pdfmd

Convert scanned PDF documents to structured Markdown with OCR and signature detection.

Features

• OCR Text Extraction: Converts scanned PDFs to text using Tesseract
• Structured Markdown: Detects headings, paragraphs, and lists
• Image Extraction: Extracts and deduplicates embedded images
• Signature Detection: Identifies signature regions and matches against expected signers
• Metadata Generation: Creates metadata.json with conversion stats
• Windows Support: Handles paths with spaces and special characters

Installation

Prerequisites

• Python 3.11+
• Tesseract OCR (see INSTALL.md for installation guide)

Install pdfmd

cd PDFtoMD
pip install -e .

Usage

Basic conversion

pdfmd convert document.pdf

Custom output path

pdfmd convert document.pdf -o ./output/document.md

With signature detection

pdfmd convert document.pdf --signers "Juan Pérez,María García"

Convert specific pages

pdfmd convert document.pdf --pages "0,1,2,3,4"

Show PDF info

pdfmd info document.pdf

Show version

pdfmd --version

Output Files

File	Description
document.md	Structured Markdown with OCR text
metadata.json	Conversion stats (pages, confidence, time)
signature_report.md	Signature detection report
signature_report.json	Signature data in JSON format
images/	Extracted images (if any)

CLI Options

Usage: pdfmd [OPTIONS] INPUT_PATH

Options:
  -o, --output PATH     Output Markdown file path
  -l, --lang TEXT        OCR language (default: spa)
  -d, --dpi INTEGER      Render DPI (default: 300)
  -s, --signers TEXT     Comma-separated list of expected signers
  -p, --pages TEXT       Pages to process (e.g., '0,1,2' or '0-5')
  -v, --version          Show version
  --help                 Show this message and exit

Development

Run tests

pytest tests/ -v

Run lint

ruff check pdfmd/

License

MIT