kert 7dffb76801 add Pydantic field metadata, pipeline schema validation, col: tag namespace, and stack.toml config ...

- SQLTable: add column_meta(), field_descriptions(), to_ddl(), _resolve_type()
- generate_models.py: read DuckDB column comments, emit Field(description=...)
- run_pipeline: validate output columns against SQLTable contract (SchemaError)
- Pipeline.run: pass output class through to runner
- Tag.col(): new namespace for column-level descriptions
- bib/meta: collect_column_comments() and apply_column_comments() for Zotero→DuckDB flow
- stack.toml + src/conf/: centralised config loader with attribute access and path()
- 100% test coverage on all changed files

2026-03-12 13:03:44 -04:00

29 KiB

Raw Blame History

View Raw