Changes
New features:
- A PPTX (MS PowerPoint) document parser and loader has been added.
Changes:
- The chromadb loader has been re-configured and re-built to support PDF
and PPTX loading. *This will break backwards compatibility*
- Some lower-level utilities have been moved to the libs/utilities module.
Dependency requirements:
- Updated the installation requirements to loosen the install burden.
- For example, the PPTX parser does not need the PDF parser's supporting
libraries are not and visa-versa. The parser-specific libraries can be
installed by the user on use.
- A user-friendly message has been added if a parser is imported where
the supporting libraries are not yet installed.
Fixes:
- The chroma Document object creation has been updated to only add unique
documents, addressing the issue of duplicated documents causing the
loader to fall over.
Documentation:
- The documentation suite has been added.
Released: 2025-02-12