Fix score corruption after undoing local time signature deletion

[0xMashiro]

Resolves: #29789

Deleting a local time signature and undoing restores the visible signature but can leave the full-score staff time-signature map at the preceding meter. Saving then reports an incomplete measure and score corruption.

Register and unregister time signatures when Measure adds or removes a complete time-signature segment. Restoring a segment during undo does not pass through the individual-item registration in Segment::add, so this keeps both mutation paths consistent.

The regression uses the existing timeSig-11.mscz fixture: Clarinet measure 2 has local 7/8 and measure 3 has local 5/4. It covers deletion from the full score and linked part, undo, redo, integrity validation, and MSCZ save/reopen including excerpts.

Validation

• Unmodified official main a64e0d0c8725a89af2b4b9f8e058a75be20c01d3: the new regression fails. The restored object is 5/4, while the full-score staff lookup remains 7/8; the save integrity check reports incomplete measures.
• Candidate: all 16 native time signature tests pass (Engraving_TimesigTests.*).
• Linux desktop, Qt 6.10.2, GCC 14, Debug/ASan: agent-operated delete → undo → save reproduces the corruption warning before the fix and saves successfully after it.
• Official Uncrustify configuration and git diff --check pass.

Reproduction instructions, scores and native logs

Before video:

local-timesig-before.mp4

After video:

local-timesig-after.mp4

Related work

This addresses the existing report #29789, rather than claiming a newly discovered issue. #26100 provides related local-meter work and the reused fixture. #17418 concerns deleting a whole measure, and #11941 concerns an initial time signature. No matching active fix PR was found in the issue timeline and repository searches; this is not an exhaustive claim about all prior work.

Contributor checklist

• I signed the CLA. The contributor confirmed signing; the MuseScore.org username remains to be supplied.
• The title of the PR describes the problem it addresses.
• Each commit's message describes its purpose and effects, and references the issue it resolves. If changes are extensive, there is a sequence of easily reviewable commits.
• The code in the PR follows the coding rules.
• I understand all aspects of the code I'm contributing and I'm able to explain it if requested.
• The code compiles and runs on my machine, preferably after each commit individually. I have manually tested and verified that my changes fulfil their intended purpose. (Desktop verification was operated by the coding agent.)
• No prior attempts to resolve this problem exist, or if they do, I listed them in my PR description and described how I avoided repeating past mistakes. (Search scope and related work are described above.)
• There are no unnecessary changes.
• I created a unit test or vtest to verify the changes I made (if applicable).

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Advanced

Run ID: 6c87e503-decf-4b8a-acdb-ee94234346d8

📥 Commits

Reviewing files that changed from the base of the PR and between a64e0d0 and add1935.

📒 Files selected for processing (2)

• src/engraving/dom/measure.cpp
• src/engraving/tests/timesig_tests.cpp

Included review availability: Your plan provides up to 10 included reviews per hour; 9 remain after this review.

---

📝 Walkthrough

Walkthrough

Measure::add registers each TimeSig in an inserted time-signature segment with its staff. Measure::remove unregisters each TimeSig when the segment is removed. A new test covers local time-signature deletion from scores and excerpts, undo and redo behavior, and persistence after saving and reopening the score.

Priority: ➖ Normal — Impact reflects medium issue severity.

Severity of issue fixed: Medium

Merge Risk: ⚪ Minimal · up to add19

Local time signatures are now registered when restored and unregistered when removed, preventing stale timing state after deletion and undo. The covered score, excerpt, undo/redo, and persistence flows indicate the corruption fix is ready to merge.

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 50.00% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 4 functions across 2 files.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Linked Issues check	✅ Passed	The implementation directly addresses issue #29789 by registering and unregistering staff time signatures when complete time-signature segments are added or removed. The regression test covers deletio…
Out of Scope Changes check	✅ Passed	The changes are limited to the required Measure time-signature registration logic and focused regression coverage. No unrelated code or test changes are present.
Description check	✅ Passed	The description clearly explains the score-corruption bug, the implementation, regression coverage, validation results, related work, and checklist status. The CLA username is not supplied, but the de…
Title check	✅ Passed	The title clearly and concisely describes the primary change: fixing score corruption caused by undoing local time signature deletion.

• Fix all pre-merge checks with AI

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

Warning

Linked repositories: Public OSS repositories can only analyze public repositories installed in this organization. No linked repositories were analyzed; skipped musescore/muse_framework.git.

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}