visual-parser 2.2.4__py3-none-any.whl → 2.2.5__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (25) hide show
  1. visual_parser/__init__.py +2 -2
  2. visual_parser/__main__.py +8 -8
  3. visual_parser/cli.py +1 -1
  4. visual_parser/cli_main.py +328 -328
  5. visual_parser/config.py +1 -1
  6. visual_parser/figure_describer.py +350 -353
  7. visual_parser/image_describer.py +333 -333
  8. visual_parser/image_tracker.py +68 -68
  9. visual_parser/kb_redo.py +153 -153
  10. visual_parser/metadata_extractor.py +10 -10
  11. visual_parser/model_catalog.py +190 -190
  12. visual_parser/nougat_engine.py +73 -11
  13. visual_parser/ollama_local.py +349 -349
  14. visual_parser/openai_gateway.py +84 -84
  15. visual_parser/pdf_engine.py +164 -0
  16. visual_parser/pipeline.py +5 -2
  17. visual_parser/text_extractor.py +45 -56
  18. visual_parser/vision_llm.py +427 -427
  19. {visual_parser-2.2.4.dist-info → visual_parser-2.2.5.dist-info}/METADATA +27 -11
  20. visual_parser-2.2.5.dist-info/RECORD +27 -0
  21. {visual_parser-2.2.4.dist-info → visual_parser-2.2.5.dist-info}/WHEEL +1 -1
  22. {visual_parser-2.2.4.dist-info → visual_parser-2.2.5.dist-info}/licenses/LICENSE +201 -201
  23. visual_parser-2.2.4.dist-info/RECORD +0 -26
  24. {visual_parser-2.2.4.dist-info → visual_parser-2.2.5.dist-info}/entry_points.txt +0 -0
  25. {visual_parser-2.2.4.dist-info → visual_parser-2.2.5.dist-info}/top_level.txt +0 -0
visual_parser/__init__.py CHANGED
@@ -2,7 +2,7 @@
2
2
  visual_parser — Standalone Visual-RAG PDF & Image Parser
3
3
  =========================================================
4
4
  Detects new PDFs (and standalone images alongside them) in a user-supplied
5
- directory, extracts text (via Nougat or lightweight PyMuPDF/PyPDFLoader),
5
+ directory, extracts text (via Nougat or lightweight pypdfium2/PyPDFLoader),
6
6
  describes every figure/chart/schematic and every standalone image using a
7
7
  Vision LLM (OpenAI GPT, Google Gemini, or a local model via Ollama), and
8
8
  writes JSONL knowledge bases ready for any downstream RAG system:
@@ -19,4 +19,4 @@ from visual_parser.config import ParserConfig
19
19
  from visual_parser.pipeline import run_pipeline
20
20
 
21
21
  __all__ = ["ParserConfig", "run_pipeline"]
22
- __version__ = "2.2.4"
22
+ __version__ = "2.2.5"
visual_parser/__main__.py CHANGED
@@ -1,8 +1,8 @@
1
- """
2
- Enables: python -m visual_parser --input-dir ./my_pdfs ...
3
- """
4
- import sys
5
- from visual_parser.cli_main import main
6
-
7
- if __name__ == "__main__":
8
- sys.exit(main())
1
+ """
2
+ Enables: python -m visual_parser --input-dir ./my_pdfs ...
3
+ """
4
+ import sys
5
+ from visual_parser.cli_main import main
6
+
7
+ if __name__ == "__main__":
8
+ sys.exit(main())
visual_parser/cli.py CHANGED
@@ -88,7 +88,7 @@ def _build_arg_parser() -> argparse.ArgumentParser:
88
88
  default="nougat",
89
89
  help=(
90
90
  "nougat — Nougat OCR model (best for scanned/complex PDFs, GPU recommended).\n"
91
- "lightweight — PyMuPDF text layer + PyPDFLoader fallback (fast, no GPU needed)."
91
+ "lightweight — PDF text layer (pypdfium2) + PyPDFLoader fallback (fast, no GPU needed)."
92
92
  ),
93
93
  )
94
94
  text_group.add_argument(