anthonypdawson / vector-inspector

80%

Build Type

push

github

Committed by web-flow

Commit Message

Fix LanceDB flat schema metadata extraction and column width limits (#40)

* fix: LanceDB flat schema metadata extraction and column width limits

Fixes metadata display for LanceDB databases with flat schemas (e.g., Contextus)
where metadata is stored as individual columns rather than nested in a 'metadata' column.

## Changes

### LanceDB Connection
- Extract metadata from PyArrow schema first (all non-reserved columns)
- Support flat schema format in get_all_items() - build metadata dicts from column values
- Detect and exclude content column from metadata to prevent duplication
- Maintain backward compatibility with nested 'metadata' column format
- Fix count detection when schema extraction succeeds but count is unavailable

### UI
- Add 600px max-width constraint to table columns to prevent excessive width
- Prevent 'document' column from appearing twice (once as content, once as metadata)

### Tests
- Add comprehensive flat schema metadata extraction tests
- Update existing tests to reflect content column exclusion from metadata
- Verify backward compatibility with nested metadata format

Closes #XX

* docs: update CHANGELOG for v0.8.4

* ci: add PR checks for tests and linting

- Enable CI test workflow on all PRs to master/main
- Add new lint workflow with ruff (check + format)
- Linting is advisory-only (won't block merges) until codebase is cleaned up
- Tests remain blocking as expected

* chore: finalize v0.8.4 release prep

- Update RELEASE_REASON.md to v0.8.4 (matches CHANGELOG and pyproject.toml)
- Add permissions blocks to CI workflows per GitHub security recommendation

* chore: bump __version__ to 0.8.4 in __init__.py

Coverage Stats

73 of 88 new or added lines in 3 files covered. (82.95%)

4 existing lines in 1 file now uncovered.

15751 of 19601 relevant lines covered (80.36%)

0.8 hits per line