fix: pdf parsing, pass as str instead of posix path (#4448)

* FIX: PyMuPDF for parsing, pass as str instead of posix path

* Switch back to pypdf due to licensing

* Allow specification of the number of threads

* [autofix.ci] apply automated fixes

* Update file.py

* Switch StrInput to IntInput

* ✨ (Document QA.json, Vector Store RAG.json): Update FileComponent class to include IntInput for concurrency_multithreading to control the number of workers for parallel processing. This enhancement allows for better control and optimization of processing tasks.

* ⬆️ (pyproject.toml): upgrade langchain-core dependency from version 0.3.10 to version 0.3.12 to ensure compatibility and leverage new features

* ⬆️ (pyproject.toml): update langchain-core dependency to allow versions greater than or equal to 0.3.12 for compatibility reasons

---------

Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
Co-authored-by: cristhianzl <cristhian.lousa@gmail.com>
Co-authored-by: Gabriel Luiz Freitas Almeida <gabriel@langflow.org>
This commit is contained in:
Eric Hare 2024-11-08 08:55:06 -08:00 • committed by GitHub
commit a4863bbe57
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
6 changed files with 70 additions and 59 deletions

View file

@ -135,7 +135,6 @@ dependencies = [
"duckdb>=1.0.0",
"python-docx>=1.1.0",
"jq>=1.7.0; sys_platform != 'win32'",
"pypdf>=4.2.0",
"nest-asyncio>=1.6.0",
"emoji>=2.12.0",
"cryptography>=42.0.5,<44.0.0",
@ -162,6 +161,7 @@ dependencies = [
"assemblyai>=0.33.0",
"fastapi-pagination>=0.12.29",
"defusedxml>=0.7.1",
"pypdf~=5.1.0",
]
[project.urls]