Read a Finding Entry Titled With the Suppressed Marker as No Section in pr_review.py #3173

The shape

SUPPRESSED is matched anywhere in a line, so a finding entry nested inside another collapsed section reads as a suppressed section of its own when its title carries the marker words. marker_blocks scans each <details> region, the nested entry being one, and accepts the entry's own <summary> line because HEADING matches it.

Constructed body: a previously-missed section whose second nested entry is titled 2 low confidence results are dropped, each entry a <details> opening on a severity badge summary.

<details>
<summary><strong>Previously missed (3)</strong></summary>

In code that hasn't changed since last review

<details>
<summary><severity badge> Constructed finding 1</summary>
...
</details>
<details>
<summary><severity badge> 2 low confidence results are dropped</summary>
...
</details>
...
</details>

suppressed_blocks returns one block holding that entry, so the digest reports suppressed= for a round that suppressed nothing and prints a previously-missed finding a second time under the suppressed banner, which asserts the opposite triage. unrecognized_in returns nothing, so shapes=ok.

The same holds on develop at 3bb8896 with either previously-missed spelling. It predates #3172, which found it while checking a section bound.

Expected

A finding entry's own summary is not a section heading, whatever words its title carries. One direction is to read a section heading only where it states a count, as sibling_heading and finding_entry already do, or to anchor SUPPRESSED the way PREVIOUSLY_MISSED is anchored.