Fix trimmed ranges of deeply nested children

[jhuanxx44]

Link the Issue(s) this Pull Request is related to.

Fixes #2043

Summarize your change.

Composition::trimmed_range_of_child fails with NOT_A_CHILD_OF (exposed as NotAChildError in Python) when it traverses three or more parent compositions. The first iteration advances current, but subsequent iterations leave it pointing to an earlier descendant, so the next parent cannot find it among its direct children.

Advance current after accumulating each parent range, matching the traversal in range_of_child.

Reference associated tests.

• C++ regression: a Stack → Track → Stack → Clip hierarchy with a nonzero track offset; checks both the error status and the returned range.
• Python regression: one through four parent compositions, accumulated track offsets, and an outer source range that either clips the result or excludes the clip entirely.
• Both regressions fail before the production change and pass after it.

Validated locally on macOS arm64 with AppleClang 21 and Python 3.14.6:

• C++ static Debug build: all 10 CTest targets pass.
• make test: 482 tests run, 2 existing skips.
• make lint, make manifest, clang-format --dry-run --Werror on the changed C++ files, and git diff --check pass.

Assisted-by: Codex / GPT-6

[linux-foundation-easycla]

The committers listed above are authorized under a signed CLA.

• ✅ login: jhuanxx44 / name: jinghuan (0c65eec)

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 83.52%. Comparing base (a639246) to head (65cc72d).

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2048      +/-   ##
==========================================
+ Coverage   83.48%   83.52%   +0.04%     
==========================================
  Files         182      182              
  Lines       13517    13533      +16     
  Branches     1254     1255       +1     
==========================================
+ Hits        11284    11303      +19     
+ Misses       2060     2057       -3     
  Partials      173      173              

Flag	Coverage Δ
py-unittests	83.52% <100.00%> (+0.04%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
src/opentimelineio/composition.cpp	80.32% <100.00%> (+1.05%)	⬆️
tests/test_composition.py	99.72% <100.00%> (+<0.01%)	⬆️

---

Continue to review full report in Codecov by Harness.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update a639246...65cc72d. Read the comment docs.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[jminor]

Hi @jhuanxx44 - thanks, this PR is looking pretty good. We'll arrange for an OTIO committer to review it. Out of curiosity, how did you come across this issue, and what lead you to contribute to OTIO?

[jhuanxx44]

Hi @jminor, thanks for taking a look!

I’m exploring a video editing project of my own and will need a timeline model, which is how I came across OTIO. Around the same time, I saw #2043, where you were looking for someone to verify a bug identified by an LLM-based review, and it seemed like something I could help with.

To be upfront, I used Codex to help investigate and implement the change, hence the Assisted-by: trailer in the PR. I reviewed the diff myself and ran the full local test and lint suite; the new tests failed before the fix and passed afterward.

I’m happy to answer any questions from the committer and follow whatever conventions OTIO prefers for AI-assisted contributions.

[jminor]

We're actually right in the midst of establishing our AI policy, and this PR matches very closely to what we're aiming for. Your timing is perfect. See details here: #2046