Hi, A follow-up to my previous mail, where I said I could not run upstream's test.py. I have now run it (test/test.py from the 2.5.4 tag) against unmodified 2.5.4 and against 2.5.4 with the attached PR #742 lib patch:
2.5.4 (unpatched): Ran 240 tests, FAILED (14 failures) 2.5.4 + PR #742: Ran 240 tests, FAILED (14 failures) The failure count is the same, so these 14 look pre-existing or environment-related. I only captured the summary, so I have not confirmed that the two runs fail exactly the same 14 tests. For reference, PR #742 also adds a second ReDoS case (issue_metadata_list_redos: "items:\n - " followed by 32000 spaces). Its test_redos.py hunk does not apply to 2.5.4, as I noted before. Regards
