Tags: Unstructured-IO/unstructured-api
Tags
perf: upgrade pdfminer-six to 20260107 (#536) Fixes ~15-18% performance regression introduced in 20251230 where f-strings were evaluated eagerly even when logging was disabled. See: pdfminer/pdfminer.six#1233 Fix: pdfminer/pdfminer.six#1234 <!-- CURSOR_SUMMARY --> --- > [!NOTE] > Addresses a pdf parsing performance issue by updating dependency and aligning versioning. > > - Upgrade `pdfminer-six` to `20260107` in `requirements/base.txt` and constrain it in `requirements/constraints.in` (perf fix) > - Bump version to `0.0.92` in `prepline_general/api/__version__.py` and `preprocessing-pipeline-family.yaml` > - Update `CHANGELOG.md` with the perf fix note > > <sup>Written by [Cursor Bugbot](https://cursor.com/dashboard?tab=bugbot) for commit cd8a7ef. This will update automatically on new commits. Configure [here](https://cursor.com/dashboard?tab=bugbot).</sup> <!-- /CURSOR_SUMMARY -->
Update Version and bump requirements (#528) This branch updates multiple dependencies to address open CVE (Common Vulnerabilities and Exposures) security vulnerabilities and brings the project up to date with the latest unstructured package release. Key Changes: - Version bump: 0.0.89 → 0.0.90 - Major dependency updates: - unstructured: 0.17.2 → 0.18.18 - torch: 2.7.1 → 2.9.0 - transformers: 4.52.4 → 4.57.1 - fastapi: 0.115.12 → 0.121.0 - cryptography: 45.0.4 → 46.0.3 - pillow: 11.3.0 → 12.0.0 - scipy: 1.15.3 → 1.16.3 - networkx: 3.4.2 → 3.5 Additional improvements: - Enhanced Makefile pip-compile workflow for better requirement management - Added constraints.txt for improved dependency resolution - Streamlined test requirements (reduced from 723 to much fewer lines) --------- Co-authored-by: Claude <[email protected]>
PreviousNext