Merge fix/baileys-lid-group-senderkey-patch - Correct patch format #13
Conversation
Baileys v6.7.19 includes sender's own devices in senderKeyJids when sending to LID-addressed groups, causing WhatsApp to reject with 'not-acceptable' error. This patch filters sender's own devices from the senderKeyJids list before sending group messages, following the same pattern Baileys uses for non-group messages (meJids vs otherJids filtering). Changes: - Add Baileys patch in patches/baileys+6.7.19+lid-group-send.patch - Update Dockerfile to apply patch after npm ci - Extract meUser from meId and meLid from authState.creds - Skip sender's own devices in senderKeyJids loop Fixes: #issue-number Related: WhiskeySockets/Baileys#1690, EvolutionAPI#1768
…tion Alpine Linux does not include 'patch' by default. Without this, the Baileys LID group send patch would silently fail during build, leaving the fix unapplied in production images. Changes: - Add 'patch' to apk install list - Remove '|| echo' fallback to fail fast on patch errors
…eys-lid-group-senderkey-patch
- Use unified diff format without git headers - Change from -p1 to -p0 for correct path stripping - Ensures patch applies correctly during Docker build
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| # [WIDGET-WORKS] Apply Baileys patch for LID group send fix | ||
| COPY ./patches ./patches | ||
| RUN patch -p1 < patches/baileys+6.7.19+lid-group-send.patch | ||
| RUN patch -p0 < patches/baileys+6.7.19+lid-group-send.patch |
There was a problem hiding this comment.
Switching the patch invocation to patch -p0 while the patch file still uses a/node_modules/.../b/... paths means the build step now looks for files under a/node_modules/... and fails before the image is built (the actual files live under node_modules). With -p1 the leading a/ was stripped and the patch applied; with -p0 the patch cannot be applied during docker build, breaking the image.
Useful? React with 👍 / 👎.
No description provided.