Session Handoff [auto-1299]: Fold Typographic Punctuation in pr_review.py reply --match (#1299)

[ptr727]

Next steps, in priority order

1. Fix scripts/pr_review.py reply --match Cannot Select a Thread Whose Body Uses Typographic Quotes #1299: pr_review.py reply --match cannot select a thread whose body uses typographic punctuation. In scripts/pr_review.py matching_threads(), normalise both the --match pattern and each candidate body before the substring compare, folding at least the typographic single and double quotes, the en and em dashes, and the ellipsis to their ASCII equivalents. Keep the refusal design unchanged: no match and more than one match still refuse. The issue's secondary suggestion, saying on NO_MATCH how many unresolved threads the pull request has, is in scope where it stays small. Done looks like a test in scripts/tests/test_pr_review.py selecting a thread whose body carries typographic quotes from an ASCII pattern and the reverse, failing when the normalisation is reverted, and a pull request into develop carrying Closes on promotion: #1299.

External blockers

None known at pick time.

Internal dependencies

None.

State

• Created by an unattended-handoff picker at scope develop. No branch, worktree, or pull request exists for this lane yet.
• Re-derive matching_threads() and its callers from origin/develop rather than trusting this summary.

The parked decision queue

Empty for this lane.

What the last round did

Nothing on this lane yet. The picker chose #1299 as a correctness defect in a script every review loop runs, with its fix stated in the issue and no open choice.

What not to repeat

Nothing yet.

New learnings

None yet.

[ptr727]

Merged: PR #1878 (squash commit a50dcc5, feature/auto-1299 -> develop). Fixes #1299.

matching_threads() in scripts/pr_review.py now folds typographic single/double quotes, en/em dash, and ellipsis to ASCII before the reply --match substring compare, on both the pattern and each candidate thread body. NO_MATCH now states the total unresolved-thread count.

Review loop: two rounds. Round 1 (local-strict-review, 5 findings) fixed before the first push: the new coverage test was tautological (iterated the implementation's own fold table rather than an independent expected set), a docstring carried a semicolon, and another docstring overclaimed ASCII-safety beyond what the table covers. Round 2 (GitHub Copilot on the PR) found a real latent bug in the round-1 fix itself: the new parameterized test's stdout assertion read a buffer shared across all 7 subTest iterations rather than each iteration's own output; fixed in 7818762 and swept the file for sibling instances of the same shape (found none).

Two related gaps found during local review are filed separately as out of scope for this fix:

• pr_review.py reply --match: an ASCII double-hyphen and an em dash still miss each other #1876 (ASCII -- and a real em dash still miss each other)
• pr_review.py reply --match: a pattern copied from a printed unresolved: line can still miss its own thread #1877 (a --match pattern copied from this script's own truncated/whitespace-collapsed unresolved: line can still miss)

Lane closed, no successor needed.