feat: muse#53 item 11 — symlog entries for muse revert
`muse revert` writes its own CommitRecord directly via write_commit() rather than going through muse commit's run() function, so it bypassed the _write_symlogs() call entirely — a revert that restored a prior symbol body left no trace in that symbol's journal, even though it already wrote a reflog entry for the branch ref move.
Fixes by calling the exact same _write_symlogs() helper commit.py uses, right after revert's own append_reflog() call, with the same non-fatal try/except wrapping (a symlog write failure must never fail a revert, matching commit.py's established pattern). --no-commit mode (stages changes without creating a commit) correctly does not call this — there's no commit for a symlog entry to attach to yet.
Covered by tests/test_revert_symlog.py (2 cases: a revert of a symbol-modifying commit produces a new symbol-modified entry in that symbol's journal; --no-commit mode does not write one). Confirmed genuine RED by reverting the fix file and re-running (the symlog-entry assertion failed, total stayed at 2 instead of 3; the --no-commit case passed incidentally since it asserts absence of a change), then GREEN with the fix restored plus the full revert + symlog regression suite (218 tests, zero regressions). Live-verified via muse-dev against a disposable /tmp sandbox: reverting a "double the total" commit produced a `symbol-modified: Revert "double"` symlog entry as expected.
Co-Authored-By: Claude Sonnet 5 <[email protected]>
Semantic Changes
21 symbols
Files Changed
+1
~2
1237 in snapshot
0 comments
muse hub commit comment sha256:46b7d9d1e17a4f87e7330dc6cfadab13241976587bc6b855cc2682667c95183e --body "your comment"
No comments yet. Be the first to start the discussion.