adopt-project.py: under a failing clean filter that a rule names for .gitattributes, check also prints plan-file-reencoded for a file the split stores as its blob, beside the right plan-attributes-reencoded (#1136) #1192

Activity

1. brettheap commented on Oct 23, 2026

   

   brettheap

   on Oct 23, 2026

   ContributorAuthor

   More actions

   (lane openRepoShape-1, 2026-10-23T08:22:26Z) A likely cause, measured by r1169b while re-reviewing #1169. See issuecomment-6360685539 on #1169.

   • The cause: the rehearsal picks the files it asks about with check-attr --cached (_SplitRehearsal.reached), and git 2.43 does not read attr.tree for --cached.
   • The test: with --cached dropped (r1169b's X1, a temporary change in a scratch clone), this issue's case gives only the right plan-attributes-reencoded. Head gives both findings.
   • So this issue likely shares adopt-project.py: check's rehearsal and _stored_as_written take a parent's -c settings, attr.tree too, and the source's objects, GIT_ATTR_SOURCE is refused by name before any question, and each throwaway reads its environment first (#1136) #1169's cause, more than the update-index --refresh this issue guessed at.

   Lane 1's next round on #1169 makes that change. This issue gets a row there, or stays open with what is left.

   Lane: openreposhape-1 (openRepoShape-1)
2. brettheap

   added a commit that references this issue on Oct 23, 2026

   checkpoint: adopt-project.py: the rehearsal's `check-attr` drops `--c…

   ...

   1e958ca
3. brettheap

   added this to the v1.2.1 milestone on Oct 23, 2026
4. brettheap commented on Oct 23, 2026

   

   brettheap

   on Oct 23, 2026

   ContributorAuthor

   More actions

   (lane openRepoShape-1, 2026-10-23T20:44:08Z) The cause is now traced, and the case is wider than this issue says. From r1169d's review of #1169 at 76c9b767 (#1169 issuecomment-6374351753, SHOULD-2), with git 2.43.

   • The trace (r1169c's expB): under the failing required filter, update-index -q --refresh over the rehearsal's TAKEN index exits 128 with fatal: .gitattributes: clean filter 'w1136' failed. The exit code is never read (adopt-project.py:13021 at 76c9b767). diff-files --name-only then exits 0 and lists .gitattributes specs/spec.md. With no filter, the refresh exits 0 and diff-files lists nothing.
   • Both rules in the side tree. The own root says *.png binary; the side tree says /.gitattributes filter=w1136 and *.md text, and attr.tree names it. In all three forms (global config, GIT_CONFIG_PARAMETERS, GIT_CONFIG_COUNT):
     • adopt-project.py: check's rehearsal and _stored_as_written take a parent's -c settings, attr.tree too, and the source's objects, GIT_ATTR_SOURCE is refused by name before any question, and each throwaway reads its environment first (#1136) #1169's head 76c9b767 gives exit 1, with plan-attributes-reencoded and the misnamed plan-file-reencoded for specs/001-feature/spec.md.
     • 25805102 gives plan-attributes-reencoded alone.
     • bba168bd says plan ok.
     • So adopt-project.py: check's rehearsal and _stored_as_written take a parent's -c settings, attr.tree too, and the source's objects, GIT_ATTR_SOURCE is refused by name before any question, and each throwaway reads its environment first (#1136) #1169's read without --cached brings this case into all three forms. It is not wrong behaviour: check refuses with the right first finding, where the base said plan ok.
   • Control: the tree's *.md text alone, with no filter, says plan ok on every tree. So spec.md is stored as its blob, and the second finding names a file that is not reencoded.
   • The fix this issue asks for: read the refresh's exit code, and where it fails, name only the filter's failure.

   Lane: openreposhape-1 (openRepoShape-1)