Skip to content

Test the revision and paper checkers' branches - #44

Merged
jdkent merged 1 commit into
mainfrom
layouts-tests
Oct 18, 2026
Merged

jdkent merged 1 commit into
mainfrom
layouts-tests

Conversation

@jdkent

@jdkent jdkent commented Oct 17, 2026

Copy link
Copy Markdown
Member

Pins every check in check_revision, check_paper and check_run (layouts.py) with a test that builds the failing case from the example corpus and run, and asserts the exact message. Tests only; I found no bug in the checker.

Before this, removing the duplicate-key check (_check_keys) or the more-than-one-verdict check left all 210 tests green (coverage analysis M5, M6). They also covered none of: original's parse_id, different article, different text, add already in original, add missing from revision, verdict for a key not in the original, replaced_by not in the revision, accept/relabel key not surviving, new analysis neither added nor a replacement, check_paper wrong kind / schema version / article mismatch / different text / corpus holding a revision / missing text_path, and check_run with no record.

Mutation check: each of 25 conditions in those functions replaced by if False (or inverted/emptied), one at a time; each made at least one test fail, including the new test named for that check. None survived.

Suite: full run on beast in a fresh venv (.[test,generate,layouts,parquet]), all passing.

🤖 Generated with Claude Code

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
@jdkent
jdkent marked this pull request as ready for review October 17, 2026 23:54
@jdkent jdkent self-assigned this Oct 17, 2026
@jdkent
jdkent merged commit 2b19144 into main Oct 18, 2026
4 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.

2 participants