pr_review.py reply --match: a pattern copied from a printed unresolved: line can still miss its own thread

[ptr727]

Context

Split out of #1299 during its own local-strict-review pass, which flagged this as a related but out-of-scope gap in the same fix.

The defect

reply_to_thread() prints each unresolved thread via describe(), which collapses the thread body's whitespace to single spaces (" ".join(body.split())) and truncates it to 120 characters before printing it as the unresolved: line a NO_MATCH refusal shows.

matching_threads() itself compares --match against the thread's raw, uncollapsed, untruncated body.

So a --match string copied verbatim from one of those printed unresolved: lines can still return NO_MATCH, in two ways:

• The pattern spans a point in the body that carried a line break, tab, or run of multiple spaces in the original, now printed back as a single space; the raw body still carries the original whitespace, so the collapsed substring is not literally present in it.
• The pattern is copied at or past the 120-character cutoff, past which the printed line carries nothing to copy from at all.

This is the same shape of failure #1299 fixed: a --match string that a caller copied straight from what this script itself printed still fails to select the thread it was copied from.

Suggested fix

Worth considering: fold whitespace the same way on both sides before the substring compare (matching describe()'s own collapse), so a pattern copied from the printed line matches the same way the original would have. The 120-character truncation is a display concern rather than a matching one and may not need a corresponding change beyond documenting the limit.