fix(schema): align to the tier4_perception_dataset document

[ktro2828]

What

There is a subtle difference in the definition of autolabel_metadata from the document of tier4_perception_dataset.

This pull request introduces a new AutolabelMetadata type to encapsulate metadata about models used for auto-labeling, replacing the previous usage of a raw list of AutolabelModel instances. The change is reflected consistently across the schema, codebase, and tests, improving clarity and extensibility of the auto-labeling metadata representation.

Schema and documentation updates:

• Introduced the AutolabelMetadata type, which wraps a list of AutolabelModel entries, and updated documentation to describe its structure and usage.
• Updated all schema tables (sample_annotation, sample_data, object_ann, surface_ann) to use autolabel_metadata: <option[AutolabelMetadata]> instead of a list of AutolabelModel. [1] [2] [3] [4]

Codebase changes:

• Added the AutolabelMetadata dataclass and its conversion logic in autolabel_metadata.py, and updated the AutolabelMixin and all relevant schema tables to use this new type. [1] [2] [3] [4]
• Improved docstrings and error handling for type conversion functions. [1] [2]

Test updates:

• Updated test fixtures and test cases to use the new AutolabelMetadata structure instead of a list of models. [1] [2] [3]
• Added new tests for AutolabelMetadata conversion and error handling, and updated AutolabelMixin tests to use the new type. [1] [2] [3]

These changes make the handling of auto-labeling metadata more robust and future-proof by introducing a clear, extensible structure.

[Copilot]

Pull request overview

This PR aligns the autolabel_metadata representation with the tier4_perception_dataset specification by introducing an AutolabelMetadata wrapper type. Previously, autolabel_metadata was represented as a raw list of AutolabelModel instances; now it's encapsulated in a structured AutolabelMetadata object containing a models field.

Changes:

• Introduced AutolabelMetadata dataclass with a models field to wrap lists of AutolabelModel instances
• Updated schema tables (SampleData, SampleAnnotation, ObjectAnn, SurfaceAnn) to use AutolabelMetadata instead of list[AutolabelModel]
• Updated documentation to describe the new AutolabelMetadata structure and its usage

Reviewed changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated no comments.

Show a summary per file

File	Description
t4_devkit/schema/tables/autolabel_metadata.py	Added AutolabelMetadata dataclass with conversion logic and updated AutolabelMixin to use the new type
t4_devkit/schema/tables/sample_data.py	Updated SampleData table to use AutolabelMetadata type for autolabel_metadata field
docs/schema/table.md	Added AutolabelMetadata documentation section and updated all table schemas to reference the new type
tests/schema/conftest.py	Updated test fixtures to use the new nested {"models": [...]} structure
tests/schema/tables/test_autolabel_metadata.py	Added TestAutolabelMetadata test class and updated existing tests to use AutolabelMetadata

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[github-actions]

☂️ Python Coverage

current status: ✅

Overall Coverage

Lines	Covered	Coverage	Threshold	Status
4139	3429	83%	50%	🟢

New Files

No new covered files...

Modified Files

File	Coverage	Status
t4_devkit/schema/tables/autolabel_metadata.py	95%	🟢
t4_devkit/schema/tables/sample_data.py	100%	🟢
TOTAL	97%	🟢

updated for commit: 8413caf by action🐍