Skip to main content

​
Common Issues

  • PyMuPDF not installed — pip install pymupdf. The skill raises RuntimeError with install instructions if missing.
  • Password-protected PDFs — Not currently supported. Decrypt with qpdf --decrypt input.pdf output.pdf first.
  • Scanned-only PDFs (no text layer) — OCR fallback uses pytesseract if installed (pip install pytesseract). Without it, extraction returns [OCR not available].
  • Tables not detected — PyMuPDF’s find_tables() requires structured table layouts. Hand-drawn or image-based tables won’t extract. Consider exporting from the source app instead.
  • OCR disabled / slow — OCR is only triggered when a page has fewer than 20 characters of extractable text. For large scanned docs, expect slower extraction. Set page_start / page_end to limit scope.