user-docs

#!/usr/bin/env python3
"""
PyMuPDF-based PDF parser for robust text extraction.
Fallback for when Docling misses content due to encoding issues.
"""

import sys
import json
import traceback

def parse_pdf(file_path: str) -> dict:
    """Parse PDF using PyMuPDF (fitz) for robust text extraction."""
    try:
        import fitz  # PyMuPDF
    except ImportError:
        return {
            "success": False,
            "error": "PyMuPDF not installed. Run: pip install pymupdf"
        }
    
    try:
        doc = fitz.open(file_path)
        
        # Extract text from all pages
        full_text = ""
        for page_num, page in enumerate(doc):
            text = page.get_text("text")  # Plain text extraction
            if text.strip():
                full_text += f"\n--- Page {page_num + 1} ---\n"
                full_text += text
        
        # Get metadata BEFORE closing document!
        metadata = doc.metadata or {}
        title = metadata.get("title", "")
        page_count = doc.page_count
        char_count = len(full_text)
        
        # Clean up
        doc.close()
        
        return {
            "success": True,
            "content": full_text.strip(),
            "metadata": {
                "title": title if title else None,
                "page_count": page_count,
                "char_count": char_count
            }
        }
        
    except Exception as e:
        return {
            "success": False,
            "error": f"PyMuPDF parsing failed: {str(e)}\n{traceback.format_exc()}"
        }


def main():
    if len(sys.argv) < 2:
        print(json.dumps({
            "success": False,
            "error": "Usage: pymupdf_parser.py <pdf_path>"
        }))
        sys.exit(1)
    
    file_path = sys.argv[1]
    result = parse_pdf(file_path)
    print(json.dumps(result, ensure_ascii=False))


if __name__ == "__main__":
    main()