Skip to content

Create layer_2_semantic collection and warn on unknown vector collect… - #265

Merged
pradeeban merged 1 commit into
KathiraveluLab:mainfrom
saeed-taj:fix/layer_2_semantic_collection
Oct 18, 2026
Merged

pradeeban merged 1 commit into
KathiraveluLab:mainfrom
saeed-taj:fix/layer_2_semantic_collection

Conversation

@saeed-taj

Copy link
Copy Markdown
Contributor

Fixes #247

  • VectorStore now creates the layer_2_semantic collection (Chroma name pipeline_layer_2_semantic), so the location embedding that ingestion/routes.py already pushes there is actually stored.
  • store_vector() now logs a warning when the collection name isn't registered. It still returns False, so callers behave as before. A missing client still returns False quietly.

Notes

  • The collection name was kept as layer_2_semantic, as discussed on the issue.
  • Backfilling embeddings for existing posts is left out of this PR.

Tests

New tests/test_vector_store.py, using a temporary persist directory so it doesn't touch ./data/chroma_db:

  • the layer_2_semantic collection is created
  • a location vector is stored and store_vector returns True
  • an unknown collection name logs a warning and returns False

@madhavzanwar

Copy link
Copy Markdown
Contributor

Thanks for working on this. I fetched the branch and ran it: the 3 new tests pass, the full suite passes (352 passed, 1 skipped), and the CI flake8 gate is clean. It matches what I tried on my side.

Small things:

  • In test_location_vector_is_stored the metadata key is location_test, but the route uses location_text.
  • "layer_2_semantic" : has a space before the colon, and tests/test_vector_store.py has no newline at the end of the file.

@pradeeban
pradeeban merged commit e4f2400 into KathiraveluLab:main Oct 18, 2026
1 of 3 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Location embeddings are not stored: layer_2_semantic collection is never created in VectorStore

3 participants