Fix fMP4 zero-duration samples

[kixelated]

Summary

• Treat explicit zero fMP4 sample durations as unknown during decode and import.
• Strip explicit zero duration fields when republishing fMP4 passthrough fragments.
• Fill missing fMP4 export durations from the successor frame when PTS is monotonic, otherwise fall back to the default frame duration.

Tests

• cargo test -p moq-mux
• cargo test -p moq-hls
• cargo test -p moq-mux container::fmp4 -- --nocapture
• git diff --check

(Written by Codex)

[sourcery-ai]

Hey - I've left some high level feedback:

• The infer_missing_durations/pts_monotonic logic disables PTS-based inference for the whole fragment if any timestamp pair is out of order; consider scoping monotonicity to local neighbors so well-behaved regions still benefit from PTS-derived durations.
• In the fMP4 import path, treating zero durations as None and erroring on any non-final sample with missing duration (MissingSampleDuration) changes behavior; if certain inputs intentionally use zero-length samples mid-fragment, it may be useful to handle them explicitly or surface a more descriptive error to distinguish malformed from intentional gaps.

Prompt for AI Agents

Please address the comments from this code review:

## Overall Comments
- The `infer_missing_durations`/`pts_monotonic` logic disables PTS-based inference for the whole fragment if any timestamp pair is out of order; consider scoping monotonicity to local neighbors so well-behaved regions still benefit from PTS-derived durations.
- In the fMP4 import path, treating zero durations as `None` and erroring on any non-final sample with missing duration (`MissingSampleDuration`) changes behavior; if certain inputs intentionally use zero-length samples mid-fragment, it may be useful to handle them explicitly or surface a more descriptive error to distinguish malformed from intentional gaps.

---

Sourcery is free for open source - if you like our reviews please consider sharing them ✨

• X
• Mastodon
• LinkedIn
• Facebook

_{Help me be more useful! Please click 👍 or 👎 on each comment and I'll use the feedback to improve your reviews.}

[coderabbitai]

Walkthrough

The pull request adds missing/zero sample duration inference to the fMP4 pipeline in rs/moq-mux: mod.rs now treats a resolved duration of 0 as None; import.rs validates that only the final sample in a track fragment may have a missing duration and normalizes zero fields before re-encoding; export.rs gains an infer_missing_durations helper that fills missing durations from PTS deltas (using an optional successor frame) and updates emit_fragment accordingly, with new unit tests. Separately, all justfiles are updated from bare set fallback to explicit set fallback := true, with minor whitespace and documentation formatting edits throughout the repository.

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 78.95% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title is concise and directly matches the main fMP4 zero-duration sample handling change.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Description check	✅ Passed	The description accurately matches the fMP4 zero-duration handling and export duration inference changes in the pull request.

✨ Finishing Touches

✨ Simplify code

• Create PR with simplified code
• Commit simplified code in branch codex/fmp4-zero-duration-vod

---

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.}

[coderabbitai]

Actionable comments posted: 2

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@js/watch/README.md`:
- Line 73: The Markdown table row for the `visible` option is being split into
too many columns because the pipe characters inside the value list are
interpreted as delimiters. Update the `visible` row in the README table so the
values are escaped or otherwise rendered as a single cell, keeping the table
consistent with the 4-column layout used by the surrounding rows.

In `@rs/moq-mux/src/container/fmp4/import.rs`:
- Around line 569-579: The duration normalization in the fMP4 import path is
dropping explicit zero values in a way that can fall back to inherited non-zero
defaults. Update the logic in the import routine that mutates
`traf_mut.tfhd.default_sample_duration` and each `trun_mut.entries` duration so
zero is only cleared when no non-zero `trex`/`tfhd` default can apply, or
resolve inherited durations first before removing the override. Keep the fix
localized to the import flow that handles `traf_mut`, `tfhd`, and `trun_mut`.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Run ID: 4a54ca79-3fc5-485c-862f-750b7fc2d0d2

📥 Commits

Reviewing files that changed from the base of the PR and between 02273a0 and 6048dc2.

📒 Files selected for processing (27)

• .github/justfile
• demo/boy/justfile
• demo/justfile
• demo/pub/justfile
• demo/relay/justfile
• demo/sub/justfile
• demo/web/justfile
• doc/bin/cli.md
• flake.nix
• go/justfile
• infra/apt/justfile
• infra/justfile
• infra/rpm/justfile
• js/CLAUDE.md
• js/justfile
• js/watch/README.md
• justfile
• kt/justfile
• nix/modules/moq-relay.nix
• py/justfile
• rs/CLAUDE.md
• rs/justfile
• rs/moq-mux/src/container/fmp4/export.rs
• rs/moq-mux/src/container/fmp4/import.rs
• rs/moq-mux/src/container/fmp4/mod.rs
• swift/justfile
• test/justfile

[coderabbitai]

🗄️ Data Integrity & Integration | 🟠 Major | 🏗️ Heavy lift

Avoid letting defaults replace an explicit zero duration.

Clearing tfhd.default_sample_duration == Some(0) or entry.duration == Some(0) can expose a non-zero fallback default from trex/tfhd, so a sample that import treated as “missing duration” may be republished with a real duration. Only strip the zero when no non-zero default can apply, or materialize inherited durations before removing the zero override.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@rs/moq-mux/src/container/fmp4/import.rs` around lines 569 - 579, The duration
normalization in the fMP4 import path is dropping explicit zero values in a way
that can fall back to inherited non-zero defaults. Update the logic in the
import routine that mutates `traf_mut.tfhd.default_sample_duration` and each
`trun_mut.entries` duration so zero is only cleared when no non-zero
`trex`/`tfhd` default can apply, or resolve inherited durations first before
removing the override. Keep the fix localized to the import flow that handles
`traf_mut`, `tfhd`, and `trun_mut`.