->header_state = h_matching_transfer_encoding_chunked; - } else { - parser->header_state = h_general; - } - break; - - case h_content_length: - if (!IS_NUM(ch)) { - SET_ERRNO(HPE_INVALID_CONTENT_LENGTH); - goto error; - } - - parser->content_length = ch - '0'; - break; - - case h_connection: - /* looking for 'Connection: keep-alive' */ - if (c == 'k') { - parser->header_state = h_matching_connection_keep_alive; - /* looking for 'Connection: close' */ - } else if (c == 'c') { - parser->header_state = h_matching_connection_close; - } else { - parser->header_state = h_general; - } - break; - - default: - parser->header_state = h_general; - break; - } - break; - } - - case s_header_value: - { - - if (ch == CR) { - parser->state = s_header_almost_done; - CALLBACK_DATA(header_value); - break; - } - - if (ch == LF) { - parser->state = s_header_almost_done; - CALLBACK_DATA_NOADVANCE(header_value); - goto reexecute_byte; - } - - c = LOWER(ch); - - switch (parser->header_state) { - case h_general: - break; - - case h_connection: - case h_transfer_encoding: - assert(0 && "Shouldn't get here."); - break; - - case h_content_length: - { - uint64_t t; - - if (ch == ' ') break; - - if (!IS_NUM(ch)) { - SET_ERRNO(HPE_INVALID_CONTENT_LENGTH); - goto error; - } - - t = parser->content_length; - t *= 10; - t += ch - '0'; - - /* Overflow? */ - if (t < parser->content_length || t == ULLONG_MAX) { - SET_ERRNO(HPE_INVALID_CONTENT_LENGTH); - goto error; - } - - parser->content_length = t; - break; - } - - /* Transfer-Encoding: chunked */ - case h_matching_transfer_encoding_chunked: - parser->index++; - if (parser->index > sizeof(CHUNKED)-1 - || c != CHUNKED[parser->index]) { - parser->header_state = h_general; - } else if (parser->index == sizeof(CHUNKED)-2) { - parser->header_state = h_transfer_encoding_chunked; - } - break; - - /* looking for 'Connection: keep-alive' */ - case h_matching_connection_keep_alive: - parser->index++; - if (parser->index > sizeof(KEEP_ALIVE)-1 - || c != KEEP_ALIVE[parser->index]) { - parser->header_state = h_general; - } else if (parser->index == sizeof(KEEP_ALIVE)-2) { - parser->header_state = h_connection_keep_alive; - } - break; - - /* looking for 'Connection: close' */ - case h_matching_connection_close: - parser->index++; - if (parser->index > sizeof(CLOSE)-1 || c != CLOSE[parser->index]) { - parser->header_state = h_general; - } else if (parser->index == sizeof(CLOSE)-2) { - parser->header_state = h_connection_close; - } - break; - - case h_transfer_encoding_chunked: - case h_connection_keep_alive: - case h_connection_close: - if (ch != ' ') parser->header_state = h_general; - break; - - default: - parser->state = s_header_value; - parser->header_state = h_general; - break; - } - break; - } - - case s_header_almost_done: - { - STRICT_CHECK(ch != LF); - - parser->state = s_header_value_lws; - - switch (parser->header_state) { - case h_connection_keep_alive: - parser->flags |= F_CONNECTION_KEEP_ALIVE; - break; - case h_connection_close: - parser->flags |= F_CONNECTION_CLOSE; - break; - case h_transfer_encoding_chunked: - parser->flags |= F_CHUNKED; - break; - default: - break; - } - - break; - } - - case s_header_value_lws: - { - if (ch == ' ' || ch == '\t') - parser->state = s_header_value_start; - else - { - parser->state = s_header_field_start; - goto reexecute_byte; - } - break; - } - - case s_headers_almost_done: - { - STRICT_CHECK(ch != LF); - - if (parser->flags & F_TRAILING) { - /* End of a chunked request */ - parser->state = NEW_MESSAGE(); - CALLBACK_NOTIFY(message_complete); - break; - } - - parser->state = s_headers_done; - - /* Set this here so that on_headers_complete() callbacks can see it */ - parser->upgrade = - (parser->flags & F_UPGRADE || parser->method == HTTP_CONNECT); - - /* Here we call the headers_complete callback. This is somewhat - * different than other callbacks because if the user returns 1, we - * will interpret that as saying that this message has no body. This - * is needed for the annoying case of recieving a response to a HEAD - * request. - * - * We'd like to use CALLBACK_NOTIFY_NOADVANCE() here but we cannot, so - * we have to simulate it by handling a change in errno below. - */ - if (settings->on_headers_complete) { - switch (settings->on_headers_complete(parser)) { - case 0: - break; - - case 1: - parser->flags |= F_SKIPBODY; - break; - - default: - SET_ERRNO(HPE_CB_headers_complete); - return p - data; /* Error */ - } - } - - if (HTTP_PARSER_ERRNO(parser) != HPE_OK) { - return p - data; - } - - goto reexecute_byte; - } - - case s_headers_done: - { - STRICT_CHECK(ch != LF); - - parser->nread = 0; - - /* Exit, the rest of the connect is in a different protocol. */ - if (parser->upgrade) { - parser->state = NEW_MESSAGE(); - CALLBACK_NOTIFY(message_complete); - return (p - data) + 1; - } - - if (parser->flags & F_SKIPBODY) { - parser->state = NEW_MESSAGE(); - CALLBACK_NOTIFY(message_complete); - } else if (parser->flags & F_CHUNKED) { - /* chunked encoding - ignore Content-Length header */ - parser->state = s_chunk_size_start; - } else { - if (parser->content_length == 0) { - /* Content-Length header given but zero: Content-Length: 0\r\n */ - parser->state = NEW_MESSAGE(); - CALLBACK_NOTIFY(message_complete); - } else if (parser->content_length != ULLONG_MAX) { - /* Content-Length header given and non-zero */ - parser->state = s_body_identity; - } else { - if (parser->type == HTTP_REQUEST || - !http_message_needs_eof(parser)) { - /* Assume content-length 0 - read the next */ - parser->state = NEW_MESSAGE(); - CALLBACK_NOTIFY(message_complete); - } else { - /* Read body until EOF */ - parser->state = s_body_identity_eof; - } - } - } - - break; - } - - case s_body_identity: - { - uint64_t to_read = MIN(parser->content_length, - (uint64_t) ((data + len) - p)); - - assert(parser->content_length != 0 - && parser->content_length != ULLONG_MAX); - - /* The difference between advancing content_length and p is because - * the latter will automaticaly advance on the next loop iteration. - * Further, if content_length ends up at 0, we want to see the last - * byte again for our message complete callback. - */ - MARK(body); - parser->content_length -= to_read; - p += to_read - 1; - - if (parser->content_length == 0) { - parser->state = s_message_done; - - /* Mimic CALLBACK_DATA_NOADVANCE() but with one extra byte. - * - * The alternative to doing this is to wait for the next byte to - * trigger the data callback, just as in every other case. The - * problem with this is that this makes it difficult for the test - * harness to distinguish between complete-on-EOF and - * complete-on-length. It's not clear that this distinction is - * important for applications, but let's keep it for now. - */ - CALLBACK_DATA_(body, p - body_mark + 1, p - data); - goto reexecute_byte; - } - - break; - } - - /* read until EOF */ - case s_body_identity_eof: - MARK(body); - p = data + len - 1; - - break; - - case s_message_done: - parser->state = NEW_MESSAGE(); - CALLBACK_NOTIFY(message_complete); - break; - - case s_chunk_size_start: - { - assert(parser->nread == 1); - assert(parser->flags & F_CHUNKED); - - unhex_val = unhex[(unsigned char)ch]; - if (unhex_val == -1) { - SET_ERRNO(HPE_INVALID_CHUNK_SIZE); - goto error; - } - - parser->content_length = unhex_val; - parser->state = s_chunk_size; - break; - } - - case s_chunk_size: - { - uint64_t t; - - assert(parser->flags & F_CHUNKED); - - if (ch == CR) { - parser->state = s_chunk_size_almost_done; - break; - } - - unhex_val = unhex[(unsigned char)ch]; - - if (unhex_val == -1) { - if (ch == ';' || ch == ' ') { - parser->state = s_chunk_parameters; - break; - } - - SET_ERRNO(HPE_INVALID_CHUNK_SIZE); - goto error; - } - - t = parser->content_length; - t *= 16; - t += unhex_val; - - /* Overflow? */ - if (t < parser->content_length || t == ULLONG_MAX) { - SET_ERRNO(HPE_INVALID_CONTENT_LENGTH); - goto error; - } - - parser->content_length = t; - break; - } - - case s_chunk_parameters: - { - assert(parser->flags & F_CHUNKED); - /* just ignore this shit. TODO check for overflow */ - if (ch == CR) { - parser->state = s_chunk_size_almost_done; - break; - } - break; - } - - case s_chunk_size_almost_done: - { - assert(parser->flags & F_CHUNKED); - STRICT_CHECK(ch != LF); - - parser->nread = 0; - - if (parser->content_length == 0) { - parser->flags |= F_TRAILING; - parser->state = s_header_field_start; - } else { - parser->state = s_chunk_data; - } - break; - } - - case s_chunk_data: - { - uint64_t to_read = MIN(parser->content_length, - (uint64_t) ((data + len) - p)); - - assert(parser->flags & F_CHUNKED); - assert(parser->content_length != 0 - && parser->content_length != ULLONG_MAX); - - /* See the explanation in s_body_identity for why the content - * length and data pointers are managed this way. - */ - MARK(body); - parser->content_length -= to_read; - p += to_read - 1; - - if (parser->content_length == 0) { - parser->state = s_chunk_data_almost_done; - } - - break; - } - - case s_chunk_data_almost_done: - assert(parser->flags & F_CHUNKED); - assert(parser->content_length == 0); - STRICT_CHECK(ch != CR); - parser->state = s_chunk_data_done; - CALLBACK_DATA(body); - break; - - case s_chunk_data_done: - assert(parser->flags & F_CHUNKED); - STRICT_CHECK(ch != LF); - parser->nread = 0; - parser->state = s_chunk_size_start; - break; - - default: - assert(0 && "unhandled state"); - SET_ERRNO(HPE_INVALID_INTERNAL_STATE); - goto error; - } - } - - /* Run callbacks for any marks that we have leftover after we ran our of - * bytes. There should be at most one of these set, so it's OK to invoke - * them in series (unset marks will not result in callbacks). - * - * We use the NOADVANCE() variety of callbacks here because 'p' has already - * overflowed 'data' and this allows us to correct for the off-by-one that - * we'd otherwise have (since CALLBACK_DATA() is meant to be run with a 'p' - * value that's in-bounds). - */ - - assert(((header_field_mark ? 1 : 0) + - (header_value_mark ? 1 : 0) + - (url_mark ? 1 : 0) + - (body_mark ? 1 : 0)) <= 1); - - CALLBACK_DATA_NOADVANCE(header_field); - CALLBACK_DATA_NOADVANCE(header_value); - CALLBACK_DATA_NOADVANCE(url); - CALLBACK_DATA_NOADVANCE(body); - - return len; - -error: - if (HTTP_PARSER_ERRNO(parser) == HPE_OK) { - SET_ERRNO(HPE_UNKNOWN); - } - - return (p - data); -} - - -/* Does the parser need to see an EOF to find the end of the message? */ -int -http_message_needs_eof (const http_parser *parser) -{ - if (parser->type == HTTP_REQUEST) { - return 0; - } - - /* See RFC 2616 section 4.4 */ - if (parser->status_code / 100 == 1 || /* 1xx e.g. Continue */ - parser->status_code == 204 || /* No Content */ - parser->status_code == 304 || /* Not Modified */ - parser->flags & F_SKIPBODY) { /* response to a HEAD request */ - return 0; - } - - if ((parser->flags & F_CHUNKED) || parser->content_length != ULLONG_MAX) { - return 0; - } - - return 1; -} - - -int -http_should_keep_alive (const http_parser *parser) -{ - if (parser->http_major > 0 && parser->http_minor > 0) { - /* HTTP/1.1 */ - if (parser->flags & F_CONNECTION_CLOSE) { - return 0; - } - } else { - /* HTTP/1.0 or earlier */ - if (!(parser->flags & F_CONNECTION_KEEP_ALIVE)) { - return 0; - } - } - - return !http_message_needs_eof(parser); -} - - -const char * -http_method_str (enum http_method m) -{ - return ELEM_AT(method_strings, m, ""); -} - - -void -http_parser_init (http_parser *parser, enum http_parser_type t) -{ - void *data = parser->data; /* preserve application data */ - memset(parser, 0, sizeof(*parser)); - parser->data = data; - parser->type = t; - parser->state = (t == HTTP_REQUEST ? s_start_req : (t == HTTP_RESPONSE ? s_start_res : s_start_req_or_res)); - parser->http_errno = HPE_OK; -} - -const char * -http_errno_name(enum http_errno err) { - assert(err < (sizeof(http_strerror_tab)/sizeof(http_strerror_tab[0]))); - return http_strerror_tab[err].name; -} - -const char * -http_errno_description(enum http_errno err) { - assert(err < (sizeof(http_strerror_tab)/sizeof(http_strerror_tab[0]))); - return http_strerror_tab[err].description; -} - -static enum http_host_state -http_parse_host_char(enum http_host_state s, const char ch) { - switch(s) { - case s_http_userinfo: - case s_http_userinfo_start: - if (ch == '@') { - return s_http_host_start; - } - - if (IS_USERINFO_CHAR(ch)) { - return s_http_userinfo; - } - break; - - case s_http_host_start: - if (ch == '[') { - return s_http_host_v6_start; - } - - if (IS_HOST_CHAR(ch)) { - return s_http_host; - } - - break; - - case s_http_host: - if (IS_HOST_CHAR(ch)) { - return s_http_host; - } - - /* FALLTHROUGH */ - case s_http_host_v6_end: - if (ch == ':') { - return s_http_host_port_start; - } - - break; - - case s_http_host_v6: - if (ch == ']') { - return s_http_host_v6_end; - } - - /* FALLTHROUGH */ - case s_http_host_v6_start: - if (IS_HEX(ch) || ch == ':' || ch == '.') { - return s_http_host_v6; - } - - break; - - case s_http_host_port: - case s_http_host_port_start: - if (IS_NUM(ch)) { - return s_http_host_port; - } - - break; - - default: - break; - } - return s_http_host_dead; -} - -static int -http_parse_host(const char * buf, struct http_parser_url *u, int found_at) { - enum http_host_state s; - - const char *p; - size_t buflen = u->field_data[UF_HOST].off + u->field_data[UF_HOST].len; - - u->field_data[UF_HOST].len = 0; - - s = found_at ? s_http_userinfo_start : s_http_host_start; - - for (p = buf + u->field_data[UF_HOST].off; p < buf + buflen; p++) { - enum http_host_state new_s = http_parse_host_char(s, *p); - - if (new_s == s_http_host_dead) { - return 1; - } - - switch(new_s) { - case s_http_host: - if (s != s_http_host) { - u->field_data[UF_HOST].off = p - buf; - } - u->field_data[UF_HOST].len++; - break; - - case s_http_host_v6: - if (s != s_http_host_v6) { - u->field_data[UF_HOST].off = p - buf; - } - u->field_data[UF_HOST].len++; - break; - - case s_http_host_port: - if (s != s_http_host_port) { - u->field_data[UF_PORT].off = p - buf; - u->field_data[UF_PORT].len = 0; - u->field_set |= (1 << UF_PORT); - } - u->field_data[UF_PORT].len++; - break; - - case s_http_userinfo: - if (s != s_http_userinfo) { - u->field_data[UF_USERINFO].off = p - buf ; - u->field_data[UF_USERINFO].len = 0; - u->field_set |= (1 << UF_USERINFO); - } - u->field_data[UF_USERINFO].len++; - break; - - default: - break; - } - s = new_s; - } - - /* Make sure we don't end somewhere unexpected */ - switch (s) { - case s_http_host_start: - case s_http_host_v6_start: - case s_http_host_v6: - case s_http_host_port_start: - case s_http_userinfo: - case s_http_userinfo_start: - return 1; - default: - break; - } - - return 0; -} - -int -http_parser_parse_url(const char *buf, size_t buflen, int is_connect, - struct http_parser_url *u) -{ - enum state s; - const char *p; - enum http_parser_url_fields uf, old_uf; - int found_at = 0; - - u->port = u->field_set = 0; - s = is_connect ? s_req_server_start : s_req_spaces_before_url; - uf = old_uf = UF_MAX; - - for (p = buf; p < buf + buflen; p++) { - s = parse_url_char(s, *p); - - /* Figure out the next field that we're operating on */ - switch (s) { - case s_dead: - return 1; - - /* Skip delimeters */ - case s_req_schema_slash: - case s_req_schema_slash_slash: - case s_req_server_start: - case s_req_query_string_start: - case s_req_fragment_start: - continue; - - case s_req_schema: - uf = UF_SCHEMA; - break; - - case s_req_server_with_at: - found_at = 1; - - /* FALLTROUGH */ - case s_req_server: - uf = UF_HOST; - break; - - case s_req_path: - uf = UF_PATH; - break; - - case s_req_query_string: - uf = UF_QUERY; - break; - - case s_req_fragment: - uf = UF_FRAGMENT; - break; - - default: - assert(!"Unexpected state"); - return 1; - } - - /* Nothing's changed; soldier on */ - if (uf == old_uf) { - u->field_data[uf].len++; - continue; - } - - u->field_data[uf].off = p - buf; - u->field_data[uf].len = 1; - - u->field_set |= (1 << uf); - old_uf = uf; - } - - /* host must be present if there is a schema */ - /* parsing http:///toto will fail */ - if ((u->field_set & ((1 << UF_SCHEMA) | (1 << UF_HOST))) != 0) { - if (http_parse_host(buf, u, found_at) != 0) { - return 1; - } - } - - /* CONNECT requests can only contain "hostname:port" */ - if (is_connect && u->field_set != ((1 << UF_HOST)|(1 << UF_PORT))) { - return 1; - } - - if (u->field_set & (1 << UF_PORT)) { - /* Don't bother with endp; we've already validated the string */ - unsigned long v = strtoul(buf + u->field_data[UF_PORT].off, NULL, 10); - - /* Ports have a max value of 2^16 */ - if (v > 0xffff) { - return 1; - } - - u->port = (uint16_t) v; - } - - return 0; -} - -void -http_parser_pause(http_parser *parser, int paused) { - /* Users should only be pausing/unpausing a parser that is not in an error - * state. In non-debug builds, there's not much that we can do about this - * other than ignore it. - */ - if (HTTP_PARSER_ERRNO(parser) == HPE_OK || - HTTP_PARSER_ERRNO(parser) == HPE_PAUSED) { - SET_ERRNO((paused) ? HPE_PAUSED : HPE_OK); - } else { - assert(0 && "Attempting to pause parser in error state"); - } -} - -int -http_body_is_final(const struct http_parser *parser) { - return parser->state == s_message_done; -} diff --git a/ext/http-parser/http_parser.gyp b/ext/http-parser/http_parser.gyp deleted file mode 100644 index ef34ecae..00000000 --- a/ext/http-parser/http_parser.gyp +++ /dev/null @@ -1,111 +0,0 @@ -# This file is used with the GYP meta build system. -# http://code.google.com/p/gyp/ -# To build try this: -# svn co http://gyp.googlecode.com/svn/trunk gyp -# ./gyp/gyp -f make --depth=`pwd` http_parser.gyp -# ./out/Debug/test -{ - 'target_defaults': { - 'default_configuration': 'Debug', - 'configurations': { - # TODO: hoist these out and put them somewhere common, because - # RuntimeLibrary MUST MATCH across the entire project - 'Debug': { - 'defines': [ 'DEBUG', '_DEBUG' ], - 'cflags': [ '-Wall', '-Wextra', '-O0', '-g', '-ftrapv' ], - 'msvs_settings': { - 'VCCLCompilerTool': { - 'RuntimeLibrary': 1, # static debug - }, - }, - }, - 'Release': { - 'defines': [ 'NDEBUG' ], - 'cflags': [ '-Wall', '-Wextra', '-O3' ], - 'msvs_settings': { - 'VCCLCompilerTool': { - 'RuntimeLibrary': 0, # static release - }, - }, - } - }, - 'msvs_settings': { - 'VCCLCompilerTool': { - }, - 'VCLibrarianTool': { - }, - 'VCLinkerTool': { - 'GenerateDebugInformation': 'true', - }, - }, - 'conditions': [ - ['OS == "win"', { - 'defines': [ - 'WIN32' - ], - }] - ], - }, - - 'targets': [ - { - 'target_name': 'http_parser', - 'type': 'static_library', - 'include_dirs': [ '.' ], - 'direct_dependent_settings': { - 'defines': [ 'HTTP_PARSER_STRICT=0' ], - 'include_dirs': [ '.' ], - }, - 'defines': [ 'HTTP_PARSER_STRICT=0' ], - 'sources': [ './http_parser.c', ], - 'conditions': [ - ['OS=="win"', { - 'msvs_settings': { - 'VCCLCompilerTool': { - # Compile as C++. http_parser.c is actually C99, but C++ is - # close enough in this case. - 'CompileAs': 2, - }, - }, - }] - ], - }, - - { - 'target_name': 'http_parser_strict', - 'type': 'static_library', - 'include_dirs': [ '.' ], - 'direct_dependent_settings': { - 'defines': [ 'HTTP_PARSER_STRICT=1' ], - 'include_dirs': [ '.' ], - }, - 'defines': [ 'HTTP_PARSER_STRICT=1' ], - 'sources': [ './http_parser.c', ], - 'conditions': [ - ['OS=="win"', { - 'msvs_settings': { - 'VCCLCompilerTool': { - # Compile as C++. http_parser.c is actually C99, but C++ is - # close enough in this case. - 'CompileAs': 2, - }, - }, - }] - ], - }, - - { - 'target_name': 'test-nonstrict', - 'type': 'executable', - 'dependencies': [ 'http_parser' ], - 'sources': [ 'test.c' ] - }, - - { - 'target_name': 'test-strict', - 'type': 'executable', - 'dependencies': [ 'http_parser_strict' ], - 'sources': [ 'test.c' ] - } - ] -} diff --git a/ext/http-parser/http_parser.h b/ext/http-parser/http_parser.h deleted file mode 100644 index a992c742..00000000 --- a/ext/http-parser/http_parser.h +++ /dev/null @@ -1,304 +0,0 @@ -/* Copyright Joyent, Inc. and other Node contributors. All rights reserved. - * - * Permission is hereby granted, free of charge, to any person obtaining a copy - * of this software and associated documentation files (the "Software"), to - * deal in the Software without restriction, including without limitation the - * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or - * sell copies of the Software, and to permit persons to whom the Software is - * furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE - * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER - * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING - * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS - * IN THE SOFTWARE. - */ -#ifndef http_parser_h -#define http_parser_h -#ifdef __cplusplus -extern "C" { -#endif - -#define HTTP_PARSER_VERSION_MAJOR 2 -#define HTTP_PARSER_VERSION_MINOR 0 - -#include -#if defined(_WIN32) && !defined(__MINGW32__) && (!defined(_MSC_VER) || _MSC_VER<1600) -#include -#include -typedef __int8 int8_t; -typedef unsigned __int8 uint8_t; -typedef __int16 int16_t; -typedef unsigned __int16 uint16_t; -typedef __int32 int32_t; -typedef unsigned __int32 uint32_t; -typedef __int64 int64_t; -typedef unsigned __int64 uint64_t; -#else -#include -#endif - -/* Compile with -DHTTP_PARSER_STRICT=0 to make less checks, but run - * faster - */ -#ifndef HTTP_PARSER_STRICT -# define HTTP_PARSER_STRICT 1 -#endif - -/* Maximium header size allowed */ -#define HTTP_MAX_HEADER_SIZE (80*1024) - - -typedef struct http_parser http_parser; -typedef struct http_parser_settings http_parser_settings; - - -/* Callbacks should return non-zero to indicate an error. The parser will - * then halt execution. - * - * The one exception is on_headers_complete. In a HTTP_RESPONSE parser - * returning '1' from on_headers_complete will tell the parser that it - * should not expect a body. This is used when receiving a response to a - * HEAD request which may contain 'Content-Length' or 'Transfer-Encoding: - * chunked' headers that indicate the presence of a body. - * - * http_data_cb does not return data chunks. It will be call arbitrarally - * many times for each string. E.G. you might get 10 callbacks for "on_url" - * each providing just a few characters more data. - */ -typedef int (*http_data_cb) (http_parser*, const char *at, size_t length); -typedef int (*http_cb) (http_parser*); - - -/* Request Methods */ -#define HTTP_METHOD_MAP(XX) \ - XX(0, DELETE, DELETE) \ - XX(1, GET, GET) \ - XX(2, HEAD, HEAD) \ - XX(3, POST, POST) \ - XX(4, PUT, PUT) \ - /* pathological */ \ - XX(5, CONNECT, CONNECT) \ - XX(6, OPTIONS, OPTIONS) \ - XX(7, TRACE, TRACE) \ - /* webdav */ \ - XX(8, COPY, COPY) \ - XX(9, LOCK, LOCK) \ - XX(10, MKCOL, MKCOL) \ - XX(11, MOVE, MOVE) \ - XX(12, PROPFIND, PROPFIND) \ - XX(13, PROPPATCH, PROPPATCH) \ - XX(14, SEARCH, SEARCH) \ - XX(15, UNLOCK, UNLOCK) \ - /* subversion */ \ - XX(16, REPORT, REPORT) \ - XX(17, MKACTIVITY, MKACTIVITY) \ - XX(18, CHECKOUT, CHECKOUT) \ - XX(19, MERGE, MERGE) \ - /* upnp */ \ - XX(20, MSEARCH, M-SEARCH) \ - XX(21, NOTIFY, NOTIFY) \ - XX(22, SUBSCRIBE, SUBSCRIBE) \ - XX(23, UNSUBSCRIBE, UNSUBSCRIBE) \ - /* RFC-5789 */ \ - XX(24, PATCH, PATCH) \ - XX(25, PURGE, PURGE) \ - -enum http_method - { -#define XX(num, name, string) HTTP_##name = num, - HTTP_METHOD_MAP(XX) -#undef XX - }; - - -enum http_parser_type { HTTP_REQUEST, HTTP_RESPONSE, HTTP_BOTH }; - - -/* Flag values for http_parser.flags field */ -enum flags - { F_CHUNKED = 1 << 0 - , F_CONNECTION_KEEP_ALIVE = 1 << 1 - , F_CONNECTION_CLOSE = 1 << 2 - , F_TRAILING = 1 << 3 - , F_UPGRADE = 1 << 4 - , F_SKIPBODY = 1 << 5 - }; - - -/* Map for errno-related constants - * - * The provided argument should be a macro that takes 2 arguments. - */ -#define HTTP_ERRNO_MAP(XX) \ - /* No error */ \ - XX(OK, "success") \ - \ - /* Callback-related errors */ \ - XX(CB_message_begin, "the on_message_begin callback failed") \ - XX(CB_status_complete, "the on_status_complete callback failed") \ - XX(CB_url, "the on_url callback failed") \ - XX(CB_header_field, "the on_header_field callback failed") \ - XX(CB_header_value, "the on_header_value callback failed") \ - XX(CB_headers_complete, "the on_headers_complete callback failed") \ - XX(CB_body, "the on_body callback failed") \ - XX(CB_message_complete, "the on_message_complete callback failed") \ - \ - /* Parsing-related errors */ \ - XX(INVALID_EOF_STATE, "stream ended at an unexpected time") \ - XX(HEADER_OVERFLOW, \ - "too many header bytes seen; overflow detected") \ - XX(CLOSED_CONNECTION, \ - "data received after completed connection: close message") \ - XX(INVALID_VERSION, "invalid HTTP version") \ - XX(INVALID_STATUS, "invalid HTTP status code") \ - XX(INVALID_METHOD, "invalid HTTP method") \ - XX(INVALID_URL, "invalid URL") \ - XX(INVALID_HOST, "invalid host") \ - XX(INVALID_PORT, "invalid port") \ - XX(INVALID_PATH, "invalid path") \ - XX(INVALID_QUERY_STRING, "invalid query string") \ - XX(INVALID_FRAGMENT, "invalid fragment") \ - XX(LF_EXPECTED, "LF character expected") \ - XX(INVALID_HEADER_TOKEN, "invalid character in header") \ - XX(INVALID_CONTENT_LENGTH, \ - "invalid character in content-length header") \ - XX(INVALID_CHUNK_SIZE, \ - "invalid character in chunk size header") \ - XX(INVALID_CONSTANT, "invalid constant string") \ - XX(INVALID_INTERNAL_STATE, "encountered unexpected internal state")\ - XX(STRICT, "strict mode assertion failed") \ - XX(PAUSED, "parser is paused") \ - XX(UNKNOWN, "an unknown error occurred") - - -/* Define HPE_* values for each errno value above */ -#define HTTP_ERRNO_GEN(n, s) HPE_##n, -enum http_errno { - HTTP_ERRNO_MAP(HTTP_ERRNO_GEN) -}; -#undef HTTP_ERRNO_GEN - - -/* Get an http_errno value from an http_parser */ -#define HTTP_PARSER_ERRNO(p) ((enum http_errno) (p)->http_errno) - - -struct http_parser { - /** PRIVATE **/ - unsigned char type : 2; /* enum http_parser_type */ - unsigned char flags : 6; /* F_* values from 'flags' enum; semi-public */ - unsigned char state; /* enum state from http_parser.c */ - unsigned char header_state; /* enum header_state from http_parser.c */ - unsigned char index; /* index into current matcher */ - - uint32_t nread; /* # bytes read in various scenarios */ - uint64_t content_length; /* # bytes in body (0 if no Content-Length header) */ - - /** READ-ONLY **/ - unsigned short http_major; - unsigned short http_minor; - unsigned short status_code; /* responses only */ - unsigned char method; /* requests only */ - unsigned char http_errno : 7; - - /* 1 = Upgrade header was present and the parser has exited because of that. - * 0 = No upgrade header present. - * Should be checked when http_parser_execute() returns in addition to - * error checking. - */ - unsigned char upgrade : 1; - - /** PUBLIC **/ - void *data; /* A pointer to get hook to the "connection" or "socket" object */ -}; - - -struct http_parser_settings { - http_cb on_message_begin; - http_data_cb on_url; - http_cb on_status_complete; - http_data_cb on_header_field; - http_data_cb on_header_value; - http_cb on_headers_complete; - http_data_cb on_body; - http_cb on_message_complete; -}; - - -enum http_parser_url_fields - { UF_SCHEMA = 0 - , UF_HOST = 1 - , UF_PORT = 2 - , UF_PATH = 3 - , UF_QUERY = 4 - , UF_FRAGMENT = 5 - , UF_USERINFO = 6 - , UF_MAX = 7 - }; - - -/* Result structure for http_parser_parse_url(). - * - * Callers should index into field_data[] with UF_* values iff field_set - * has the relevant (1 << UF_*) bit set. As a courtesy to clients (and - * because we probably have padding left over), we convert any port to - * a uint16_t. - */ -struct http_parser_url { - uint16_t field_set; /* Bitmask of (1 << UF_*) values */ - uint16_t port; /* Converted UF_PORT string */ - - struct { - uint16_t off; /* Offset into buffer in which field starts */ - uint16_t len; /* Length of run in buffer */ - } field_data[UF_MAX]; -}; - - -void http_parser_init(http_parser *parser, enum http_parser_type type); - - -size_t http_parser_execute(http_parser *parser, - const http_parser_settings *settings, - const char *data, - size_t len); - - -/* If http_should_keep_alive() in the on_headers_complete or - * on_message_complete callback returns 0, then this should be - * the last message on the connection. - * If you are the server, respond with the "Connection: close" header. - * If you are the client, close the connection. - */ -int http_should_keep_alive(const http_parser *parser); - -/* Returns a string version of the HTTP method. */ -const char *http_method_str(enum http_method m); - -/* Return a string name of the given error */ -const char *http_errno_name(enum http_errno err); - -/* Return a string description of the given error */ -const char *http_errno_description(enum http_errno err); - -/* Parse a URL; return nonzero on failure */ -int http_parser_parse_url(const char *buf, size_t buflen, - int is_connect, - struct http_parser_url *u); - -/* Pause or un-pause the parser; a nonzero value pauses */ -void http_parser_pause(http_parser *parser, int paused); - -/* Checks if this is the final chunk of the body. */ -int http_body_is_final(const http_parser *parser); - -#ifdef __cplusplus -} -#endif -#endif diff --git a/ext/http-parser/test.c b/ext/http-parser/test.c deleted file mode 100644 index 83723b7f..00000000 --- a/ext/http-parser/test.c +++ /dev/null @@ -1,3425 +0,0 @@ -/* Copyright Joyent, Inc. and other Node contributors. All rights reserved. - * - * Permission is hereby granted, free of charge, to any person obtaining a copy - * of this software and associated documentation files (the "Software"), to - * deal in the Software without restriction, including without limitation the - * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or - * sell copies of the Software, and to permit persons to whom the Software is - * furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE - * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER - * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING - * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS - * IN THE SOFTWARE. - */ -#include "http_parser.h" -#include -#include -#include -#include /* rand */ -#include -#include - -#undef TRUE -#define TRUE 1 -#undef FALSE -#define FALSE 0 - -#define MAX_HEADERS 13 -#define MAX_ELEMENT_SIZE 2048 - -#define MIN(a,b) ((a) < (b) ? (a) : (b)) - -static http_parser *parser; - -struct message { - const char *name; // for debugging purposes - const char *raw; - enum http_parser_type type; - enum http_method method; - int status_code; - char request_path[MAX_ELEMENT_SIZE]; - char request_url[MAX_ELEMENT_SIZE]; - char fragment[MAX_ELEMENT_SIZE]; - char query_string[MAX_ELEMENT_SIZE]; - char body[MAX_ELEMENT_SIZE]; - size_t body_size; - const char *host; - const char *userinfo; - uint16_t port; - int num_headers; - enum { NONE=0, FIELD, VALUE } last_header_element; - char headers [MAX_HEADERS][2][MAX_ELEMENT_SIZE]; - int should_keep_alive; - - const char *upgrade; // upgraded body - - unsigned short http_major; - unsigned short http_minor; - - int message_begin_cb_called; - int headers_complete_cb_called; - int message_complete_cb_called; - int message_complete_on_eof; - int body_is_final; -}; - -static int currently_parsing_eof; - -static struct message messages[5]; -static int num_messages; -static http_parser_settings *current_pause_parser; - -/* * R E Q U E S T S * */ -const struct message requests[] = -#define CURL_GET 0 -{ {.name= "curl get" - ,.type= HTTP_REQUEST - ,.raw= "GET /test HTTP/1.1\r\n" - "User-Agent: curl/7.18.0 (i486-pc-linux-gnu) libcurl/7.18.0 OpenSSL/0.9.8g zlib/1.2.3.3 libidn/1.1\r\n" - "Host: 0.0.0.0=5000\r\n" - "Accept: */*\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/test" - ,.request_url= "/test" - ,.num_headers= 3 - ,.headers= - { { "User-Agent", "curl/7.18.0 (i486-pc-linux-gnu) libcurl/7.18.0 OpenSSL/0.9.8g zlib/1.2.3.3 libidn/1.1" } - , { "Host", "0.0.0.0=5000" } - , { "Accept", "*/*" } - } - ,.body= "" - } - -#define FIREFOX_GET 1 -, {.name= "firefox get" - ,.type= HTTP_REQUEST - ,.raw= "GET /favicon.ico HTTP/1.1\r\n" - "Host: 0.0.0.0=5000\r\n" - "User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9) Gecko/2008061015 Firefox/3.0\r\n" - "Accept: text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8\r\n" - "Accept-Language: en-us,en;q=0.5\r\n" - "Accept-Encoding: gzip,deflate\r\n" - "Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7\r\n" - "Keep-Alive: 300\r\n" - "Connection: keep-alive\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/favicon.ico" - ,.request_url= "/favicon.ico" - ,.num_headers= 8 - ,.headers= - { { "Host", "0.0.0.0=5000" } - , { "User-Agent", "Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9) Gecko/2008061015 Firefox/3.0" } - , { "Accept", "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8" } - , { "Accept-Language", "en-us,en;q=0.5" } - , { "Accept-Encoding", "gzip,deflate" } - , { "Accept-Charset", "ISO-8859-1,utf-8;q=0.7,*;q=0.7" } - , { "Keep-Alive", "300" } - , { "Connection", "keep-alive" } - } - ,.body= "" - } - -#define DUMBFUCK 2 -, {.name= "dumbfuck" - ,.type= HTTP_REQUEST - ,.raw= "GET /dumbfuck HTTP/1.1\r\n" - "aaaaaaaaaaaaa:++++++++++\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/dumbfuck" - ,.request_url= "/dumbfuck" - ,.num_headers= 1 - ,.headers= - { { "aaaaaaaaaaaaa", "++++++++++" } - } - ,.body= "" - } - -#define FRAGMENT_IN_URI 3 -, {.name= "fragment in url" - ,.type= HTTP_REQUEST - ,.raw= "GET /forums/1/topics/2375?page=1#posts-17408 HTTP/1.1\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "page=1" - ,.fragment= "posts-17408" - ,.request_path= "/forums/1/topics/2375" - /* XXX request url does include fragment? */ - ,.request_url= "/forums/1/topics/2375?page=1#posts-17408" - ,.num_headers= 0 - ,.body= "" - } - -#define GET_NO_HEADERS_NO_BODY 4 -, {.name= "get no headers no body" - ,.type= HTTP_REQUEST - ,.raw= "GET /get_no_headers_no_body/world HTTP/1.1\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE /* would need Connection: close */ - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/get_no_headers_no_body/world" - ,.request_url= "/get_no_headers_no_body/world" - ,.num_headers= 0 - ,.body= "" - } - -#define GET_ONE_HEADER_NO_BODY 5 -, {.name= "get one header no body" - ,.type= HTTP_REQUEST - ,.raw= "GET /get_one_header_no_body HTTP/1.1\r\n" - "Accept: */*\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE /* would need Connection: close */ - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/get_one_header_no_body" - ,.request_url= "/get_one_header_no_body" - ,.num_headers= 1 - ,.headers= - { { "Accept" , "*/*" } - } - ,.body= "" - } - -#define GET_FUNKY_CONTENT_LENGTH 6 -, {.name= "get funky content length body hello" - ,.type= HTTP_REQUEST - ,.raw= "GET /get_funky_content_length_body_hello HTTP/1.0\r\n" - "conTENT-Length: 5\r\n" - "\r\n" - "HELLO" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 0 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/get_funky_content_length_body_hello" - ,.request_url= "/get_funky_content_length_body_hello" - ,.num_headers= 1 - ,.headers= - { { "conTENT-Length" , "5" } - } - ,.body= "HELLO" - } - -#define POST_IDENTITY_BODY_WORLD 7 -, {.name= "post identity body world" - ,.type= HTTP_REQUEST - ,.raw= "POST /post_identity_body_world?q=search#hey HTTP/1.1\r\n" - "Accept: */*\r\n" - "Transfer-Encoding: identity\r\n" - "Content-Length: 5\r\n" - "\r\n" - "World" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_POST - ,.query_string= "q=search" - ,.fragment= "hey" - ,.request_path= "/post_identity_body_world" - ,.request_url= "/post_identity_body_world?q=search#hey" - ,.num_headers= 3 - ,.headers= - { { "Accept", "*/*" } - , { "Transfer-Encoding", "identity" } - , { "Content-Length", "5" } - } - ,.body= "World" - } - -#define POST_CHUNKED_ALL_YOUR_BASE 8 -, {.name= "post - chunked body: all your base are belong to us" - ,.type= HTTP_REQUEST - ,.raw= "POST /post_chunked_all_your_base HTTP/1.1\r\n" - "Transfer-Encoding: chunked\r\n" - "\r\n" - "1e\r\nall your base are belong to us\r\n" - "0\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_POST - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/post_chunked_all_your_base" - ,.request_url= "/post_chunked_all_your_base" - ,.num_headers= 1 - ,.headers= - { { "Transfer-Encoding" , "chunked" } - } - ,.body= "all your base are belong to us" - } - -#define TWO_CHUNKS_MULT_ZERO_END 9 -, {.name= "two chunks ; triple zero ending" - ,.type= HTTP_REQUEST - ,.raw= "POST /two_chunks_mult_zero_end HTTP/1.1\r\n" - "Transfer-Encoding: chunked\r\n" - "\r\n" - "5\r\nhello\r\n" - "6\r\n world\r\n" - "000\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_POST - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/two_chunks_mult_zero_end" - ,.request_url= "/two_chunks_mult_zero_end" - ,.num_headers= 1 - ,.headers= - { { "Transfer-Encoding", "chunked" } - } - ,.body= "hello world" - } - -#define CHUNKED_W_TRAILING_HEADERS 10 -, {.name= "chunked with trailing headers. blech." - ,.type= HTTP_REQUEST - ,.raw= "POST /chunked_w_trailing_headers HTTP/1.1\r\n" - "Transfer-Encoding: chunked\r\n" - "\r\n" - "5\r\nhello\r\n" - "6\r\n world\r\n" - "0\r\n" - "Vary: *\r\n" - "Content-Type: text/plain\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_POST - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/chunked_w_trailing_headers" - ,.request_url= "/chunked_w_trailing_headers" - ,.num_headers= 3 - ,.headers= - { { "Transfer-Encoding", "chunked" } - , { "Vary", "*" } - , { "Content-Type", "text/plain" } - } - ,.body= "hello world" - } - -#define CHUNKED_W_BULLSHIT_AFTER_LENGTH 11 -, {.name= "with bullshit after the length" - ,.type= HTTP_REQUEST - ,.raw= "POST /chunked_w_bullshit_after_length HTTP/1.1\r\n" - "Transfer-Encoding: chunked\r\n" - "\r\n" - "5; ihatew3;whatthefuck=aretheseparametersfor\r\nhello\r\n" - "6; blahblah; blah\r\n world\r\n" - "0\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_POST - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/chunked_w_bullshit_after_length" - ,.request_url= "/chunked_w_bullshit_after_length" - ,.num_headers= 1 - ,.headers= - { { "Transfer-Encoding", "chunked" } - } - ,.body= "hello world" - } - -#define WITH_QUOTES 12 -, {.name= "with quotes" - ,.type= HTTP_REQUEST - ,.raw= "GET /with_\"stupid\"_quotes?foo=\"bar\" HTTP/1.1\r\n\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "foo=\"bar\"" - ,.fragment= "" - ,.request_path= "/with_\"stupid\"_quotes" - ,.request_url= "/with_\"stupid\"_quotes?foo=\"bar\"" - ,.num_headers= 0 - ,.headers= { } - ,.body= "" - } - -#define APACHEBENCH_GET 13 -/* The server receiving this request SHOULD NOT wait for EOF - * to know that content-length == 0. - * How to represent this in a unit test? message_complete_on_eof - * Compare with NO_CONTENT_LENGTH_RESPONSE. - */ -, {.name = "apachebench get" - ,.type= HTTP_REQUEST - ,.raw= "GET /test HTTP/1.0\r\n" - "Host: 0.0.0.0:5000\r\n" - "User-Agent: ApacheBench/2.3\r\n" - "Accept: */*\r\n\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 0 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/test" - ,.request_url= "/test" - ,.num_headers= 3 - ,.headers= { { "Host", "0.0.0.0:5000" } - , { "User-Agent", "ApacheBench/2.3" } - , { "Accept", "*/*" } - } - ,.body= "" - } - -#define QUERY_URL_WITH_QUESTION_MARK_GET 14 -/* Some clients include '?' characters in query strings. - */ -, {.name = "query url with question mark" - ,.type= HTTP_REQUEST - ,.raw= "GET /test.cgi?foo=bar?baz HTTP/1.1\r\n\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "foo=bar?baz" - ,.fragment= "" - ,.request_path= "/test.cgi" - ,.request_url= "/test.cgi?foo=bar?baz" - ,.num_headers= 0 - ,.headers= {} - ,.body= "" - } - -#define PREFIX_NEWLINE_GET 15 -/* Some clients, especially after a POST in a keep-alive connection, - * will send an extra CRLF before the next request - */ -, {.name = "newline prefix get" - ,.type= HTTP_REQUEST - ,.raw= "\r\nGET /test HTTP/1.1\r\n\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/test" - ,.request_url= "/test" - ,.num_headers= 0 - ,.headers= { } - ,.body= "" - } - -#define UPGRADE_REQUEST 16 -, {.name = "upgrade request" - ,.type= HTTP_REQUEST - ,.raw= "GET /demo HTTP/1.1\r\n" - "Host: example.com\r\n" - "Connection: Upgrade\r\n" - "Sec-WebSocket-Key2: 12998 5 Y3 1 .P00\r\n" - "Sec-WebSocket-Protocol: sample\r\n" - "Upgrade: WebSocket\r\n" - "Sec-WebSocket-Key1: 4 @1 46546xW%0l 1 5\r\n" - "Origin: http://example.com\r\n" - "\r\n" - "Hot diggity dogg" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/demo" - ,.request_url= "/demo" - ,.num_headers= 7 - ,.upgrade="Hot diggity dogg" - ,.headers= { { "Host", "example.com" } - , { "Connection", "Upgrade" } - , { "Sec-WebSocket-Key2", "12998 5 Y3 1 .P00" } - , { "Sec-WebSocket-Protocol", "sample" } - , { "Upgrade", "WebSocket" } - , { "Sec-WebSocket-Key1", "4 @1 46546xW%0l 1 5" } - , { "Origin", "http://example.com" } - } - ,.body= "" - } - -#define CONNECT_REQUEST 17 -, {.name = "connect request" - ,.type= HTTP_REQUEST - ,.raw= "CONNECT 0-home0.netscape.com:443 HTTP/1.0\r\n" - "User-agent: Mozilla/1.1N\r\n" - "Proxy-authorization: basic aGVsbG86d29ybGQ=\r\n" - "\r\n" - "some data\r\n" - "and yet even more data" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 0 - ,.method= HTTP_CONNECT - ,.query_string= "" - ,.fragment= "" - ,.request_path= "" - ,.request_url= "0-home0.netscape.com:443" - ,.num_headers= 2 - ,.upgrade="some data\r\nand yet even more data" - ,.headers= { { "User-agent", "Mozilla/1.1N" } - , { "Proxy-authorization", "basic aGVsbG86d29ybGQ=" } - } - ,.body= "" - } - -#define REPORT_REQ 18 -, {.name= "report request" - ,.type= HTTP_REQUEST - ,.raw= "REPORT /test HTTP/1.1\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_REPORT - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/test" - ,.request_url= "/test" - ,.num_headers= 0 - ,.headers= {} - ,.body= "" - } - -#define NO_HTTP_VERSION 19 -, {.name= "request with no http version" - ,.type= HTTP_REQUEST - ,.raw= "GET /\r\n" - "\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 0 - ,.http_minor= 9 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/" - ,.request_url= "/" - ,.num_headers= 0 - ,.headers= {} - ,.body= "" - } - -#define MSEARCH_REQ 20 -, {.name= "m-search request" - ,.type= HTTP_REQUEST - ,.raw= "M-SEARCH * HTTP/1.1\r\n" - "HOST: 239.255.255.250:1900\r\n" - "MAN: \"ssdp:discover\"\r\n" - "ST: \"ssdp:all\"\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_MSEARCH - ,.query_string= "" - ,.fragment= "" - ,.request_path= "*" - ,.request_url= "*" - ,.num_headers= 3 - ,.headers= { { "HOST", "239.255.255.250:1900" } - , { "MAN", "\"ssdp:discover\"" } - , { "ST", "\"ssdp:all\"" } - } - ,.body= "" - } - -#define LINE_FOLDING_IN_HEADER 21 -, {.name= "line folding in header value" - ,.type= HTTP_REQUEST - ,.raw= "GET / HTTP/1.1\r\n" - "Line1: abc\r\n" - "\tdef\r\n" - " ghi\r\n" - "\t\tjkl\r\n" - " mno \r\n" - "\t \tqrs\r\n" - "Line2: \t line2\t\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/" - ,.request_url= "/" - ,.num_headers= 2 - ,.headers= { { "Line1", "abcdefghijklmno qrs" } - , { "Line2", "line2\t" } - } - ,.body= "" - } - - -#define QUERY_TERMINATED_HOST 22 -, {.name= "host terminated by a query string" - ,.type= HTTP_REQUEST - ,.raw= "GET http://hypnotoad.org?hail=all HTTP/1.1\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "hail=all" - ,.fragment= "" - ,.request_path= "" - ,.request_url= "http://hypnotoad.org?hail=all" - ,.host= "hypnotoad.org" - ,.num_headers= 0 - ,.headers= { } - ,.body= "" - } - -#define QUERY_TERMINATED_HOSTPORT 23 -, {.name= "host:port terminated by a query string" - ,.type= HTTP_REQUEST - ,.raw= "GET http://hypnotoad.org:1234?hail=all HTTP/1.1\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "hail=all" - ,.fragment= "" - ,.request_path= "" - ,.request_url= "http://hypnotoad.org:1234?hail=all" - ,.host= "hypnotoad.org" - ,.port= 1234 - ,.num_headers= 0 - ,.headers= { } - ,.body= "" - } - -#define SPACE_TERMINATED_HOSTPORT 24 -, {.name= "host:port terminated by a space" - ,.type= HTTP_REQUEST - ,.raw= "GET http://hypnotoad.org:1234 HTTP/1.1\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "" - ,.fragment= "" - ,.request_path= "" - ,.request_url= "http://hypnotoad.org:1234" - ,.host= "hypnotoad.org" - ,.port= 1234 - ,.num_headers= 0 - ,.headers= { } - ,.body= "" - } - -#define PATCH_REQ 25 -, {.name = "PATCH request" - ,.type= HTTP_REQUEST - ,.raw= "PATCH /file.txt HTTP/1.1\r\n" - "Host: www.example.com\r\n" - "Content-Type: application/example\r\n" - "If-Match: \"e0023aa4e\"\r\n" - "Content-Length: 10\r\n" - "\r\n" - "cccccccccc" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_PATCH - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/file.txt" - ,.request_url= "/file.txt" - ,.num_headers= 4 - ,.headers= { { "Host", "www.example.com" } - , { "Content-Type", "application/example" } - , { "If-Match", "\"e0023aa4e\"" } - , { "Content-Length", "10" } - } - ,.body= "cccccccccc" - } - -#define CONNECT_CAPS_REQUEST 26 -, {.name = "connect caps request" - ,.type= HTTP_REQUEST - ,.raw= "CONNECT HOME0.NETSCAPE.COM:443 HTTP/1.0\r\n" - "User-agent: Mozilla/1.1N\r\n" - "Proxy-authorization: basic aGVsbG86d29ybGQ=\r\n" - "\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 0 - ,.method= HTTP_CONNECT - ,.query_string= "" - ,.fragment= "" - ,.request_path= "" - ,.request_url= "HOME0.NETSCAPE.COM:443" - ,.num_headers= 2 - ,.upgrade="" - ,.headers= { { "User-agent", "Mozilla/1.1N" } - , { "Proxy-authorization", "basic aGVsbG86d29ybGQ=" } - } - ,.body= "" - } - -#if !HTTP_PARSER_STRICT -#define UTF8_PATH_REQ 27 -, {.name= "utf-8 path request" - ,.type= HTTP_REQUEST - ,.raw= "GET /δ¶/δt/pope?q=1#narf HTTP/1.1\r\n" - "Host: github.com\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.query_string= "q=1" - ,.fragment= "narf" - ,.request_path= "/δ¶/δt/pope" - ,.request_url= "/δ¶/δt/pope?q=1#narf" - ,.num_headers= 1 - ,.headers= { {"Host", "github.com" } - } - ,.body= "" - } - -#define HOSTNAME_UNDERSCORE 28 -, {.name = "hostname underscore" - ,.type= HTTP_REQUEST - ,.raw= "CONNECT home_0.netscape.com:443 HTTP/1.0\r\n" - "User-agent: Mozilla/1.1N\r\n" - "Proxy-authorization: basic aGVsbG86d29ybGQ=\r\n" - "\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 0 - ,.method= HTTP_CONNECT - ,.query_string= "" - ,.fragment= "" - ,.request_path= "" - ,.request_url= "home_0.netscape.com:443" - ,.num_headers= 2 - ,.upgrade="" - ,.headers= { { "User-agent", "Mozilla/1.1N" } - , { "Proxy-authorization", "basic aGVsbG86d29ybGQ=" } - } - ,.body= "" - } -#endif /* !HTTP_PARSER_STRICT */ - -/* see https://github.com/ry/http-parser/issues/47 */ -#define EAT_TRAILING_CRLF_NO_CONNECTION_CLOSE 29 -, {.name = "eat CRLF between requests, no \"Connection: close\" header" - ,.raw= "POST / HTTP/1.1\r\n" - "Host: www.example.com\r\n" - "Content-Type: application/x-www-form-urlencoded\r\n" - "Content-Length: 4\r\n" - "\r\n" - "q=42\r\n" /* note the trailing CRLF */ - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_POST - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/" - ,.request_url= "/" - ,.num_headers= 3 - ,.upgrade= 0 - ,.headers= { { "Host", "www.example.com" } - , { "Content-Type", "application/x-www-form-urlencoded" } - , { "Content-Length", "4" } - } - ,.body= "q=42" - } - -/* see https://github.com/ry/http-parser/issues/47 */ -#define EAT_TRAILING_CRLF_WITH_CONNECTION_CLOSE 30 -, {.name = "eat CRLF between requests even if \"Connection: close\" is set" - ,.raw= "POST / HTTP/1.1\r\n" - "Host: www.example.com\r\n" - "Content-Type: application/x-www-form-urlencoded\r\n" - "Content-Length: 4\r\n" - "Connection: close\r\n" - "\r\n" - "q=42\r\n" /* note the trailing CRLF */ - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE /* input buffer isn't empty when on_message_complete is called */ - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_POST - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/" - ,.request_url= "/" - ,.num_headers= 4 - ,.upgrade= 0 - ,.headers= { { "Host", "www.example.com" } - , { "Content-Type", "application/x-www-form-urlencoded" } - , { "Content-Length", "4" } - , { "Connection", "close" } - } - ,.body= "q=42" - } - -#define PURGE_REQ 31 -, {.name = "PURGE request" - ,.type= HTTP_REQUEST - ,.raw= "PURGE /file.txt HTTP/1.1\r\n" - "Host: www.example.com\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_PURGE - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/file.txt" - ,.request_url= "/file.txt" - ,.num_headers= 1 - ,.headers= { { "Host", "www.example.com" } } - ,.body= "" - } - -#define SEARCH_REQ 32 -, {.name = "SEARCH request" - ,.type= HTTP_REQUEST - ,.raw= "SEARCH / HTTP/1.1\r\n" - "Host: www.example.com\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_SEARCH - ,.query_string= "" - ,.fragment= "" - ,.request_path= "/" - ,.request_url= "/" - ,.num_headers= 1 - ,.headers= { { "Host", "www.example.com" } } - ,.body= "" - } - -#define PROXY_WITH_BASIC_AUTH 33 -, {.name= "host:port and basic_auth" - ,.type= HTTP_REQUEST - ,.raw= "GET http://a%12:b!&*$@hypnotoad.org:1234/toto HTTP/1.1\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.method= HTTP_GET - ,.fragment= "" - ,.request_path= "/toto" - ,.request_url= "http://a%12:b!&*$@hypnotoad.org:1234/toto" - ,.host= "hypnotoad.org" - ,.userinfo= "a%12:b!&*$" - ,.port= 1234 - ,.num_headers= 0 - ,.headers= { } - ,.body= "" - } - - -, {.name= NULL } /* sentinel */ -}; - -/* * R E S P O N S E S * */ -const struct message responses[] = -#define GOOGLE_301 0 -{ {.name= "google 301" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 301 Moved Permanently\r\n" - "Location: http://www.google.com/\r\n" - "Content-Type: text/html; charset=UTF-8\r\n" - "Date: Sun, 26 Apr 2009 11:11:49 GMT\r\n" - "Expires: Tue, 26 May 2009 11:11:49 GMT\r\n" - "X-$PrototypeBI-Version: 1.6.0.3\r\n" /* $ char in header field */ - "Cache-Control: public, max-age=2592000\r\n" - "Server: gws\r\n" - "Content-Length: 219 \r\n" - "\r\n" - "\n" - "301 Moved\n" - "

301 Moved

\n" - "The document has moved\n" - "here.\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 301 - ,.num_headers= 8 - ,.headers= - { { "Location", "http://www.google.com/" } - , { "Content-Type", "text/html; charset=UTF-8" } - , { "Date", "Sun, 26 Apr 2009 11:11:49 GMT" } - , { "Expires", "Tue, 26 May 2009 11:11:49 GMT" } - , { "X-$PrototypeBI-Version", "1.6.0.3" } - , { "Cache-Control", "public, max-age=2592000" } - , { "Server", "gws" } - , { "Content-Length", "219 " } - } - ,.body= "\n" - "301 Moved\n" - "

301 Moved

\n" - "The document has moved\n" - "here.\r\n" - "\r\n" - } - -#define NO_CONTENT_LENGTH_RESPONSE 1 -/* The client should wait for the server's EOF. That is, when content-length - * is not specified, and "Connection: close", the end of body is specified - * by the EOF. - * Compare with APACHEBENCH_GET - */ -, {.name= "no content-length response" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 200 OK\r\n" - "Date: Tue, 04 Aug 2009 07:59:32 GMT\r\n" - "Server: Apache\r\n" - "X-Powered-By: Servlet/2.5 JSP/2.1\r\n" - "Content-Type: text/xml; charset=utf-8\r\n" - "Connection: close\r\n" - "\r\n" - "\n" - "\n" - " \n" - " \n" - " SOAP-ENV:Client\n" - " Client Error\n" - " \n" - " \n" - "" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= TRUE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 200 - ,.num_headers= 5 - ,.headers= - { { "Date", "Tue, 04 Aug 2009 07:59:32 GMT" } - , { "Server", "Apache" } - , { "X-Powered-By", "Servlet/2.5 JSP/2.1" } - , { "Content-Type", "text/xml; charset=utf-8" } - , { "Connection", "close" } - } - ,.body= "\n" - "\n" - " \n" - " \n" - " SOAP-ENV:Client\n" - " Client Error\n" - " \n" - " \n" - "" - } - -#define NO_HEADERS_NO_BODY_404 2 -, {.name= "404 no headers no body" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 404 Not Found\r\n\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= TRUE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 404 - ,.num_headers= 0 - ,.headers= {} - ,.body_size= 0 - ,.body= "" - } - -#define NO_REASON_PHRASE 3 -, {.name= "301 no response phrase" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 301\r\n\r\n" - ,.should_keep_alive = FALSE - ,.message_complete_on_eof= TRUE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 301 - ,.num_headers= 0 - ,.headers= {} - ,.body= "" - } - -#define TRAILING_SPACE_ON_CHUNKED_BODY 4 -, {.name="200 trailing space on chunked body" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 200 OK\r\n" - "Content-Type: text/plain\r\n" - "Transfer-Encoding: chunked\r\n" - "\r\n" - "25 \r\n" - "This is the data in the first chunk\r\n" - "\r\n" - "1C\r\n" - "and this is the second one\r\n" - "\r\n" - "0 \r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 200 - ,.num_headers= 2 - ,.headers= - { {"Content-Type", "text/plain" } - , {"Transfer-Encoding", "chunked" } - } - ,.body_size = 37+28 - ,.body = - "This is the data in the first chunk\r\n" - "and this is the second one\r\n" - - } - -#define NO_CARRIAGE_RET 5 -, {.name="no carriage ret" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 200 OK\n" - "Content-Type: text/html; charset=utf-8\n" - "Connection: close\n" - "\n" - "these headers are from http://news.ycombinator.com/" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= TRUE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 200 - ,.num_headers= 2 - ,.headers= - { {"Content-Type", "text/html; charset=utf-8" } - , {"Connection", "close" } - } - ,.body= "these headers are from http://news.ycombinator.com/" - } - -#define PROXY_CONNECTION 6 -, {.name="proxy connection" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 200 OK\r\n" - "Content-Type: text/html; charset=UTF-8\r\n" - "Content-Length: 11\r\n" - "Proxy-Connection: close\r\n" - "Date: Thu, 31 Dec 2009 20:55:48 +0000\r\n" - "\r\n" - "hello world" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 200 - ,.num_headers= 4 - ,.headers= - { {"Content-Type", "text/html; charset=UTF-8" } - , {"Content-Length", "11" } - , {"Proxy-Connection", "close" } - , {"Date", "Thu, 31 Dec 2009 20:55:48 +0000"} - } - ,.body= "hello world" - } - -#define UNDERSTORE_HEADER_KEY 7 - // shown by - // curl -o /dev/null -v "http://ad.doubleclick.net/pfadx/DARTSHELLCONFIGXML;dcmt=text/xml;" -, {.name="underscore header key" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 200 OK\r\n" - "Server: DCLK-AdSvr\r\n" - "Content-Type: text/xml\r\n" - "Content-Length: 0\r\n" - "DCLK_imp: v7;x;114750856;0-0;0;17820020;0/0;21603567/21621457/1;;~okv=;dcmt=text/xml;;~cs=o\r\n\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 200 - ,.num_headers= 4 - ,.headers= - { {"Server", "DCLK-AdSvr" } - , {"Content-Type", "text/xml" } - , {"Content-Length", "0" } - , {"DCLK_imp", "v7;x;114750856;0-0;0;17820020;0/0;21603567/21621457/1;;~okv=;dcmt=text/xml;;~cs=o" } - } - ,.body= "" - } - -#define BONJOUR_MADAME_FR 8 -/* The client should not merge two headers fields when the first one doesn't - * have a value. - */ -, {.name= "bonjourmadame.fr" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.0 301 Moved Permanently\r\n" - "Date: Thu, 03 Jun 2010 09:56:32 GMT\r\n" - "Server: Apache/2.2.3 (Red Hat)\r\n" - "Cache-Control: public\r\n" - "Pragma: \r\n" - "Location: http://www.bonjourmadame.fr/\r\n" - "Vary: Accept-Encoding\r\n" - "Content-Length: 0\r\n" - "Content-Type: text/html; charset=UTF-8\r\n" - "Connection: keep-alive\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 0 - ,.status_code= 301 - ,.num_headers= 9 - ,.headers= - { { "Date", "Thu, 03 Jun 2010 09:56:32 GMT" } - , { "Server", "Apache/2.2.3 (Red Hat)" } - , { "Cache-Control", "public" } - , { "Pragma", "" } - , { "Location", "http://www.bonjourmadame.fr/" } - , { "Vary", "Accept-Encoding" } - , { "Content-Length", "0" } - , { "Content-Type", "text/html; charset=UTF-8" } - , { "Connection", "keep-alive" } - } - ,.body= "" - } - -#define RES_FIELD_UNDERSCORE 9 -/* Should handle spaces in header fields */ -, {.name= "field underscore" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 200 OK\r\n" - "Date: Tue, 28 Sep 2010 01:14:13 GMT\r\n" - "Server: Apache\r\n" - "Cache-Control: no-cache, must-revalidate\r\n" - "Expires: Mon, 26 Jul 1997 05:00:00 GMT\r\n" - ".et-Cookie: PlaxoCS=1274804622353690521; path=/; domain=.plaxo.com\r\n" - "Vary: Accept-Encoding\r\n" - "_eep-Alive: timeout=45\r\n" /* semantic value ignored */ - "_onnection: Keep-Alive\r\n" /* semantic value ignored */ - "Transfer-Encoding: chunked\r\n" - "Content-Type: text/html\r\n" - "Connection: close\r\n" - "\r\n" - "0\r\n\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 200 - ,.num_headers= 11 - ,.headers= - { { "Date", "Tue, 28 Sep 2010 01:14:13 GMT" } - , { "Server", "Apache" } - , { "Cache-Control", "no-cache, must-revalidate" } - , { "Expires", "Mon, 26 Jul 1997 05:00:00 GMT" } - , { ".et-Cookie", "PlaxoCS=1274804622353690521; path=/; domain=.plaxo.com" } - , { "Vary", "Accept-Encoding" } - , { "_eep-Alive", "timeout=45" } - , { "_onnection", "Keep-Alive" } - , { "Transfer-Encoding", "chunked" } - , { "Content-Type", "text/html" } - , { "Connection", "close" } - } - ,.body= "" - } - -#define NON_ASCII_IN_STATUS_LINE 10 -/* Should handle non-ASCII in status line */ -, {.name= "non-ASCII in status line" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 500 Oriëntatieprobleem\r\n" - "Date: Fri, 5 Nov 2010 23:07:12 GMT+2\r\n" - "Content-Length: 0\r\n" - "Connection: close\r\n" - "\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 500 - ,.num_headers= 3 - ,.headers= - { { "Date", "Fri, 5 Nov 2010 23:07:12 GMT+2" } - , { "Content-Length", "0" } - , { "Connection", "close" } - } - ,.body= "" - } - -#define HTTP_VERSION_0_9 11 -/* Should handle HTTP/0.9 */ -, {.name= "http version 0.9" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/0.9 200 OK\r\n" - "\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= TRUE - ,.http_major= 0 - ,.http_minor= 9 - ,.status_code= 200 - ,.num_headers= 0 - ,.headers= - {} - ,.body= "" - } - -#define NO_CONTENT_LENGTH_NO_TRANSFER_ENCODING_RESPONSE 12 -/* The client should wait for the server's EOF. That is, when neither - * content-length nor transfer-encoding is specified, the end of body - * is specified by the EOF. - */ -, {.name= "neither content-length nor transfer-encoding response" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 200 OK\r\n" - "Content-Type: text/plain\r\n" - "\r\n" - "hello world" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= TRUE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 200 - ,.num_headers= 1 - ,.headers= - { { "Content-Type", "text/plain" } - } - ,.body= "hello world" - } - -#define NO_BODY_HTTP10_KA_200 13 -, {.name= "HTTP/1.0 with keep-alive and EOF-terminated 200 status" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.0 200 OK\r\n" - "Connection: keep-alive\r\n" - "\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= TRUE - ,.http_major= 1 - ,.http_minor= 0 - ,.status_code= 200 - ,.num_headers= 1 - ,.headers= - { { "Connection", "keep-alive" } - } - ,.body_size= 0 - ,.body= "" - } - -#define NO_BODY_HTTP10_KA_204 14 -, {.name= "HTTP/1.0 with keep-alive and a 204 status" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.0 204 No content\r\n" - "Connection: keep-alive\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 0 - ,.status_code= 204 - ,.num_headers= 1 - ,.headers= - { { "Connection", "keep-alive" } - } - ,.body_size= 0 - ,.body= "" - } - -#define NO_BODY_HTTP11_KA_200 15 -, {.name= "HTTP/1.1 with an EOF-terminated 200 status" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 200 OK\r\n" - "\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= TRUE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 200 - ,.num_headers= 0 - ,.headers={} - ,.body_size= 0 - ,.body= "" - } - -#define NO_BODY_HTTP11_KA_204 16 -, {.name= "HTTP/1.1 with a 204 status" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 204 No content\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 204 - ,.num_headers= 0 - ,.headers={} - ,.body_size= 0 - ,.body= "" - } - -#define NO_BODY_HTTP11_NOKA_204 17 -, {.name= "HTTP/1.1 with a 204 status and keep-alive disabled" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 204 No content\r\n" - "Connection: close\r\n" - "\r\n" - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 204 - ,.num_headers= 1 - ,.headers= - { { "Connection", "close" } - } - ,.body_size= 0 - ,.body= "" - } - -#define NO_BODY_HTTP11_KA_CHUNKED_200 18 -, {.name= "HTTP/1.1 with chunked endocing and a 200 response" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 200 OK\r\n" - "Transfer-Encoding: chunked\r\n" - "\r\n" - "0\r\n" - "\r\n" - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 200 - ,.num_headers= 1 - ,.headers= - { { "Transfer-Encoding", "chunked" } - } - ,.body_size= 0 - ,.body= "" - } - -#if !HTTP_PARSER_STRICT -#define SPACE_IN_FIELD_RES 19 -/* Should handle spaces in header fields */ -, {.name= "field space" - ,.type= HTTP_RESPONSE - ,.raw= "HTTP/1.1 200 OK\r\n" - "Server: Microsoft-IIS/6.0\r\n" - "X-Powered-By: ASP.NET\r\n" - "en-US Content-Type: text/xml\r\n" /* this is the problem */ - "Content-Type: text/xml\r\n" - "Content-Length: 16\r\n" - "Date: Fri, 23 Jul 2010 18:45:38 GMT\r\n" - "Connection: keep-alive\r\n" - "\r\n" - "hello" /* fake body */ - ,.should_keep_alive= TRUE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 1 - ,.status_code= 200 - ,.num_headers= 7 - ,.headers= - { { "Server", "Microsoft-IIS/6.0" } - , { "X-Powered-By", "ASP.NET" } - , { "en-US Content-Type", "text/xml" } - , { "Content-Type", "text/xml" } - , { "Content-Length", "16" } - , { "Date", "Fri, 23 Jul 2010 18:45:38 GMT" } - , { "Connection", "keep-alive" } - } - ,.body= "hello" - } -#endif /* !HTTP_PARSER_STRICT */ - -, {.name= NULL } /* sentinel */ -}; - -/* strnlen() is a POSIX.2008 addition. Can't rely on it being available so - * define it ourselves. - */ -size_t -strnlen(const char *s, size_t maxlen) -{ - const char *p; - - p = memchr(s, '\0', maxlen); - if (p == NULL) - return maxlen; - - return p - s; -} - -size_t -strlncat(char *dst, size_t len, const char *src, size_t n) -{ - size_t slen; - size_t dlen; - size_t rlen; - size_t ncpy; - - slen = strnlen(src, n); - dlen = strnlen(dst, len); - - if (dlen < len) { - rlen = len - dlen; - ncpy = slen < rlen ? slen : (rlen - 1); - memcpy(dst + dlen, src, ncpy); - dst[dlen + ncpy] = '\0'; - } - - assert(len > slen + dlen); - return slen + dlen; -} - -size_t -strlcat(char *dst, const char *src, size_t len) -{ - return strlncat(dst, len, src, (size_t) -1); -} - -size_t -strlncpy(char *dst, size_t len, const char *src, size_t n) -{ - size_t slen; - size_t ncpy; - - slen = strnlen(src, n); - - if (len > 0) { - ncpy = slen < len ? slen : (len - 1); - memcpy(dst, src, ncpy); - dst[ncpy] = '\0'; - } - - assert(len > slen); - return slen; -} - -size_t -strlcpy(char *dst, const char *src, size_t len) -{ - return strlncpy(dst, len, src, (size_t) -1); -} - -int -request_url_cb (http_parser *p, const char *buf, size_t len) -{ - assert(p == parser); - strlncat(messages[num_messages].request_url, - sizeof(messages[num_messages].request_url), - buf, - len); - return 0; -} - -int -status_complete_cb (http_parser *p) { - assert(p == parser); - p->data++; - return 0; -} - -int -header_field_cb (http_parser *p, const char *buf, size_t len) -{ - assert(p == parser); - struct message *m = &messages[num_messages]; - - if (m->last_header_element != FIELD) - m->num_headers++; - - strlncat(m->headers[m->num_headers-1][0], - sizeof(m->headers[m->num_headers-1][0]), - buf, - len); - - m->last_header_element = FIELD; - - return 0; -} - -int -header_value_cb (http_parser *p, const char *buf, size_t len) -{ - assert(p == parser); - struct message *m = &messages[num_messages]; - - strlncat(m->headers[m->num_headers-1][1], - sizeof(m->headers[m->num_headers-1][1]), - buf, - len); - - m->last_header_element = VALUE; - - return 0; -} - -void -check_body_is_final (const http_parser *p) -{ - if (messages[num_messages].body_is_final) { - fprintf(stderr, "\n\n *** Error http_body_is_final() should return 1 " - "on last on_body callback call " - "but it doesn't! ***\n\n"); - assert(0); - abort(); - } - messages[num_messages].body_is_final = http_body_is_final(p); -} - -int -body_cb (http_parser *p, const char *buf, size_t len) -{ - assert(p == parser); - strlncat(messages[num_messages].body, - sizeof(messages[num_messages].body), - buf, - len); - messages[num_messages].body_size += len; - check_body_is_final(p); - // printf("body_cb: '%s'\n", requests[num_messages].body); - return 0; -} - -int -count_body_cb (http_parser *p, const char *buf, size_t len) -{ - assert(p == parser); - assert(buf); - messages[num_messages].body_size += len; - check_body_is_final(p); - return 0; -} - -int -message_begin_cb (http_parser *p) -{ - assert(p == parser); - messages[num_messages].message_begin_cb_called = TRUE; - return 0; -} - -int -headers_complete_cb (http_parser *p) -{ - assert(p == parser); - messages[num_messages].method = parser->method; - messages[num_messages].status_code = parser->status_code; - messages[num_messages].http_major = parser->http_major; - messages[num_messages].http_minor = parser->http_minor; - messages[num_messages].headers_complete_cb_called = TRUE; - messages[num_messages].should_keep_alive = http_should_keep_alive(parser); - return 0; -} - -int -message_complete_cb (http_parser *p) -{ - assert(p == parser); - if (messages[num_messages].should_keep_alive != http_should_keep_alive(parser)) - { - fprintf(stderr, "\n\n *** Error http_should_keep_alive() should have same " - "value in both on_message_complete and on_headers_complete " - "but it doesn't! ***\n\n"); - assert(0); - abort(); - } - - if (messages[num_messages].body_size && - http_body_is_final(p) && - !messages[num_messages].body_is_final) - { - fprintf(stderr, "\n\n *** Error http_body_is_final() should return 1 " - "on last on_body callback call " - "but it doesn't! ***\n\n"); - assert(0); - abort(); - } - - messages[num_messages].message_complete_cb_called = TRUE; - - messages[num_messages].message_complete_on_eof = currently_parsing_eof; - - num_messages++; - return 0; -} - -/* These dontcall_* callbacks exist so that we can verify that when we're - * paused, no additional callbacks are invoked */ -int -dontcall_message_begin_cb (http_parser *p) -{ - if (p) { } // gcc - fprintf(stderr, "\n\n*** on_message_begin() called on paused parser ***\n\n"); - abort(); -} - -int -dontcall_header_field_cb (http_parser *p, const char *buf, size_t len) -{ - if (p || buf || len) { } // gcc - fprintf(stderr, "\n\n*** on_header_field() called on paused parser ***\n\n"); - abort(); -} - -int -dontcall_header_value_cb (http_parser *p, const char *buf, size_t len) -{ - if (p || buf || len) { } // gcc - fprintf(stderr, "\n\n*** on_header_value() called on paused parser ***\n\n"); - abort(); -} - -int -dontcall_request_url_cb (http_parser *p, const char *buf, size_t len) -{ - if (p || buf || len) { } // gcc - fprintf(stderr, "\n\n*** on_request_url() called on paused parser ***\n\n"); - abort(); -} - -int -dontcall_body_cb (http_parser *p, const char *buf, size_t len) -{ - if (p || buf || len) { } // gcc - fprintf(stderr, "\n\n*** on_body_cb() called on paused parser ***\n\n"); - abort(); -} - -int -dontcall_headers_complete_cb (http_parser *p) -{ - if (p) { } // gcc - fprintf(stderr, "\n\n*** on_headers_complete() called on paused " - "parser ***\n\n"); - abort(); -} - -int -dontcall_message_complete_cb (http_parser *p) -{ - if (p) { } // gcc - fprintf(stderr, "\n\n*** on_message_complete() called on paused " - "parser ***\n\n"); - abort(); -} - -static http_parser_settings settings_dontcall = - {.on_message_begin = dontcall_message_begin_cb - ,.on_header_field = dontcall_header_field_cb - ,.on_header_value = dontcall_header_value_cb - ,.on_url = dontcall_request_url_cb - ,.on_body = dontcall_body_cb - ,.on_headers_complete = dontcall_headers_complete_cb - ,.on_message_complete = dontcall_message_complete_cb - }; - -/* These pause_* callbacks always pause the parser and just invoke the regular - * callback that tracks content. Before returning, we overwrite the parser - * settings to point to the _dontcall variety so that we can verify that - * the pause actually did, you know, pause. */ -int -pause_message_begin_cb (http_parser *p) -{ - http_parser_pause(p, 1); - *current_pause_parser = settings_dontcall; - return message_begin_cb(p); -} - -int -pause_header_field_cb (http_parser *p, const char *buf, size_t len) -{ - http_parser_pause(p, 1); - *current_pause_parser = settings_dontcall; - return header_field_cb(p, buf, len); -} - -int -pause_header_value_cb (http_parser *p, const char *buf, size_t len) -{ - http_parser_pause(p, 1); - *current_pause_parser = settings_dontcall; - return header_value_cb(p, buf, len); -} - -int -pause_request_url_cb (http_parser *p, const char *buf, size_t len) -{ - http_parser_pause(p, 1); - *current_pause_parser = settings_dontcall; - return request_url_cb(p, buf, len); -} - -int -pause_body_cb (http_parser *p, const char *buf, size_t len) -{ - http_parser_pause(p, 1); - *current_pause_parser = settings_dontcall; - return body_cb(p, buf, len); -} - -int -pause_headers_complete_cb (http_parser *p) -{ - http_parser_pause(p, 1); - *current_pause_parser = settings_dontcall; - return headers_complete_cb(p); -} - -int -pause_message_complete_cb (http_parser *p) -{ - http_parser_pause(p, 1); - *current_pause_parser = settings_dontcall; - return message_complete_cb(p); -} - -static http_parser_settings settings_pause = - {.on_message_begin = pause_message_begin_cb - ,.on_header_field = pause_header_field_cb - ,.on_header_value = pause_header_value_cb - ,.on_url = pause_request_url_cb - ,.on_body = pause_body_cb - ,.on_headers_complete = pause_headers_complete_cb - ,.on_message_complete = pause_message_complete_cb - }; - -static http_parser_settings settings = - {.on_message_begin = message_begin_cb - ,.on_header_field = header_field_cb - ,.on_header_value = header_value_cb - ,.on_url = request_url_cb - ,.on_body = body_cb - ,.on_headers_complete = headers_complete_cb - ,.on_message_complete = message_complete_cb - }; - -static http_parser_settings settings_count_body = - {.on_message_begin = message_begin_cb - ,.on_header_field = header_field_cb - ,.on_header_value = header_value_cb - ,.on_url = request_url_cb - ,.on_body = count_body_cb - ,.on_headers_complete = headers_complete_cb - ,.on_message_complete = message_complete_cb - }; - -static http_parser_settings settings_null = - {.on_message_begin = 0 - ,.on_header_field = 0 - ,.on_header_value = 0 - ,.on_url = 0 - ,.on_body = 0 - ,.on_headers_complete = 0 - ,.on_message_complete = 0 - }; - -void -parser_init (enum http_parser_type type) -{ - num_messages = 0; - - assert(parser == NULL); - - parser = malloc(sizeof(http_parser)); - - http_parser_init(parser, type); - - memset(&messages, 0, sizeof messages); - -} - -void -parser_free () -{ - assert(parser); - free(parser); - parser = NULL; -} - -size_t parse (const char *buf, size_t len) -{ - size_t nparsed; - currently_parsing_eof = (len == 0); - nparsed = http_parser_execute(parser, &settings, buf, len); - return nparsed; -} - -size_t parse_count_body (const char *buf, size_t len) -{ - size_t nparsed; - currently_parsing_eof = (len == 0); - nparsed = http_parser_execute(parser, &settings_count_body, buf, len); - return nparsed; -} - -size_t parse_pause (const char *buf, size_t len) -{ - size_t nparsed; - http_parser_settings s = settings_pause; - - currently_parsing_eof = (len == 0); - current_pause_parser = &s; - nparsed = http_parser_execute(parser, current_pause_parser, buf, len); - return nparsed; -} - -static inline int -check_str_eq (const struct message *m, - const char *prop, - const char *expected, - const char *found) { - if ((expected == NULL) != (found == NULL)) { - printf("\n*** Error: %s in '%s' ***\n\n", prop, m->name); - printf("expected %s\n", (expected == NULL) ? "NULL" : expected); - printf(" found %s\n", (found == NULL) ? "NULL" : found); - return 0; - } - if (expected != NULL && 0 != strcmp(expected, found)) { - printf("\n*** Error: %s in '%s' ***\n\n", prop, m->name); - printf("expected '%s'\n", expected); - printf(" found '%s'\n", found); - return 0; - } - return 1; -} - -static inline int -check_num_eq (const struct message *m, - const char *prop, - int expected, - int found) { - if (expected != found) { - printf("\n*** Error: %s in '%s' ***\n\n", prop, m->name); - printf("expected %d\n", expected); - printf(" found %d\n", found); - return 0; - } - return 1; -} - -#define MESSAGE_CHECK_STR_EQ(expected, found, prop) \ - if (!check_str_eq(expected, #prop, expected->prop, found->prop)) return 0 - -#define MESSAGE_CHECK_NUM_EQ(expected, found, prop) \ - if (!check_num_eq(expected, #prop, expected->prop, found->prop)) return 0 - -#define MESSAGE_CHECK_URL_EQ(u, expected, found, prop, fn) \ -do { \ - char ubuf[256]; \ - \ - if ((u)->field_set & (1 << (fn))) { \ - memcpy(ubuf, (found)->request_url + (u)->field_data[(fn)].off, \ - (u)->field_data[(fn)].len); \ - ubuf[(u)->field_data[(fn)].len] = '\0'; \ - } else { \ - ubuf[0] = '\0'; \ - } \ - \ - check_str_eq(expected, #prop, expected->prop, ubuf); \ -} while(0) - -int -message_eq (int index, const struct message *expected) -{ - int i; - struct message *m = &messages[index]; - - MESSAGE_CHECK_NUM_EQ(expected, m, http_major); - MESSAGE_CHECK_NUM_EQ(expected, m, http_minor); - - if (expected->type == HTTP_REQUEST) { - MESSAGE_CHECK_NUM_EQ(expected, m, method); - } else { - MESSAGE_CHECK_NUM_EQ(expected, m, status_code); - } - - MESSAGE_CHECK_NUM_EQ(expected, m, should_keep_alive); - MESSAGE_CHECK_NUM_EQ(expected, m, message_complete_on_eof); - - assert(m->message_begin_cb_called); - assert(m->headers_complete_cb_called); - assert(m->message_complete_cb_called); - - - MESSAGE_CHECK_STR_EQ(expected, m, request_url); - - /* Check URL components; we can't do this w/ CONNECT since it doesn't - * send us a well-formed URL. - */ - if (*m->request_url && m->method != HTTP_CONNECT) { - struct http_parser_url u; - - if (http_parser_parse_url(m->request_url, strlen(m->request_url), 0, &u)) { - fprintf(stderr, "\n\n*** failed to parse URL %s ***\n\n", - m->request_url); - abort(); - } - - if (expected->host) { - MESSAGE_CHECK_URL_EQ(&u, expected, m, host, UF_HOST); - } - - if (expected->userinfo) { - MESSAGE_CHECK_URL_EQ(&u, expected, m, userinfo, UF_USERINFO); - } - - m->port = (u.field_set & (1 << UF_PORT)) ? - u.port : 0; - - MESSAGE_CHECK_URL_EQ(&u, expected, m, query_string, UF_QUERY); - MESSAGE_CHECK_URL_EQ(&u, expected, m, fragment, UF_FRAGMENT); - MESSAGE_CHECK_URL_EQ(&u, expected, m, request_path, UF_PATH); - MESSAGE_CHECK_NUM_EQ(expected, m, port); - } - - if (expected->body_size) { - MESSAGE_CHECK_NUM_EQ(expected, m, body_size); - } else { - MESSAGE_CHECK_STR_EQ(expected, m, body); - } - - MESSAGE_CHECK_NUM_EQ(expected, m, num_headers); - - int r; - for (i = 0; i < m->num_headers; i++) { - r = check_str_eq(expected, "header field", expected->headers[i][0], m->headers[i][0]); - if (!r) return 0; - r = check_str_eq(expected, "header value", expected->headers[i][1], m->headers[i][1]); - if (!r) return 0; - } - - MESSAGE_CHECK_STR_EQ(expected, m, upgrade); - - return 1; -} - -/* Given a sequence of varargs messages, return the number of them that the - * parser should successfully parse, taking into account that upgraded - * messages prevent all subsequent messages from being parsed. - */ -size_t -count_parsed_messages(const size_t nmsgs, ...) { - size_t i; - va_list ap; - - va_start(ap, nmsgs); - - for (i = 0; i < nmsgs; i++) { - struct message *m = va_arg(ap, struct message *); - - if (m->upgrade) { - va_end(ap); - return i + 1; - } - } - - va_end(ap); - return nmsgs; -} - -/* Given a sequence of bytes and the number of these that we were able to - * parse, verify that upgrade bodies are correct. - */ -void -upgrade_message_fix(char *body, const size_t nread, const size_t nmsgs, ...) { - va_list ap; - size_t i; - size_t off = 0; - - va_start(ap, nmsgs); - - for (i = 0; i < nmsgs; i++) { - struct message *m = va_arg(ap, struct message *); - - off += strlen(m->raw); - - if (m->upgrade) { - off -= strlen(m->upgrade); - - /* Check the portion of the response after its specified upgrade */ - if (!check_str_eq(m, "upgrade", body + off, body + nread)) { - abort(); - } - - /* Fix up the response so that message_eq() will verify the beginning - * of the upgrade */ - *(body + nread + strlen(m->upgrade)) = '\0'; - messages[num_messages -1 ].upgrade = body + nread; - - va_end(ap); - return; - } - } - - va_end(ap); - printf("\n\n*** Error: expected a message with upgrade ***\n"); - - abort(); -} - -static void -print_error (const char *raw, size_t error_location) -{ - fprintf(stderr, "\n*** %s ***\n\n", - http_errno_description(HTTP_PARSER_ERRNO(parser))); - - int this_line = 0, char_len = 0; - size_t i, j, len = strlen(raw), error_location_line = 0; - for (i = 0; i < len; i++) { - if (i == error_location) this_line = 1; - switch (raw[i]) { - case '\r': - char_len = 2; - fprintf(stderr, "\\r"); - break; - - case '\n': - char_len = 2; - fprintf(stderr, "\\n\n"); - - if (this_line) goto print; - - error_location_line = 0; - continue; - - default: - char_len = 1; - fputc(raw[i], stderr); - break; - } - if (!this_line) error_location_line += char_len; - } - - fprintf(stderr, "[eof]\n"); - - print: - for (j = 0; j < error_location_line; j++) { - fputc(' ', stderr); - } - fprintf(stderr, "^\n\nerror location: %u\n", (unsigned int)error_location); -} - -void -test_preserve_data (void) -{ - char my_data[] = "application-specific data"; - http_parser parser; - parser.data = my_data; - http_parser_init(&parser, HTTP_REQUEST); - if (parser.data != my_data) { - printf("\n*** parser.data not preserved accross http_parser_init ***\n\n"); - abort(); - } -} - -struct url_test { - const char *name; - const char *url; - int is_connect; - struct http_parser_url u; - int rv; -}; - -const struct url_test url_tests[] = -{ {.name="proxy request" - ,.url="http://hostname/" - ,.is_connect=0 - ,.u= - {.field_set=(1 << UF_SCHEMA) | (1 << UF_HOST) | (1 << UF_PATH) - ,.port=0 - ,.field_data= - {{ 0, 4 } /* UF_SCHEMA */ - ,{ 7, 8 } /* UF_HOST */ - ,{ 0, 0 } /* UF_PORT */ - ,{ 15, 1 } /* UF_PATH */ - ,{ 0, 0 } /* UF_QUERY */ - ,{ 0, 0 } /* UF_FRAGMENT */ - ,{ 0, 0 } /* UF_USERINFO */ - } - } - ,.rv=0 - } - -, {.name="proxy request with port" - ,.url="http://hostname:444/" - ,.is_connect=0 - ,.u= - {.field_set=(1 << UF_SCHEMA) | (1 << UF_HOST) | (1 << UF_PORT) | (1 << UF_PATH) - ,.port=444 - ,.field_data= - {{ 0, 4 } /* UF_SCHEMA */ - ,{ 7, 8 } /* UF_HOST */ - ,{ 16, 3 } /* UF_PORT */ - ,{ 19, 1 } /* UF_PATH */ - ,{ 0, 0 } /* UF_QUERY */ - ,{ 0, 0 } /* UF_FRAGMENT */ - ,{ 0, 0 } /* UF_USERINFO */ - } - } - ,.rv=0 - } - -, {.name="CONNECT request" - ,.url="hostname:443" - ,.is_connect=1 - ,.u= - {.field_set=(1 << UF_HOST) | (1 << UF_PORT) - ,.port=443 - ,.field_data= - {{ 0, 0 } /* UF_SCHEMA */ - ,{ 0, 8 } /* UF_HOST */ - ,{ 9, 3 } /* UF_PORT */ - ,{ 0, 0 } /* UF_PATH */ - ,{ 0, 0 } /* UF_QUERY */ - ,{ 0, 0 } /* UF_FRAGMENT */ - ,{ 0, 0 } /* UF_USERINFO */ - } - } - ,.rv=0 - } - -, {.name="CONNECT request but not connect" - ,.url="hostname:443" - ,.is_connect=0 - ,.rv=1 - } - -, {.name="proxy ipv6 request" - ,.url="http://[1:2::3:4]/" - ,.is_connect=0 - ,.u= - {.field_set=(1 << UF_SCHEMA) | (1 << UF_HOST) | (1 << UF_PATH) - ,.port=0 - ,.field_data= - {{ 0, 4 } /* UF_SCHEMA */ - ,{ 8, 8 } /* UF_HOST */ - ,{ 0, 0 } /* UF_PORT */ - ,{ 17, 1 } /* UF_PATH */ - ,{ 0, 0 } /* UF_QUERY */ - ,{ 0, 0 } /* UF_FRAGMENT */ - ,{ 0, 0 } /* UF_USERINFO */ - } - } - ,.rv=0 - } - -, {.name="proxy ipv6 request with port" - ,.url="http://[1:2::3:4]:67/" - ,.is_connect=0 - ,.u= - {.field_set=(1 << UF_SCHEMA) | (1 << UF_HOST) | (1 << UF_PORT) | (1 << UF_PATH) - ,.port=67 - ,.field_data= - {{ 0, 4 } /* UF_SCHEMA */ - ,{ 8, 8 } /* UF_HOST */ - ,{ 18, 2 } /* UF_PORT */ - ,{ 20, 1 } /* UF_PATH */ - ,{ 0, 0 } /* UF_QUERY */ - ,{ 0, 0 } /* UF_FRAGMENT */ - ,{ 0, 0 } /* UF_USERINFO */ - } - } - ,.rv=0 - } - -, {.name="CONNECT ipv6 address" - ,.url="[1:2::3:4]:443" - ,.is_connect=1 - ,.u= - {.field_set=(1 << UF_HOST) | (1 << UF_PORT) - ,.port=443 - ,.field_data= - {{ 0, 0 } /* UF_SCHEMA */ - ,{ 1, 8 } /* UF_HOST */ - ,{ 11, 3 } /* UF_PORT */ - ,{ 0, 0 } /* UF_PATH */ - ,{ 0, 0 } /* UF_QUERY */ - ,{ 0, 0 } /* UF_FRAGMENT */ - ,{ 0, 0 } /* UF_USERINFO */ - } - } - ,.rv=0 - } - -, {.name="ipv4 in ipv6 address" - ,.url="http://[2001:0000:0000:0000:0000:0000:1.9.1.1]/" - ,.is_connect=0 - ,.u= - {.field_set=(1 << UF_SCHEMA) | (1 << UF_HOST) | (1 << UF_PATH) - ,.port=0 - ,.field_data= - {{ 0, 4 } /* UF_SCHEMA */ - ,{ 8, 37 } /* UF_HOST */ - ,{ 0, 0 } /* UF_PORT */ - ,{ 46, 1 } /* UF_PATH */ - ,{ 0, 0 } /* UF_QUERY */ - ,{ 0, 0 } /* UF_FRAGMENT */ - ,{ 0, 0 } /* UF_USERINFO */ - } - } - ,.rv=0 - } - -, {.name="extra ? in query string" - ,.url="http://a.tbcdn.cn/p/fp/2010c/??fp-header-min.css,fp-base-min.css," - "fp-channel-min.css,fp-product-min.css,fp-mall-min.css,fp-category-min.css," - "fp-sub-min.css,fp-gdp4p-min.css,fp-css3-min.css,fp-misc-min.css?t=20101022.css" - ,.is_connect=0 - ,.u= - {.field_set=(1<field_set, u->port); - for (i = 0; i < UF_MAX; i++) { - if ((u->field_set & (1 << i)) == 0) { - printf("\tfield_data[%u]: unset\n", i); - continue; - } - - printf("\tfield_data[%u]: off: %u len: %u part: \"%.*s\n\"", - i, - u->field_data[i].off, - u->field_data[i].len, - u->field_data[i].len, - url + u->field_data[i].off); - } -} - -void -test_parse_url (void) -{ - struct http_parser_url u; - const struct url_test *test; - unsigned int i; - int rv; - - for (i = 0; i < (sizeof(url_tests) / sizeof(url_tests[0])); i++) { - test = &url_tests[i]; - memset(&u, 0, sizeof(u)); - - rv = http_parser_parse_url(test->url, - strlen(test->url), - test->is_connect, - &u); - - if (test->rv == 0) { - if (rv != 0) { - printf("\n*** http_parser_parse_url(\"%s\") \"%s\" test failed, " - "unexpected rv %d ***\n\n", test->url, test->name, rv); - abort(); - } - - if (memcmp(&u, &test->u, sizeof(u)) != 0) { - printf("\n*** http_parser_parse_url(\"%s\") \"%s\" failed ***\n", - test->url, test->name); - - printf("target http_parser_url:\n"); - dump_url(test->url, &test->u); - printf("result http_parser_url:\n"); - dump_url(test->url, &u); - - abort(); - } - } else { - /* test->rv != 0 */ - if (rv == 0) { - printf("\n*** http_parser_parse_url(\"%s\") \"%s\" test failed, " - "unexpected rv %d ***\n\n", test->url, test->name, rv); - abort(); - } - } - } -} - -void -test_method_str (void) -{ - assert(0 == strcmp("GET", http_method_str(HTTP_GET))); - assert(0 == strcmp("", http_method_str(1337))); -} - -void -test_message (const struct message *message) -{ - size_t raw_len = strlen(message->raw); - size_t msg1len; - for (msg1len = 0; msg1len < raw_len; msg1len++) { - parser_init(message->type); - - size_t read; - const char *msg1 = message->raw; - const char *msg2 = msg1 + msg1len; - size_t msg2len = raw_len - msg1len; - - if (msg1len) { - read = parse(msg1, msg1len); - - if (message->upgrade && parser->upgrade) { - messages[num_messages - 1].upgrade = msg1 + read; - goto test; - } - - if (read != msg1len) { - print_error(msg1, read); - abort(); - } - } - - - read = parse(msg2, msg2len); - - if (message->upgrade && parser->upgrade) { - messages[num_messages - 1].upgrade = msg2 + read; - goto test; - } - - if (read != msg2len) { - print_error(msg2, read); - abort(); - } - - read = parse(NULL, 0); - - if (read != 0) { - print_error(message->raw, read); - abort(); - } - - test: - - if (num_messages != 1) { - printf("\n*** num_messages != 1 after testing '%s' ***\n\n", message->name); - abort(); - } - - if(!message_eq(0, message)) abort(); - - parser_free(); - } -} - -void -test_message_count_body (const struct message *message) -{ - parser_init(message->type); - - size_t read; - size_t l = strlen(message->raw); - size_t i, toread; - size_t chunk = 4024; - - for (i = 0; i < l; i+= chunk) { - toread = MIN(l-i, chunk); - read = parse_count_body(message->raw + i, toread); - if (read != toread) { - print_error(message->raw, read); - abort(); - } - } - - - read = parse_count_body(NULL, 0); - if (read != 0) { - print_error(message->raw, read); - abort(); - } - - if (num_messages != 1) { - printf("\n*** num_messages != 1 after testing '%s' ***\n\n", message->name); - abort(); - } - - if(!message_eq(0, message)) abort(); - - parser_free(); -} - -void -test_simple (const char *buf, enum http_errno err_expected) -{ - parser_init(HTTP_REQUEST); - - size_t parsed; - int pass; - enum http_errno err; - - parsed = parse(buf, strlen(buf)); - pass = (parsed == strlen(buf)); - err = HTTP_PARSER_ERRNO(parser); - parsed = parse(NULL, 0); - pass &= (parsed == 0); - - parser_free(); - - /* In strict mode, allow us to pass with an unexpected HPE_STRICT as - * long as the caller isn't expecting success. - */ -#if HTTP_PARSER_STRICT - if (err_expected != err && err_expected != HPE_OK && err != HPE_STRICT) { -#else - if (err_expected != err) { -#endif - fprintf(stderr, "\n*** test_simple expected %s, but saw %s ***\n\n%s\n", - http_errno_name(err_expected), http_errno_name(err), buf); - abort(); - } -} - -void -test_header_overflow_error (int req) -{ - http_parser parser; - http_parser_init(&parser, req ? HTTP_REQUEST : HTTP_RESPONSE); - size_t parsed; - const char *buf; - buf = req ? "GET / HTTP/1.1\r\n" : "HTTP/1.0 200 OK\r\n"; - parsed = http_parser_execute(&parser, &settings_null, buf, strlen(buf)); - assert(parsed == strlen(buf)); - - buf = "header-key: header-value\r\n"; - size_t buflen = strlen(buf); - - int i; - for (i = 0; i < 10000; i++) { - parsed = http_parser_execute(&parser, &settings_null, buf, buflen); - if (parsed != buflen) { - //fprintf(stderr, "error found on iter %d\n", i); - assert(HTTP_PARSER_ERRNO(&parser) == HPE_HEADER_OVERFLOW); - return; - } - } - - fprintf(stderr, "\n*** Error expected but none in header overflow test ***\n"); - abort(); -} - -static void -test_content_length_overflow (const char *buf, size_t buflen, int expect_ok) -{ - http_parser parser; - http_parser_init(&parser, HTTP_RESPONSE); - http_parser_execute(&parser, &settings_null, buf, buflen); - - if (expect_ok) - assert(HTTP_PARSER_ERRNO(&parser) == HPE_OK); - else - assert(HTTP_PARSER_ERRNO(&parser) == HPE_INVALID_CONTENT_LENGTH); -} - -void -test_header_content_length_overflow_error (void) -{ -#define X(size) \ - "HTTP/1.1 200 OK\r\n" \ - "Content-Length: " #size "\r\n" \ - "\r\n" - const char a[] = X(18446744073709551614); /* 2^64-2 */ - const char b[] = X(18446744073709551615); /* 2^64-1 */ - const char c[] = X(18446744073709551616); /* 2^64 */ -#undef X - test_content_length_overflow(a, sizeof(a) - 1, 1); /* expect ok */ - test_content_length_overflow(b, sizeof(b) - 1, 0); /* expect failure */ - test_content_length_overflow(c, sizeof(c) - 1, 0); /* expect failure */ -} - -void -test_chunk_content_length_overflow_error (void) -{ -#define X(size) \ - "HTTP/1.1 200 OK\r\n" \ - "Transfer-Encoding: chunked\r\n" \ - "\r\n" \ - #size "\r\n" \ - "..." - const char a[] = X(FFFFFFFFFFFFFFFE); /* 2^64-2 */ - const char b[] = X(FFFFFFFFFFFFFFFF); /* 2^64-1 */ - const char c[] = X(10000000000000000); /* 2^64 */ -#undef X - test_content_length_overflow(a, sizeof(a) - 1, 1); /* expect ok */ - test_content_length_overflow(b, sizeof(b) - 1, 0); /* expect failure */ - test_content_length_overflow(c, sizeof(c) - 1, 0); /* expect failure */ -} - -void -test_no_overflow_long_body (int req, size_t length) -{ - http_parser parser; - http_parser_init(&parser, req ? HTTP_REQUEST : HTTP_RESPONSE); - size_t parsed; - size_t i; - char buf1[3000]; - size_t buf1len = sprintf(buf1, "%s\r\nConnection: Keep-Alive\r\nContent-Length: %lu\r\n\r\n", - req ? "POST / HTTP/1.0" : "HTTP/1.0 200 OK", (unsigned long)length); - parsed = http_parser_execute(&parser, &settings_null, buf1, buf1len); - if (parsed != buf1len) - goto err; - - for (i = 0; i < length; i++) { - char foo = 'a'; - parsed = http_parser_execute(&parser, &settings_null, &foo, 1); - if (parsed != 1) - goto err; - } - - parsed = http_parser_execute(&parser, &settings_null, buf1, buf1len); - if (parsed != buf1len) goto err; - return; - - err: - fprintf(stderr, - "\n*** error in test_no_overflow_long_body %s of length %lu ***\n", - req ? "REQUEST" : "RESPONSE", - (unsigned long)length); - abort(); -} - -void -test_multiple3 (const struct message *r1, const struct message *r2, const struct message *r3) -{ - int message_count = count_parsed_messages(3, r1, r2, r3); - - char total[ strlen(r1->raw) - + strlen(r2->raw) - + strlen(r3->raw) - + 1 - ]; - total[0] = '\0'; - - strcat(total, r1->raw); - strcat(total, r2->raw); - strcat(total, r3->raw); - - parser_init(r1->type); - - size_t read; - - read = parse(total, strlen(total)); - - if (parser->upgrade) { - upgrade_message_fix(total, read, 3, r1, r2, r3); - goto test; - } - - if (read != strlen(total)) { - print_error(total, read); - abort(); - } - - read = parse(NULL, 0); - - if (read != 0) { - print_error(total, read); - abort(); - } - -test: - - if (message_count != num_messages) { - fprintf(stderr, "\n\n*** Parser didn't see 3 messages only %d *** \n", num_messages); - abort(); - } - - if (!message_eq(0, r1)) abort(); - if (message_count > 1 && !message_eq(1, r2)) abort(); - if (message_count > 2 && !message_eq(2, r3)) abort(); - - parser_free(); -} - -/* SCAN through every possible breaking to make sure the - * parser can handle getting the content in any chunks that - * might come from the socket - */ -void -test_scan (const struct message *r1, const struct message *r2, const struct message *r3) -{ - char total[80*1024] = "\0"; - char buf1[80*1024] = "\0"; - char buf2[80*1024] = "\0"; - char buf3[80*1024] = "\0"; - - strcat(total, r1->raw); - strcat(total, r2->raw); - strcat(total, r3->raw); - - size_t read; - - int total_len = strlen(total); - - int total_ops = 2 * (total_len - 1) * (total_len - 2) / 2; - int ops = 0 ; - - size_t buf1_len, buf2_len, buf3_len; - int message_count = count_parsed_messages(3, r1, r2, r3); - - int i,j,type_both; - for (type_both = 0; type_both < 2; type_both ++ ) { - for (j = 2; j < total_len; j ++ ) { - for (i = 1; i < j; i ++ ) { - - if (ops % 1000 == 0) { - printf("\b\b\b\b%3.0f%%", 100 * (float)ops /(float)total_ops); - fflush(stdout); - } - ops += 1; - - parser_init(type_both ? HTTP_BOTH : r1->type); - - buf1_len = i; - strlncpy(buf1, sizeof(buf1), total, buf1_len); - buf1[buf1_len] = 0; - - buf2_len = j - i; - strlncpy(buf2, sizeof(buf1), total+i, buf2_len); - buf2[buf2_len] = 0; - - buf3_len = total_len - j; - strlncpy(buf3, sizeof(buf1), total+j, buf3_len); - buf3[buf3_len] = 0; - - read = parse(buf1, buf1_len); - - if (parser->upgrade) goto test; - - if (read != buf1_len) { - print_error(buf1, read); - goto error; - } - - read += parse(buf2, buf2_len); - - if (parser->upgrade) goto test; - - if (read != buf1_len + buf2_len) { - print_error(buf2, read); - goto error; - } - - read += parse(buf3, buf3_len); - - if (parser->upgrade) goto test; - - if (read != buf1_len + buf2_len + buf3_len) { - print_error(buf3, read); - goto error; - } - - parse(NULL, 0); - -test: - if (parser->upgrade) { - upgrade_message_fix(total, read, 3, r1, r2, r3); - } - - if (message_count != num_messages) { - fprintf(stderr, "\n\nParser didn't see %d messages only %d\n", - message_count, num_messages); - goto error; - } - - if (!message_eq(0, r1)) { - fprintf(stderr, "\n\nError matching messages[0] in test_scan.\n"); - goto error; - } - - if (message_count > 1 && !message_eq(1, r2)) { - fprintf(stderr, "\n\nError matching messages[1] in test_scan.\n"); - goto error; - } - - if (message_count > 2 && !message_eq(2, r3)) { - fprintf(stderr, "\n\nError matching messages[2] in test_scan.\n"); - goto error; - } - - parser_free(); - } - } - } - puts("\b\b\b\b100%"); - return; - - error: - fprintf(stderr, "i=%d j=%d\n", i, j); - fprintf(stderr, "buf1 (%u) %s\n\n", (unsigned int)buf1_len, buf1); - fprintf(stderr, "buf2 (%u) %s\n\n", (unsigned int)buf2_len , buf2); - fprintf(stderr, "buf3 (%u) %s\n", (unsigned int)buf3_len, buf3); - abort(); -} - -// user required to free the result -// string terminated by \0 -char * -create_large_chunked_message (int body_size_in_kb, const char* headers) -{ - int i; - size_t wrote = 0; - size_t headers_len = strlen(headers); - size_t bufsize = headers_len + (5+1024+2)*body_size_in_kb + 6; - char * buf = malloc(bufsize); - - memcpy(buf, headers, headers_len); - wrote += headers_len; - - for (i = 0; i < body_size_in_kb; i++) { - // write 1kb chunk into the body. - memcpy(buf + wrote, "400\r\n", 5); - wrote += 5; - memset(buf + wrote, 'C', 1024); - wrote += 1024; - strcpy(buf + wrote, "\r\n"); - wrote += 2; - } - - memcpy(buf + wrote, "0\r\n\r\n", 6); - wrote += 6; - assert(wrote == bufsize); - - return buf; -} - -void -test_status_complete (void) -{ - parser_init(HTTP_RESPONSE); - parser->data = 0; - http_parser_settings settings = settings_null; - settings.on_status_complete = status_complete_cb; - - char *response = "don't mind me, just a simple response"; - http_parser_execute(parser, &settings, response, strlen(response)); - assert(parser->data == (void*)0); // the status_complete callback was never called - assert(parser->http_errno == HPE_INVALID_CONSTANT); // the errno for an invalid status line -} - -/* Verify that we can pause parsing at any of the bytes in the - * message and still get the result that we're expecting. */ -void -test_message_pause (const struct message *msg) -{ - char *buf = (char*) msg->raw; - size_t buflen = strlen(msg->raw); - size_t nread; - - parser_init(msg->type); - - do { - nread = parse_pause(buf, buflen); - - // We can only set the upgrade buffer once we've gotten our message - // completion callback. - if (messages[0].message_complete_cb_called && - msg->upgrade && - parser->upgrade) { - messages[0].upgrade = buf + nread; - goto test; - } - - if (nread < buflen) { - - // Not much do to if we failed a strict-mode check - if (HTTP_PARSER_ERRNO(parser) == HPE_STRICT) { - parser_free(); - return; - } - - assert (HTTP_PARSER_ERRNO(parser) == HPE_PAUSED); - } - - buf += nread; - buflen -= nread; - http_parser_pause(parser, 0); - } while (buflen > 0); - - nread = parse_pause(NULL, 0); - assert (nread == 0); - -test: - if (num_messages != 1) { - printf("\n*** num_messages != 1 after testing '%s' ***\n\n", msg->name); - abort(); - } - - if(!message_eq(0, msg)) abort(); - - parser_free(); -} - -int -main (void) -{ - parser = NULL; - int i, j, k; - int request_count; - int response_count; - - printf("sizeof(http_parser) = %u\n", (unsigned int)sizeof(http_parser)); - - for (request_count = 0; requests[request_count].name; request_count++); - for (response_count = 0; responses[response_count].name; response_count++); - - //// API - test_preserve_data(); - test_parse_url(); - test_method_str(); - - //// OVERFLOW CONDITIONS - - test_header_overflow_error(HTTP_REQUEST); - test_no_overflow_long_body(HTTP_REQUEST, 1000); - test_no_overflow_long_body(HTTP_REQUEST, 100000); - - test_header_overflow_error(HTTP_RESPONSE); - test_no_overflow_long_body(HTTP_RESPONSE, 1000); - test_no_overflow_long_body(HTTP_RESPONSE, 100000); - - test_header_content_length_overflow_error(); - test_chunk_content_length_overflow_error(); - - //// RESPONSES - - for (i = 0; i < response_count; i++) { - test_message(&responses[i]); - } - - for (i = 0; i < response_count; i++) { - test_message_pause(&responses[i]); - } - - for (i = 0; i < response_count; i++) { - if (!responses[i].should_keep_alive) continue; - for (j = 0; j < response_count; j++) { - if (!responses[j].should_keep_alive) continue; - for (k = 0; k < response_count; k++) { - test_multiple3(&responses[i], &responses[j], &responses[k]); - } - } - } - - test_message_count_body(&responses[NO_HEADERS_NO_BODY_404]); - test_message_count_body(&responses[TRAILING_SPACE_ON_CHUNKED_BODY]); - - // test very large chunked response - { - char * msg = create_large_chunked_message(31337, - "HTTP/1.0 200 OK\r\n" - "Transfer-Encoding: chunked\r\n" - "Content-Type: text/plain\r\n" - "\r\n"); - struct message large_chunked = - {.name= "large chunked" - ,.type= HTTP_RESPONSE - ,.raw= msg - ,.should_keep_alive= FALSE - ,.message_complete_on_eof= FALSE - ,.http_major= 1 - ,.http_minor= 0 - ,.status_code= 200 - ,.num_headers= 2 - ,.headers= - { { "Transfer-Encoding", "chunked" } - , { "Content-Type", "text/plain" } - } - ,.body_size= 31337*1024 - }; - test_message_count_body(&large_chunked); - free(msg); - } - - - - printf("response scan 1/2 "); - test_scan( &responses[TRAILING_SPACE_ON_CHUNKED_BODY] - , &responses[NO_BODY_HTTP10_KA_204] - , &responses[NO_REASON_PHRASE] - ); - - printf("response scan 2/2 "); - test_scan( &responses[BONJOUR_MADAME_FR] - , &responses[UNDERSTORE_HEADER_KEY] - , &responses[NO_CARRIAGE_RET] - ); - - puts("responses okay"); - - - /// REQUESTS - - test_simple("hello world", HPE_INVALID_METHOD); - test_simple("GET / HTP/1.1\r\n\r\n", HPE_INVALID_VERSION); - - - test_simple("ASDF / HTTP/1.1\r\n\r\n", HPE_INVALID_METHOD); - test_simple("PROPPATCHA / HTTP/1.1\r\n\r\n", HPE_INVALID_METHOD); - test_simple("GETA / HTTP/1.1\r\n\r\n", HPE_INVALID_METHOD); - - // Well-formed but incomplete - test_simple("GET / HTTP/1.1\r\n" - "Content-Type: text/plain\r\n" - "Content-Length: 6\r\n" - "\r\n" - "fooba", - HPE_OK); - - static const char *all_methods[] = { - "DELETE", - "GET", - "HEAD", - "POST", - "PUT", - //"CONNECT", //CONNECT can't be tested like other methods, it's a tunnel - "OPTIONS", - "TRACE", - "COPY", - "LOCK", - "MKCOL", - "MOVE", - "PROPFIND", - "PROPPATCH", - "UNLOCK", - "REPORT", - "MKACTIVITY", - "CHECKOUT", - "MERGE", - "M-SEARCH", - "NOTIFY", - "SUBSCRIBE", - "UNSUBSCRIBE", - "PATCH", - 0 }; - const char **this_method; - for (this_method = all_methods; *this_method; this_method++) { - char buf[200]; - sprintf(buf, "%s / HTTP/1.1\r\n\r\n", *this_method); - test_simple(buf, HPE_OK); - } - - static const char *bad_methods[] = { - "C******", - "M****", - 0 }; - for (this_method = bad_methods; *this_method; this_method++) { - char buf[200]; - sprintf(buf, "%s / HTTP/1.1\r\n\r\n", *this_method); - test_simple(buf, HPE_UNKNOWN); - } - - const char *dumbfuck2 = - "GET / HTTP/1.1\r\n" - "X-SSL-Bullshit: -----BEGIN CERTIFICATE-----\r\n" - "\tMIIFbTCCBFWgAwIBAgICH4cwDQYJKoZIhvcNAQEFBQAwcDELMAkGA1UEBhMCVUsx\r\n" - "\tETAPBgNVBAoTCGVTY2llbmNlMRIwEAYDVQQLEwlBdXRob3JpdHkxCzAJBgNVBAMT\r\n" - "\tAkNBMS0wKwYJKoZIhvcNAQkBFh5jYS1vcGVyYXRvckBncmlkLXN1cHBvcnQuYWMu\r\n" - "\tdWswHhcNMDYwNzI3MTQxMzI4WhcNMDcwNzI3MTQxMzI4WjBbMQswCQYDVQQGEwJV\r\n" - "\tSzERMA8GA1UEChMIZVNjaWVuY2UxEzARBgNVBAsTCk1hbmNoZXN0ZXIxCzAJBgNV\r\n" - "\tBAcTmrsogriqMWLAk1DMRcwFQYDVQQDEw5taWNoYWVsIHBhcmQYJKoZIhvcNAQEB\r\n" - "\tBQADggEPADCCAQoCggEBANPEQBgl1IaKdSS1TbhF3hEXSl72G9J+WC/1R64fAcEF\r\n" - "\tW51rEyFYiIeZGx/BVzwXbeBoNUK41OK65sxGuflMo5gLflbwJtHBRIEKAfVVp3YR\r\n" - "\tgW7cMA/s/XKgL1GEC7rQw8lIZT8RApukCGqOVHSi/F1SiFlPDxuDfmdiNzL31+sL\r\n" - "\t0iwHDdNkGjy5pyBSB8Y79dsSJtCW/iaLB0/n8Sj7HgvvZJ7x0fr+RQjYOUUfrePP\r\n" - "\tu2MSpFyf+9BbC/aXgaZuiCvSR+8Snv3xApQY+fULK/xY8h8Ua51iXoQ5jrgu2SqR\r\n" - "\twgA7BUi3G8LFzMBl8FRCDYGUDy7M6QaHXx1ZWIPWNKsCAwEAAaOCAiQwggIgMAwG\r\n" - "\tA1UdEwEB/wQCMAAwEQYJYIZIAYb4QgHTTPAQDAgWgMA4GA1UdDwEB/wQEAwID6DAs\r\n" - "\tBglghkgBhvhCAQ0EHxYdVUsgZS1TY2llbmNlIFVzZXIgQ2VydGlmaWNhdGUwHQYD\r\n" - "\tVR0OBBYEFDTt/sf9PeMaZDHkUIldrDYMNTBZMIGaBgNVHSMEgZIwgY+AFAI4qxGj\r\n" - "\tloCLDdMVKwiljjDastqooXSkcjBwMQswCQYDVQQGEwJVSzERMA8GA1UEChMIZVNj\r\n" - "\taWVuY2UxEjAQBgNVBAsTCUF1dGhvcml0eTELMAkGA1UEAxMCQ0ExLTArBgkqhkiG\r\n" - "\t9w0BCQEWHmNhLW9wZXJhdG9yQGdyaWQtc3VwcG9ydC5hYy51a4IBADApBgNVHRIE\r\n" - "\tIjAggR5jYS1vcGVyYXRvckBncmlkLXN1cHBvcnQuYWMudWswGQYDVR0gBBIwEDAO\r\n" - "\tBgwrBgEEAdkvAQEBAQYwPQYJYIZIAYb4QgEEBDAWLmh0dHA6Ly9jYS5ncmlkLXN1\r\n" - "\tcHBvcnQuYWMudmT4sopwqlBWsvcHViL2NybC9jYWNybC5jcmwwPQYJYIZIAYb4QgEDBDAWLmh0\r\n" - "\tdHA6Ly9jYS5ncmlkLXN1cHBvcnQuYWMudWsvcHViL2NybC9jYWNybC5jcmwwPwYD\r\n" - "\tVR0fBDgwNjA0oDKgMIYuaHR0cDovL2NhLmdyaWQt5hYy51ay9wdWIv\r\n" - "\tY3JsL2NhY3JsLmNybDANBgkqhkiG9w0BAQUFAAOCAQEAS/U4iiooBENGW/Hwmmd3\r\n" - "\tXCy6Zrt08YjKCzGNjorT98g8uGsqYjSxv/hmi0qlnlHs+k/3Iobc3LjS5AMYr5L8\r\n" - "\tUO7OSkgFFlLHQyC9JzPfmLCAugvzEbyv4Olnsr8hbxF1MbKZoQxUZtMVu29wjfXk\r\n" - "\thTeApBv7eaKCWpSp7MCbvgzm74izKhu3vlDk9w6qVrxePfGgpKPqfHiOoGhFnbTK\r\n" - "\twTC6o2xq5y0qZ03JonF7OJspEd3I5zKY3E+ov7/ZhW6DqT8UFvsAdjvQbXyhV8Eu\r\n" - "\tYhixw1aKEPzNjNowuIseVogKOLXxWI5vAi5HgXdS0/ES5gDGsABo4fqovUKlgop3\r\n" - "\tRA==\r\n" - "\t-----END CERTIFICATE-----\r\n" - "\r\n"; - test_simple(dumbfuck2, HPE_OK); - -#if 0 - // NOTE(Wed Nov 18 11:57:27 CET 2009) this seems okay. we just read body - // until EOF. - // - // no content-length - // error if there is a body without content length - const char *bad_get_no_headers_no_body = "GET /bad_get_no_headers_no_body/world HTTP/1.1\r\n" - "Accept: */*\r\n" - "\r\n" - "HELLO"; - test_simple(bad_get_no_headers_no_body, 0); -#endif - /* TODO sending junk and large headers gets rejected */ - - - /* check to make sure our predefined requests are okay */ - for (i = 0; requests[i].name; i++) { - test_message(&requests[i]); - } - - for (i = 0; i < request_count; i++) { - test_message_pause(&requests[i]); - } - - for (i = 0; i < request_count; i++) { - if (!requests[i].should_keep_alive) continue; - for (j = 0; j < request_count; j++) { - if (!requests[j].should_keep_alive) continue; - for (k = 0; k < request_count; k++) { - test_multiple3(&requests[i], &requests[j], &requests[k]); - } - } - } - - printf("request scan 1/4 "); - test_scan( &requests[GET_NO_HEADERS_NO_BODY] - , &requests[GET_ONE_HEADER_NO_BODY] - , &requests[GET_NO_HEADERS_NO_BODY] - ); - - printf("request scan 2/4 "); - test_scan( &requests[POST_CHUNKED_ALL_YOUR_BASE] - , &requests[POST_IDENTITY_BODY_WORLD] - , &requests[GET_FUNKY_CONTENT_LENGTH] - ); - - printf("request scan 3/4 "); - test_scan( &requests[TWO_CHUNKS_MULT_ZERO_END] - , &requests[CHUNKED_W_TRAILING_HEADERS] - , &requests[CHUNKED_W_BULLSHIT_AFTER_LENGTH] - ); - - printf("request scan 4/4 "); - test_scan( &requests[QUERY_URL_WITH_QUESTION_MARK_GET] - , &requests[PREFIX_NEWLINE_GET ] - , &requests[CONNECT_REQUEST] - ); - - test_status_complete(); - - puts("requests okay"); - - return 0; -} diff --git a/node/Address.hpp b/node/Address.hpp index 8573a268..fce4f20c 100644 --- a/node/Address.hpp +++ b/node/Address.hpp @@ -198,7 +198,7 @@ public: inline std::string toString() const { char buf[16]; - sprintf(buf,"%.10llx",_a); + sprintf(buf,"%.10llx",(unsigned long long)_a); return std::string(buf); }; diff --git a/node/Http.cpp b/node/Http.cpp deleted file mode 100644 index 86ae2d25..00000000 --- a/node/Http.cpp +++ /dev/null @@ -1,323 +0,0 @@ -/* - * ZeroTier One - Global Peer to Peer Ethernet - * Copyright (C) 2012-2013 ZeroTier Networks LLC - * - * This program is free software: you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation, either version 3 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program. If not, see . - * - * -- - * - * ZeroTier may be used and distributed under the terms of the GPLv3, which - * are available at: http://www.gnu.org/licenses/gpl-3.0.html - * - * If you would like to embed ZeroTier into a commercial application or - * redistribute it in a modified binary form, please contact ZeroTier Networks - * LLC. Start here: http://www.zerotier.com/ - */ - -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -#ifndef _WIN32 -#include -#endif - -#include "Http.hpp" -#include "Utils.hpp" -#include "InetAddress.hpp" - -static http_parser_settings _http_parser_settings; - -namespace ZeroTier { - -static bool _sendAll(int fd,const void *buf,unsigned int len) -{ - for(;;) { - int n = (int)::send(fd,buf,len,0); - if ((n < 0)&&(errno == EINTR)) - continue; - return (n == (int)len); - } -} - -const std::map Http::EMPTY_HEADERS; - -Http::Request::Request( - Http::Method m, - const std::string &url, - const std::map &rh, - const std::string &rb, - bool (*handler)(Request *,void *,const std::string &,int,const std::map &,const std::string &), - void *arg) : - _url(url), - _requestHeaders(rh), - _handler(handler), - _arg(arg), - _method(m), - _fd(0) -{ - _http_parser_settings.on_message_begin = &Http::Request::_http_on_message_begin; - _http_parser_settings.on_url = &Http::Request::_http_on_url; - _http_parser_settings.on_status_complete = &Http::Request::_http_on_status_complete; - _http_parser_settings.on_header_field = &Http::Request::_http_on_header_field; - _http_parser_settings.on_header_value = &Http::Request::_http_on_header_value; - _http_parser_settings.on_headers_complete = &Http::Request::_http_on_headers_complete; - _http_parser_settings.on_body = &Http::Request::_http_on_body; - _http_parser_settings.on_message_complete = &Http::Request::_http_on_message_complete; - - start(); -} - -Http::Request::~Request() -{ - if (_fd > 0) - ::close(_fd); - join(); -} - -void Http::Request::main() - throw() -{ - char buf[131072]; - - try { - http_parser_init(&_parser,HTTP_RESPONSE); - _parser.data = this; - - http_parser_url urlParsed; - if (http_parser_parse_url(_url.c_str(),_url.length(),0,&urlParsed)) { - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"URL parse error"); - return; - } - if (!(urlParsed.field_set & (1 << UF_SCHEMA))) { - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"URL specifies no schema"); - return; - } - - std::string schema(_url.substr(urlParsed.field_data[UF_SCHEMA].off,urlParsed.field_data[UF_SCHEMA].len)); - - if (schema == "file") { - const std::string filePath(_url.substr(urlParsed.field_data[UF_PATH].off,urlParsed.field_data[UF_PATH].len)); - - uint64_t lm = Utils::getLastModified(filePath.c_str()); - if (lm) { - const std::map::const_iterator ifModSince(_requestHeaders.find("If-Modified-Since")); - if ((ifModSince != _requestHeaders.end())&&(ifModSince->second.length())) { - uint64_t t64 = Utils::fromRfc1123(ifModSince->second); - if ((t64)&&(lm > t64)) { - suicidalThread = !_handler(this,_arg,_url,304,_responseHeaders,""); - return; - } - } - - if (Utils::readFile(filePath.c_str(),_responseBody)) { - _responseHeaders["Last-Modified"] = Utils::toRfc1123(lm); - suicidalThread = !_handler(this,_arg,_url,200,_responseHeaders,_responseBody); - return; - } - } - - suicidalThread = !_handler(this,_arg,_url,404,_responseHeaders,"file not found or not readable"); - return; - } else if (schema == "http") { - if (!(urlParsed.field_set & (1 << UF_HOST))) { - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"URL contains no host"); - return; - } - std::string host(_url.substr(urlParsed.field_data[UF_HOST].off,urlParsed.field_data[UF_HOST].len)); - - std::list v4,v6; - { - struct addrinfo *res = (struct addrinfo *)0; - if (!getaddrinfo(host.c_str(),(const char *)0,(const struct addrinfo *)0,&res)) { - struct addrinfo *p = res; - do { - if (p->ai_family == AF_INET) - v4.push_back(InetAddress(p->ai_addr)); - else if (p->ai_family == AF_INET6) - v6.push_back(InetAddress(p->ai_addr)); - } while ((p = p->ai_next)); - freeaddrinfo(res); - } - } - - std::list *addrList; - if (v4.empty()&&v6.empty()) { - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"could not find address for host in URL"); - return; - } else if (v4.empty()) { - addrList = &v6; - } else { - addrList = &v4; - } - InetAddress *addr; - { - addrList->sort(); - addrList->unique(); - unsigned int i = 0,k = 0; - k = rand() % addrList->size(); - std::list::iterator a(addrList->begin()); - while (i++ != k) ++a; - addr = &(*a); - } - - int remotePort = ((urlParsed.field_set & (1 << UF_PORT))&&(urlParsed.port)) ? (int)urlParsed.port : (int)80; - if ((remotePort <= 0)||(remotePort > 0xffff)) { - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"URL port out of range"); - return; - } - addr->setPort(remotePort); - - _fd = socket(addr->isV6() ? AF_INET6 : AF_INET,SOCK_STREAM,0); - if (_fd <= 0) { - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"could not open socket"); - return; - } - - for(;;) { - if (connect(_fd,addr->saddr(),addr->saddrLen())) { - if (errno == EINTR) - continue; - ::close(_fd); _fd = 0; - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"connection failed to remote host"); - return; - } else break; - } - - const char *mstr = "GET"; - switch(_method) { - case HTTP_METHOD_HEAD: mstr = "HEAD"; break; - default: break; - } - int mlen = (int)snprintf(buf,sizeof(buf),"%s %s HTTP/1.1\r\nAccept-Encoding: \r\nHost: %s\r\n",mstr,_url.substr(urlParsed.field_data[UF_PATH].off,urlParsed.field_data[UF_PATH].len).c_str(),host.c_str()); - if (mlen >= (int)sizeof(buf)) { - ::close(_fd); _fd = 0; - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"URL too long"); - return; - } - if (!_sendAll(_fd,buf,mlen)) { - ::close(_fd); _fd = 0; - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"write error"); - return; - } - - for(std::map::const_iterator rh(_requestHeaders.begin());rh!=_requestHeaders.end();++rh) { - mlen = (int)snprintf(buf,sizeof(buf),"%s: %s\r\n",rh->first.c_str(),rh->second.c_str()); - if (mlen >= (int)sizeof(buf)) { - ::close(_fd); _fd = 0; - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"header too long"); - return; - } - if (!_sendAll(_fd,buf,mlen)) { - ::close(_fd); _fd = 0; - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"write error"); - return; - } - } - - if (!_sendAll(_fd,"\r\n",2)) { - ::close(_fd); _fd = 0; - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"write error"); - return; - } - - _responseStatusCode = 0; - _messageComplete = false; - for(;;) { - mlen = (int)::recv(_fd,buf,sizeof(buf),0); - if (mlen < 0) { - if (errno != EINTR) - break; - else continue; - } - if (((int)http_parser_execute(&_parser,&_http_parser_settings,buf,mlen) != mlen)||(_parser.upgrade)) { - ::close(_fd); _fd = 0; - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"invalid HTTP response from server"); - return; - } - if (_messageComplete) { - ::close(_fd); _fd = 0; - suicidalThread = !_handler(this,_arg,_url,_responseStatusCode,_responseHeaders,_responseBody); - return; - } - } - - ::close(_fd); _fd = 0; - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"empty HTTP response from server"); - return; - } else { - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"only 'file' and 'http' methods are supported"); - return; - } - } catch ( ... ) { - suicidalThread = !_handler(this,_arg,_url,0,_responseHeaders,"unexpected exception retrieving URL"); - return; - } -} - -int Http::Request::_http_on_message_begin(http_parser *parser) -{ - return 0; -} -int Http::Request::_http_on_url(http_parser *parser,const char *data,size_t length) -{ - return 0; -} -int Http::Request::_http_on_status_complete(http_parser *parser) -{ - Http::Request *r = (Http::Request *)parser->data; - r->_responseStatusCode = parser->status_code; - return 0; -} -int Http::Request::_http_on_header_field(http_parser *parser,const char *data,size_t length) -{ - Http::Request *r = (Http::Request *)parser->data; - if ((r->_currentHeaderField.length())&&(r->_responseHeaders.find(r->_currentHeaderField) != r->_responseHeaders.end())) - r->_currentHeaderField.assign(""); - r->_currentHeaderField.append(data,length); - return 0; -} -int Http::Request::_http_on_header_value(http_parser *parser,const char *data,size_t length) -{ - Http::Request *r = (Http::Request *)parser->data; - if (r->_currentHeaderField.length()) - r->_responseHeaders[r->_currentHeaderField].append(data,length); - return 0; -} -int Http::Request::_http_on_headers_complete(http_parser *parser) -{ - Http::Request *r = (Http::Request *)parser->data; - return ((r->_method == Http::HTTP_METHOD_HEAD) ? 1 : 0); -} -int Http::Request::_http_on_body(http_parser *parser,const char *data,size_t length) -{ - Http::Request *r = (Http::Request *)parser->data; - r->_responseBody.append(data,length); - return 0; -} -int Http::Request::_http_on_message_complete(http_parser *parser) -{ - Http::Request *r = (Http::Request *)parser->data; - r->_messageComplete = true; - return 0; -} - -} // namespace ZeroTier diff --git a/node/Http.hpp b/node/Http.hpp deleted file mode 100644 index a099b15d..00000000 --- a/node/Http.hpp +++ /dev/null @@ -1,129 +0,0 @@ -/* - * ZeroTier One - Global Peer to Peer Ethernet - * Copyright (C) 2012-2013 ZeroTier Networks LLC - * - * This program is free software: you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation, either version 3 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program. If not, see . - * - * -- - * - * ZeroTier may be used and distributed under the terms of the GPLv3, which - * are available at: http://www.gnu.org/licenses/gpl-3.0.html - * - * If you would like to embed ZeroTier into a commercial application or - * redistribute it in a modified binary form, please contact ZeroTier Networks - * LLC. Start here: http://www.zerotier.com/ - */ - -#ifndef _ZT_HTTP_HPP -#define _ZT_HTTP_HPP - -#include -#include -#include -#include "Thread.hpp" - -#include "../ext/http-parser/http_parser.h" - -namespace ZeroTier { - -class Http -{ -public: - /** - * HTTP request methods - */ - enum Method - { - HTTP_METHOD_GET, - HTTP_METHOD_HEAD - }; - - /** - * An empty headers map for convenience - */ - static const std::map EMPTY_HEADERS; - - /** - * HTTP request - */ - class Request : protected Thread - { - public: - /** - * Create and issue an HTTP request - * - * The supplied handler is called when the request is - * complete or if an error occurs. A code of zero indicates - * that the server could not be reached, and a description - * of the error will be in 'body'. If the handler returns - * false the Request object deletes itself. Otherwise the - * object must be deleted by other code. - * - * @param m Request method - * @param url Destination URL - * @param rh Request headers - * @param rb Request body or empty string for none (currently unused) - * @param handler Request handler function - * @param arg First argument to request handler - */ - Request( - Http::Method m, - const std::string &url, - const std::map &rh, - const std::string &rb, - bool (*handler)(Request *,void *,const std::string &,int,const std::map &,const std::string &), - void *arg); - - /** - * Destruction cancels any in-progress request - */ - virtual ~Request(); - - protected: - virtual void main() - throw(); - - private: - // HTTP parser handlers - static int _http_on_message_begin(http_parser *parser); - static int _http_on_url(http_parser *parser,const char *data,size_t length); - static int _http_on_status_complete(http_parser *parser); - static int _http_on_header_field(http_parser *parser,const char *data,size_t length); - static int _http_on_header_value(http_parser *parser,const char *data,size_t length); - static int _http_on_headers_complete(http_parser *parser); - static int _http_on_body(http_parser *parser,const char *data,size_t length); - static int _http_on_message_complete(http_parser *parser); - - http_parser _parser; - std::string _url; - - std::map _requestHeaders; - std::map _responseHeaders; - - std::string _currentHeaderField; - std::string _responseBody; - - bool (*_handler)(Request *,void *,const std::string &,int,const std::map &,const std::string &); - void *_arg; - - Http::Method _method; - int _responseStatusCode; - bool _messageComplete; - volatile int _fd; - }; -}; - -} // namespace ZeroTier - -#endif diff --git a/node/Network.hpp b/node/Network.hpp index 62c0e978..637c6664 100644 --- a/node/Network.hpp +++ b/node/Network.hpp @@ -107,14 +107,14 @@ public: inline void setNetworkId(uint64_t id) { char buf[32]; - sprintf(buf,"%llu",id); + sprintf(buf,"%.16llx",(unsigned long long)id); (*this)["nwid"] = buf; } inline uint64_t networkId() const throw(std::invalid_argument) { - return strtoull(get("nwid").c_str(),(char **)0,10); + return strtoull(get("nwid").c_str(),(char **)0,16); } inline void setPeerAddress(Address &a) @@ -137,9 +137,9 @@ public: inline void setTimestamp(uint64_t ts,uint64_t maxDelta) { char foo[32]; - sprintf(foo,"%llu",ts); + sprintf(foo,"%llu",(unsigned long long)ts); (*this)["ts"] = foo; - sprintf(foo,"%llu",maxDelta); + sprintf(foo,"%llu",(unsigned long long)maxDelta); (*this)["~ts"] = foo; } @@ -211,14 +211,14 @@ public: inline void setNetworkId(uint64_t id) { char buf[32]; - sprintf(buf,"%llu",id); + sprintf(buf,"%.16llx",(unsigned long long)id); (*this)["nwid"] = buf; } inline uint64_t networkId() const throw(std::invalid_argument) { - return strtoull(get("nwid").c_str(),(char **)0,10); + return strtoull(get("nwid").c_str(),(char **)0,16); } inline void setPeerAddress(Address &a) diff --git a/objects.mk b/objects.mk index c6fab0b6..4d828923 100644 --- a/objects.mk +++ b/objects.mk @@ -1,5 +1,4 @@ OBJS=\ - ext/http-parser/http_parser.o \ ext/kissdb/kissdb.o \ ext/lz4/lz4hc.o \ ext/lz4/lz4.o \ @@ -9,7 +8,6 @@ OBJS=\ node/EthernetTap.o \ node/Filter.o \ node/HMAC.o \ - node/Http.o \ node/Identity.o \ node/InetAddress.o \ node/Logger.o \ -- cgit v1.2.3