v0.4.4 v1.24.0
high risk
airweave-ai/airweaveSep 22, 2026 → Sep 23, 20261 day between releases
What's New
Enhanced PDF text extraction with color support and metadata fields
New Features
- Added fill_color, stroke_color, and render_mode properties to TextItem API
- Extended document metadata extraction to include Author, Subject, and Keywords
- Improved title decoding to support PDFDocEncoding and UTF-16LE
- Added extraction for text rendered using the '"' operator
- Implemented 5pt baseline spacing rule for stacked table headers
Improvements
- Improved text extraction for invisible text and clipping modes
- Improved ActualText span weight calculation based on the first glyph
- Enhanced handling of small type spacing to prevent word zipping
- Refined document metadata decoding logic
Bug Fixes
- Fixed text extraction for invisible text and clipping modes
- Fixed stacked table headers where words were incorrectly zipped
- Fixed ActualText span weight calculation
- Fixed title decoding for PDFDocEncoding and UTF-16LE
Upgrade Notes
This is a major version jump requiring careful review of integration code. The TextItem API now includes new color and render mode properties that must be handled by consuming applications. Updated metadata parsing may affect workflows relying on specific title fields.
The Two Releases
v1.24.0
1.24.0 — Text paint and document information
Sep 23, 2026· 29 days ago
Enhanced PDF text extraction by adding color information (fill/stroke) and document metadata fields to the TextItem API.
Full release notes