Add latest channel support in schema_tracker.

[shaosu-nvidia]

• Enable pedals to write the latest channel

Summary by CodeRabbit

• New Features
  • Enhanced MCAP recording to support optional latest-value tracking for pedal data, enabling efficient access to the most recent pedal data during playback.

[coderabbitai]

📝 Walkthrough

Walkthrough

This pull request adds optional support for writing an additional "latest-only" MCAP sample per update() call. The SchemaTracker class constructor now accepts an optional mcap_channel_latest_index parameter. When set, each sample processed during update() will be written to both the existing channel and the new latest channel. The PedalRecordingTraits compile-time metadata is updated to define the new "pedals_latest" channel, and the pedal tracker implementation is modified to pass the corresponding channel index to the schema tracker.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~12 minutes

🚥 Pre-merge checks | ✅ 2 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 66.67% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately describes the main change: adding support for a latest channel in the schema_tracker component, which is the core purpose of this PR.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch ssx/replay/latest_channel

---

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

[coderabbitai]

Caution

Some comments are outside the diff and can’t be posted inline due to platform limitations.

⚠️ Outside diff range comments (1)

src/core/live_trackers/cpp/inc/live_trackers/schema_tracker.hpp (1)

84-111: ⚠️ Potential issue | 🟠 Major

Guard latest-channel write on a valid decoded sample from this frame.

Line 84 initializes last_timestamp to a default value, and Line 108 only checks pointer presence. If all current samples fail GetRoot (Lines 88-91), this can write stale out_latest with an invalid timestamp.

Proposed fix

-        DeviceDataTimestamp last_timestamp{};
+        std::optional<DeviceDataTimestamp> last_timestamp;
         for (const auto& sample : samples_)
         {
             auto fb = flatbuffers::GetRoot<DataTableT>(sample.buffer.data());
             if (!fb)
             {
                 continue;
             }
@@
-            last_timestamp = sample.timestamp;
+            last_timestamp = sample.timestamp;
@@
-        if (mcap_channel_latest_index_ && mcap_channels_ && out_latest)
+        if (mcap_channel_latest_index_ && mcap_channels_ && out_latest && last_timestamp.has_value())
         {
-            mcap_channels_->write(*mcap_channel_latest_index_, last_timestamp, out_latest);
+            mcap_channels_->write(*mcap_channel_latest_index_, *last_timestamp, out_latest);
         }

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@src/core/live_trackers/cpp/inc/live_trackers/schema_tracker.hpp` around lines
84 - 111, The loop may call mcap_channels_->write for the "latest" channel even
when every sample failed decoding, causing a stale out_latest/last_timestamp to
be written; introduce a validity flag (e.g., bool have_decoded = false) set to
true only after a successful GetRoot/UnPackTo and update last_timestamp there,
then change the final guard to require have_decoded &&
mcap_channel_latest_index_ && mcap_channels_ && out_latest before calling
mcap_channels_->write; refer to symbols samples_, GetRoot<DataTableT>,
UnPackTo(out_latest.get()), out_latest, last_timestamp,
mcap_channel_latest_index_, and mcap_channels_->write.

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Outside diff comments:
In `@src/core/live_trackers/cpp/inc/live_trackers/schema_tracker.hpp`:
- Around line 84-111: The loop may call mcap_channels_->write for the "latest"
channel even when every sample failed decoding, causing a stale
out_latest/last_timestamp to be written; introduce a validity flag (e.g., bool
have_decoded = false) set to true only after a successful GetRoot/UnPackTo and
update last_timestamp there, then change the final guard to require have_decoded
&& mcap_channel_latest_index_ && mcap_channels_ && out_latest before calling
mcap_channels_->write; refer to symbols samples_, GetRoot<DataTableT>,
UnPackTo(out_latest.get()), out_latest, last_timestamp,
mcap_channel_latest_index_, and mcap_channels_->write.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro Plus

Run ID: 0b6bbe72-229e-4bea-be5b-88455c2c2a06

📥 Commits

Reviewing files that changed from the base of the PR and between af72817 and 8d0d0e9.

📒 Files selected for processing (3)

• src/core/live_trackers/cpp/inc/live_trackers/schema_tracker.hpp
• src/core/live_trackers/cpp/live_generic_3axis_pedal_tracker_impl.cpp
• src/core/mcap/cpp/inc/mcap/recording_traits.hpp