| Age | Commit message (Collapse) | Author |
|---|
|
(#2108) (#2110)
On a fresh (no-consent) load the consent banner could render completely
unstyled and expand to full page height. Cookiebot delivers its dialog CSS as
a constructed stylesheet on document.adoptedStyleSheets; ReadTheDocs'
readthedocs-addons.js reassigns that whole-array property with a destructive
replace, and when that lands after Cookiebot's network-gated adoption it drops
Cookiebot's sheet, leaving #CybotCookiebotDialog at position:static expanded to
~6700px. It is a race, so it is intermittent and clears once consent is given
(the dialog then never renders).
- layout.html: a small adoptedStyleSheets shim, installed before any page
script, that preserves Cookiebot's sheet so another library cannot drop it.
- custom.css: a leak-safe safety net that keeps the dialog contained even if
its adopted sheet is ever absent. It uses only properties Cookiebot itself
sets, so it has verifiably no effect on the normal, styled banner.
🤖 Generated by [robots](https://vyos.io)
(cherry picked from commit dce7bb521e3947af59fe18895b609ec353e0450e)
Co-authored-by: Yuriy Andamasov <yuriy@vyos.io>
|
|
🤖 Generated by [robots](https://vyos.io)
|
|
success
Two real bugs in the docs site copy-to-clipboard handler, both flagged by
copilot review on PR #1886 (deferred there because that PR is scoped to typos).
Bug 1: Copy label leaked into copied snippets on narrow screens
The handler read `currentTarget.offsetParent.innerText`. The same handler
appends a visible `.copyDiv` (with a 'Copy' <p>) into that container.
Below the 992px breakpoint the label is visible (see code-snippets.css
.copyDiv > p) so users got 'Copy' appended to every copied snippet.
Fix: extract text from the <pre> element inside the container instead,
which excludes the injected button. Also switched the lookup root from
`offsetParent` to `parentElement` so the source-of-truth is the DOM
relationship (the .copyDiv is inserted as a beforeend child of the inner
.highlight div via insertAdjacentHTML), not CSS positioning.
Bug 2: Failed clipboard writes still showed 'Copied!'
The try/catch only logged on failure but the surrounding code still flipped
the button into the copiedNotifier success state. Users got false success
when writeText rejected (insecure context, permission denied, etc.).
Fix: move the success UI flip inside the try, add an explicit failure UI
flip ('Failed' text + new `.copyFailedNotifier` class with red background)
in the catch. setTimeout still reverts both classes after 2s.
Verified in browser with a sphinx-rendered fixture (jQuery 3.7 + Pygments
output): both snippets copy their own text without the Copy label, success
flow shows Copied!, simulated writeText rejection shows Failed, both states
revert after 2s.
🤖 Generated by [robots](https://vyos.io)
|
|
Replaces the broken #1886 with a fresh, properly-converted MyST set for
the sagitta (1.4.x) docs, mirroring what landed for circinus via #1897.
This PR:
- Re-imports 210 md-*.md files for sagitta. Source: ran the pipelines
rst-to-myst converter (chrisjsewell/rst-to-myst v0.4.0, with pandoc
fallback) on sagittas RST. Post-processed via the pipelines
postprocess stage (10 ordered fixes for blanks, admonitions, label
hyphens, pandoc artifacts, structural blanks, linter markers).
Compared to the broken #1886 content (which was left over from an
earlier stage-1-only run): zero raw `<div class=>` remnants.
- For 23 stems where sagittas RST is byte-identical with currents RST
(mostly stable policy/protocol pages and the 404 page), reuses currents
already-validated md-*.md content rather than re-converting.
- Drops cli and installation/cloud/aws from sagittas swap set: their
RST has SEVERE/4 "Title level inconsistent" errors that crash
rst-to-myst; they need an independent RST-source fix and are kept as
RST-only for now.
- Adds the per-page swap mechanism: scripts/swap_sources.py,
scripts/import_myst.py, the matching tests under tests/, _swap.txt
with 210 stems, _ext/vyos.py MyST renderer fallback, Makefile
swap-wrapped targets, .readthedocs.yml swap pre/post hooks.
- Adds 187 .webp images and removes 235 superseded .jpg/.png/.jpeg
static assets; flips html_logo to vyos-logo.webp.
- Adds the MyST swap-related blocks to docs/conf.py only:
myst_enable_extensions, myst_fence_as_directive, md-*.md exclude
patterns, _swap_exclude.txt reader, _prefer_webp and _copy_md_sources
setup hooks. github_version fallback set to 'sagitta' to match the
branch (parallel to currents 'current' and circinuss 'circinus').
Deliberately excluded (per user direction):
- llms.txt and sphinx-llms-txt / sphinx-sitemap config: these will
land separately for sagitta via #1870 plus a new sagitta-specific
llms.txt template PR. The conf.py here does not pull those extensions
in, so the build does not depend on the new pip packages.
Verification before pushing:
- 210 md-*.md = 210 _swap.txt stems = 210 RST siblings on sagitta (1:1:1).
- 0 files contain raw `<div class=` (the breakage that took down /en/1.5/).
- conf.py copyright/version/release preserve sagittas values
(2024 / 1.4 / "1.4.x (sagitta)") - not currents.
- html_title from currents conf.py removed - PR #1880 is the right place
for sagittas branch-localized title.
Supersedes / closes on merge:
- #1886 (broken converter output, would break /en/1.4/ if merged).
Generated by robots https://vyos.io
|
|
|
|
(cherry picked from commit b09f2222ed6181ba377a41de37a3997559a234c8)
(cherry picked from commit 1205c1e7806d4e84e03aafc94831248cbfa9b1c8)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
