What's Changed
- fix(python): deduplicate CLI failure output by @hyunhee-jo in #495
- Add check for PUA in Alt entry by @LonelyMidoriya in #491
- chore(deps): bump urllib3 to 2.7.0 and python-multipart to 0.0.28 for security advisories by @hyunhee-jo in #498
- fix(cli): emit error for non-PDF top-level input by @hyunhee-jo in #496
- Tagged PDF - Add id to content items by @LonelyMidoriya in #501
- fix(hybrid)!: fail fast when backend left pages unprocessed with fallback disabled by @hyunhee-jo in #499
- fix(hybrid): forward --picture-description-prompt to docling VLM by @bundolee in #503
- fix(cli): announce when a folder contains zero processable PDFs (PDFDLOSP-15) by @bundolee in #507
- fix(cli): replace verapdf password exception stack trace with friendly error by @bundolee in #504
- fix: clean error for .pdf-named non-PDF input (PDFDLOSP-14) by @bundolee in #506
- fix(cli): separate markdown modifiers from --format values (PDFDLOSP-6) by @bundolee in #508
- Tagged PDF - Set ID for artifacts by @LonelyMidoriya in #509
- Add new rules for text sanitization by @LonelyMidoriya in #510
- docs(use-struct-tree): clarify that output quality depends on tag quality by @bundolee in #512
- fix(processors): log actual page count being processed, not document total by @bundolee in #513
- fix(hybrid): activate --hybrid-fallback on server-absent path (PDFDLOSP-21) by @hnc-jglee in #511
- fix(tagged-pdf): refuse encrypted documents with friendly message by @bundolee in #514
- fix(generators): emit page separator only for pages selected by --pages by @hyunhee-jo in #516
- Update verapdf version by @MaximPlusov in #517
Full Changelog: v2.4.3...v2.4.4