Package `ragindexer`

RAG Indexer

Usage

To run tests, just run:

pytest

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog and this project adheres to Semantic Versioning.

Added publication of qdrant snapshot on github pages (e54301c by Yann de The).
Added link to doc pages (080304e by Yann de The).
Added github pages publication (3fb26fb by Yann de The).
Added MIN_EXPECTED_CHAR env var (c5011fc by Yann de The).
Added test documents (9dc3019 by Yann de The).
Added logging msg to tell if OCR cache is reused (168a97d by Yann de The).
Added emails watcher (9de852c by Yann de The).
Added emails observer (511da4a by Yann de The).
Added API keys and remove of deprecated method QdrantClient.search (12d727f by Yann de The).
Added page progress (fd78e72 by Yann de The).
Added new_ocr for later integration (c56d9d7 by Yann de The).
Added a cache for OCR (c4ce7d9 by Yann de The).
Added OCR cache (8dc4205 by Yann de The).
Added unit test to check embedding relevance (d9e072f by Yann de The).
Added progress bars for file analysis. pdf tested OK (be49834 by Yann de The).
Added sqlite db to keep track of the indexed files (8f5c67f by Yann de The).

Fixed Exception raised when no text read (88c078f by Yann de The).
Fixed empty chunks (a72e49a by Yann de The).
Fixed pid computation (e3840c1 by Yann de The).
Fixed OCR cache determination (91dbc4d by Yann de The).
Fixed qdrant hostname (def96ad by Yann de The).
Fixed NLTK downloads (10e1dae by Yann de The).
Fixed bug where a whole pdf file is skipped if one page has no text (301e5eb by Yann de The).
Fixed image name (eadcf30 by Yann de The).
Fixed pdf error (cff80bb by Yann de The).
Fixed action (694bde7 by Yann de The).
Fixed docker compose stack (814adf6 by Yann de The).
Fixing pipeline (00107b2 by Yann de The).