mia 9d3d80b9f8 fix: extract text from PDF URLs in fetch_url tool ... When fetch_url fetches a PDF URL, the content was returned as garbled binary because BeautifulSoup's HTML parser corrupted the PDF bytes. Add three-layer PDF detection: - Fast-path for URLs ending in .pdf (case-insensitive) - Content-Type header detection in async SafeWebBaseLoader._fetch - %PDF binary content fallback in get_content_from_url Includes SSRF protection via validate_url(), streaming download with Content-Length pre-check, 50 MB size limit, and proper session management. No new dependencies (pypdf already in requirements.txt). Closes #23841 Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>		2026-04-19 11:07:02 +05:30
..
data	refac: mv backend files to /open_webui dir	2024-09-04 16:54:48 +02:00
open_webui	fix: extract text from PDF URLs in fetch_url tool	2026-04-19 11:07:02 +05:30
tests/retrieval	fix: extract text from PDF URLs in fetch_url tool	2026-04-19 11:07:02 +05:30
.dockerignore	fix: litellm config issue	2024-02-24 22:35:11 -08:00
.gitignore	refac	2024-09-06 04:59:20 +02:00
dev.sh	refac	2026-03-24 19:43:30 -05:00
requirements-min.txt	refac	2026-04-13 23:40:09 -05:00
requirements.txt	refac	2026-04-13 23:40:09 -05:00
start.sh	refac	2026-03-24 19:43:30 -05:00
start_windows.bat	refac	2026-03-24 19:43:30 -05:00