Helper repo to run the pdftable CLI on PDFs and export results to JSON.

It includes:

• A simple workflow to run pdftable over a PDF into an output folder (per-page PNG/PDF/HTML).
• export_pdftable_to_json.py – merges outputs into a single JSON file per run.
• Optional patch scripts we used to make the upstream code robust on CPU/Apple Silicon:
  • unify_torch_device_patch.py
  • patch_keyerror_tsr.py
  • fix_unboundlocal_tsr.py

This repo wraps the upstream pdftable project. Ensure the pdftable CLI is installed and on PATH.

Run the extractor (example):

pdftable --file_path_or_url "/path/to/your.pdf" \
  --output_dir "./cv_out" \
  --pages all \
  --lang en

Export to JSON:

python export_pdftable_to_json.py --outdir ./cv_out --outjson ./cv_out/results.json

The ./cv_out folder will contain page-*.{pdf,png,html} and model artifacts; results.json is a list of page objects with text blocks and (when present) table info.

License Apache License, Version 2.0 — see LICENSE.txt.