Skip to content

fix: strip HTML entities before matching issue citations - #794

Merged
omesser merged 1 commit into
mainfrom
cursor/strip-html-entities-776-fea6
Sep 16, 2026
Merged

omesser merged 1 commit into
mainfrom
cursor/strip-html-entities-776-fea6

Conversation

@omesser

@omesser omesser commented Sep 16, 2026 •

Copy link
Copy Markdown
Owner

What

Strip &#\d+; in scripts/comment-census.py before #\d+ / ADR matching so HTML entities are not counted as issue citations.

Why

j in platform.rs was read as issue #106 and falsely multi-cited a block for comment compaction. Same for ' / ". #758 landed the census; this closes the known false positive (#776).

How to test

  • census ignores HTML entities as issue citations (entity-only block not counted)
  • census still detects real issue citations after stripping entities (real #371 still counts)
  • Existing tests/comment-census.test.js suite still green

Closes #776

— Cursor agent (Coder), on @omesser's behalf.

Fixes #776. HTML entities like j, ', " were false-positive
matching the #\d+ issue-citation pattern. Strip &#\d+; before matching.

Test coverage: new tests verify HTML entities don't count as issue
citations, and real citations like #371 still work correctly.

Co-authored-by: Oded Messer <odedmesser@gmail.com>
@omesser
omesser marked this pull request as ready for review September 16, 2026 22:45
@omesser
omesser merged commit 6a49835 into main Sep 16, 2026
5 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

fix(scripts): Comment census reads the HTML entity &#106; as a citation of issue #106

2 participants