Cherry-pick PRs for WP 7.1.1 - #82383
Conversation
Co-authored-by: t-hamano <wildworks@git.wordpress.org> Co-authored-by: tellthemachines <isabel_brison@git.wordpress.org>
…nage_options cap (#81813) * Resolve the preloading correctly based on the capabilities * Remove read cap check and fetch global style for non-admin using view context
…82046) Co-authored-by: adamsilverstein <adamsilverstein@git.wordpress.org> Co-authored-by: andrewserong <andrewserong@git.wordpress.org>
* Accordion Panel: Reset padding-block when panel is hidden. When the accordion content style was changed to support in-browser search via `hidden="until-found"` (#74744), the browser applies `content-visibility: hidden`. Because `content-visibility: hidden` hides contents without removing the container box model, any user- or theme-defined padding on the Accordion Panel block produced an empty gap below the title when the accordion was closed. This resets `padding-block: 0 !important` on `.wp-block-accordion-panel[hidden]` so the closed panel occupies no vertical space while preserving the full padding when the panel is open. Fixes #65900 (Trac #65900). * Reset borders, box shadows, and padding on hidden accordion panel. Addresses review feedback by ensuring borders and box shadows are also reset when the accordion panel is hidden, along with padding. * Add changelog entry for Accordion Panel hidden padding fix --------- Co-authored-by: wprashed <wprashed@git.wordpress.org> Co-authored-by: t-hamano <wildworks@git.wordpress.org>
…1964) Unlinked contributors: YamabikoLab. Co-authored-by: shail-mehta <shailu25@git.wordpress.org> Co-authored-by: Mamaduka <mamaduka@git.wordpress.org> Co-authored-by: t-hamano <wildworks@git.wordpress.org> Co-authored-by: andrewserong <andrewserong@git.wordpress.org>
Co-authored-by: adamsilverstein <adamsilverstein@git.wordpress.org> Co-authored-by: andrewserong <andrewserong@git.wordpress.org> Co-authored-by: jasmussen <joen@git.wordpress.org> Co-authored-by: aaronjorbin <jorbin@git.wordpress.org>
Co-authored-by: adamsilverstein <adamsilverstein@git.wordpress.org> Co-authored-by: swissspidy <swissspidy@git.wordpress.org> Co-authored-by: kleisauke <kleisauke@git.wordpress.org> Co-authored-by: t-hamano <wildworks@git.wordpress.org> Props to @nosilver4u for the original report.
…82316) * Image block: sync / update image size settings when applying a media editor crop * Simplify * Simplify the tests a little * Further test tidying * Clarify comment * Add fallback to the cached attachment * Simplify, we don't need the third param Co-authored-by: andrewserong <andrewserong@git.wordpress.org> Co-authored-by: ramonjd <ramonopoly@git.wordpress.org> Co-authored-by: adamsilverstein <adamsilverstein@git.wordpress.org> Co-authored-by: gregbenz <gregbenz@git.wordpress.org>
|
Important Review skippedAuto reviews are disabled on base/target branches other than the default branch. Please check the settings in the CodeRabbit UI or the ⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Advanced Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
* Initial fix for gallery images with links no longer being `object-fit: cover` * Limit fix to flex galleries, avoid applying to grid gallery * Changelog * Fix dynamic gallery case * PR feedback --------- Co-authored-by: talldan <talldanwp@git.wordpress.org> Co-authored-by: andrewserong <andrewserong@git.wordpress.org> Co-authored-by: ramonjd <ramonopoly@git.wordpress.org> Co-authored-by: SainathPoojary <sainathpoojary@git.wordpress.org> Co-authored-by: Azragh <azragh@git.wordpress.org>
…andalone pages (#82254) Co-authored-by: shameemreza <shameemreza@git.wordpress.org> Co-authored-by: youknowriad <youknowriad@git.wordpress.org> Co-authored-by: annezazu <annezazu@git.wordpress.org>
fd4b429 to
0cdc115
Compare
|
Size Change: +1.08 kB (+0.01%) Total Size: 7.75 MB 📦 View Changed
|
|
The following accounts have interacted with this PR and/or linked issues. I will continue to update these lists as activity occurs. You can also manually ask me to refresh this list by adding the If you're merging code through a pull request on GitHub, copy and paste the following into the bottom of the merge commit message. To understand the WordPress project's expectations around crediting contributors, please review the Contributor Attribution page in the Core Handbook. |
|
I merged this pr and am going to cherry pick it back to the cherry-pick-wp-7.1.1 branch |
Co-authored-by: antwonw <antwonw@git.wordpress.org> Co-authored-by: adamsilverstein <adamsilverstein@git.wordpress.org> Co-authored-by: t-hamano <wildworks@git.wordpress.org>
|
Regarding the CI error, it should be resolved by cherry-picking #82478. |
This comment was marked as off-topic.
This comment was marked as off-topic.
|
Apparently I need to cherry pick a2e9438 to get the tests to pass. |
* CI: Add a shared action for a single PR automation comment Every automation posts its own comment today, so a pull request collects three to five bot comments before a human says anything. Add tools/pr-meta, an action that maintains one comment split into marker-delimited sections. Each producer owns one section and rewrites only its own, so the comment reads the same however the workflows interleave. Writers hold a shared concurrency group rather than verifying their own write, which cannot detect a lost update: two writers reading the same body both succeed, and the first section is gone. queue: max is required there, since the default cancels a pending writer once a third arrives. actionlint does not know the queue key. Its last release predates the feature and rhysd/actionlint#654 is open, so suppress that one message. The action imports nothing: depending on @actions/core would mean a full npm install in every writer job just to edit a comment. * CI: Report the bundle size in the shared PR comment compressed-size-action sets its rendered markdown as an output before it decides whether to comment, so use-check diverts the comment into a check run and the body still reaches the shared comment. use-check is conditional on the pull request not being from a fork. A fork's token cannot create a check, and unlike its comment path, its check path does not catch that failure and would fail the job. * Report Flaky Tests: Render the report instead of commenting The action now writes markdown to output-path and no longer talks to GitHub, so the report joins the shared PR comment. Its repo-token input and its GitHub API client go with it. The workflow only invoked the action when a report artifact existed, so a clean run could never clear a stale report. The writer job now runs on every pull request and treats the missing artifact as an empty section. * CI: Post the performance results on the pull request The results were only ever written to the job summary, buried in the logs of a workflow run. Limiting the number of bot comments was the stated reason for not posting them, which the shared comment removes. Each shard already renders a summary, so upload those and merge them in the results job, keeping the heading and preamble they share once. * CI: Check the type label in the workflow itself github-action-required-labels cannot report through the shared comment: its failure path exposes only a status, never the rendered message or the labels it matched. Turning its comment off also disables the branch that deletes its own warning once the labels are fixed, which would strand a warning on every pull request that ever had one. The rule is one label matching ^\[Type\], so read the labels directly. Read them at the point of writing rather than from the event payload, since runs queue behind each other and an older one can arrive after a newer one already fixed them. Also fire on opened and reopened. A pull request opened without a type label got no warning until some unrelated event happened. * Project Automation: Expose the contributor prompts as outputs Both first-time contributor tasks posted their own comment: a welcome on the pull request, and a prompt to link a WordPress.org profile once it merged. They now set outputs and the workflow posts them, so both join the shared PR comment. The account link task runs on push, whose payload carries no pull request number, so it exposes the one it resolved alongside the prompt. Its writer needs that for the comment and for the concurrency group both. * Changelogs: Reference the pull request * PR meta: Add publishConfig and emoji section headings Every other tools/* workspace carries publishConfig despite being private, and the repo-wide lint rule requires it. Lead each heading with an emoji so a reader can pick their section out of a comment holding seven of them without reading any of the others. * PR meta: Wrap the comment title in emoji Marks the comment as automation at a glance, matching the section headings below it. * PR meta: Collapse the long sections behind a fold Performance tables and flaky test traces would otherwise push the rest of the comment out of view. The fold sits outside the section delimiters, alongside the heading and the footer. Everything between them is read back as the body, so a fold kept inside would nest into itself on every write. * PR meta: Do not let a failed check clear a section Three ways a section was cleared by something other than a result: Every write re-renders every section, footers included, so a writer for a PR-state section needs the head too. Without it, updating Labels dropped the 'not the current head' marker from a stale Bundle size, presenting an old result as current. The label reporter ran even when the check step failed, where a missing body file means 'labels are fine', so a failed API call erased a warning nothing had disproved. The flaky renderer treated every readdir failure as a clean run. Only a missing directory means clean; anything else now surfaces. * PR meta: Skip the write when the head cannot be read Swallowing that failure re-rendered every section without a head, which strips the 'not the current head' footer from stale results and presents them as current. Letting it reach the outer handler skips the write instead, and adds the first tests covering the action's entry point. Also drop the two dependencies report-flaky-tests no longer uses, consolidate its unreleased changelog entries now that it renders rather than comments, and document the automation action's new outputs. * PR meta: Truncate between blocks, and fail a broken flaky run Truncation cut at a fixed offset, so a cut landing inside a stack trace left the fence open and rendered the note and its link as code. It now cuts at a paragraph break and closes whatever fences and disclosures the cut left open. The flaky renderer logged its errors and left the step green, which looks exactly like a clean run, so the writer downstream cleared a report that nothing had disproved. It now fails the step, and the writer runs only when the job producing the report succeeded. Also update the README, package description and action description, which still described the action as posting a comment. Co-authored-by: manzoorwanijk <manzoorwanijk@git.wordpress.org> Co-authored-by: ciampo <mciampini@git.wordpress.org> (cherry picked from commit a2e9438) Adapted for wp/7.1: - The performance.yml changes are not ported: they build on the sharded performance workflow that only exists on trunk. - The bundle-size.yml and end2end-test.yml changes are applied onto the branch's older job layout (no .github/setup-npm, branch-pinned action hashes); end2end folds the render step into merge-artifacts as on trunk and drops the trunk-checkout report-to-issues job. - report-flaky-tests keeps @playwright/test at ^1.61.1 so the branch's Playwright version stays pinned. - Trunk-only drift entries are dropped from the project-management-automation changelog. - packages/report-flaky-tests/src/types.ts comes along from trunk: the rewritten sources use ReportedFlakyTest, which trunk added earlier. - tools/pr-meta and report-flaky-tests extend the branch's tsconfig.base.json, as the dev/build tsconfig split does not exist on wp/7.1. - An obsolete snapshot left by the replaced test suite is pruned. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01DP49q1JdSxUhzFFE7ga2DS
The wp/7.1 adaptation of the unified PR comment tooling (#82249) brought trunk's jest 30 ranges into report-flaky-tests and tools/pr-meta, while every other workspace on this branch pins the 29 line. syncpack enforces one version per dependency across the repo, so lint:deps failed. Align @jest/test-result and jest-message-util to ^29.6.2 and @types/jest to ^29.5.14, matching the rest of the branch. Typecheck, lint and the 47 unit tests of both workspaces pass unchanged on jest 29. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01DP49q1JdSxUhzFFE7ga2DS
🤖 PR meta 🤖📦 Bundle sizeSize Change: +1.29 kB (+0.02%) Total Size: 7.75 MB 📦 View Changed
🏁 Flaky testsShow the failuresSome tests passed with failed attempts. The failures may not be related to this commit but are still reported for visibility. See the documentation for more information. Should save the changes in
|
(cherry picked from commit a474f40)
The report-flaky-tests action runs its TypeScript sources directly under Node 24, which resolves relative imports only with an explicit extension. Trunk made this change in #81218, but the branch picked up the Node 24 action config from #82249 without it, so the "Render the flaky tests report" step crashes with ERR_MODULE_NOT_FOUND whenever a test flaked. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01JrQHYVWEr61osMiYVd4AJG
|
|
de457f2 to
d62073f
Compare
|
For minor releases, the key is to cherry-pick all necessary PRs as-is and to resolve any resulting conflicts appropriately. While the |
t-hamano
left a comment
There was a problem hiding this comment.
Looks good from my end.
The following PRs have not been merged yet, and therefore not cherry-picked:
Since this seems to have already been approved, we could cherry-pick this as well, but I'll leave the decision to everyone.
Yes, I am fixing the CI now and then will merge and cherry pick |
|
I committed https://core.trac.wordpress.org/ticket/66056 and reopened for backport consideration. i'll need a second committer sign off. Thats the backport for #81947 PHP. |
|
WordPress/wordpress-develop#13412 has been cherry picked back to the 7.1 branch in https://core.trac.wordpress.org/changeset/63572 |
Re-pin to the current head of the WordPress/gutenberg#82383 cherry-pick branch, which adds the #81947, #82265, and #82249 backports on top of the previous pin. Only the `editor` and `upload-media` package hashes change in the regenerated manifest. See #66037. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01WRsjgMZe3xuN98DYxZ3Kf6
Re-pins to the `wp/7.1` tip now that the 7.1.1 cherry-pick PR (WordPress/gutenberg#82383) has landed via rebase and merge. The build content is identical to the previous pin `3f82d42`, so the regenerated manifests are unchanged and only the pinned hash differs. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01GXuW8W41AXnGkpT3hfEhzE
Core ticket: https://core.trac.wordpress.org/ticket/66037
WordPress 7.1.1 is scheduled to be released on September 17th, and 7.1.1 RC1 is scheduled to be released on Semptember 10th. In preparation for that, this PR backports the PRs with the Backport to WP Minor Release label to the wp/7.1 branch.
cc @adamsilverstein, @adrianduffell
Caution
Personal note: Be sure to change to "Rebase and merge" not the default "Squash and merge" as this allows for a full history of commits.
https://developer.wordpress.org/block-editor/contributors/code/release/package-release-and-core-updates/#:~:text=Merge%20the%20Pull%20Request%20using%20the%20%E2%80%9CRebase%20and%20Merge%E2%80%9D%20button%20to%20keep%20the%20history%20of%20the%20commits.
The following PRs are included in this PR:
postwhen the query has nopostType#82465 @ntsekouras @t-hamano (applied to the.jsfiles; trunk renamed them to.jsxin Build: Normalize JSX file extensions #80990, which is not on wp/7.1)report-flaky-testsaction out from trunk, whose interface CI: Unify the automation comments into a single PR comment #82249 changed, so every e2e run on this branch failed)The cherry-pick of #82011 needed a one-line test adaptation on wp/7.1 (fac03fe): the empty canvas is
role=buttonthere because #81231 is not on the branch.The cherry-pick of #81947 (e2a9c8d) also needed adapting, since wp/7.1 still runs Jest and its attachments controller has diverged from trunk.
Claude did the cherry-pick, its notes on what changed:
The cherry-pick of #82249 (b3acbb9) also needed adapting: wp/7.1 has neither the sharded performance workflow nor
.github/setup-npmand the dev/build tsconfig split, and it pins Playwright 1.61/Jest 29, so the workflow changes were ported onto the branch's job layout and the tooling kept on the branch's dependency lines (details in the commit message, plus follow-ups 5ea23e4 and 018ee84).The cherry-pick of #82265 (3f82d42) needed the same Jest treatment. Claude handled that one; in its words:
The following PRs have not been merged yet, and therefore not cherry-picked:
PHP changes needing Core patches
Three of the included PRs carry PHP changes that do not travel through the package build and need to land in wordpress-develop's
7.1branch for 7.1.1 (tracked by theirbackport-changelog/7.1entries):The package update itself is staged as WordPress/wordpress-develop#13422 (draft), to be re-pinned to the
wp/7.1tip once this PR is rebase-merged.How to test