docs-pdf（文档 PDF）

PDF Document Parsing

Parse PDF documents into markdown, text, and structured JSON using multi-method extraction.

Usage

Run the parsing script directly:

./scripts/parse_pdf.py <path_to_file.pdf> <output_dir>

Example:

./scripts/parse_pdf.py ~/documents/manual.pdf ./parsed/

The script uses 4 extraction methods:

• pypdf - Basic text extraction with page markers
• pdfminer - Detailed layout preservation
• pdfplumber - Table extraction and structure
• markitdown - Microsoft's markdown converter

Output Structure

output_dir/
├── file.pdf/
│   ├── parsing_summary.json
│   ├── pypdf/
│   │   └── content.md
│   ├── pdfminer/
│   │   └── content.txt
│   ├── pdfplumber/
│   │   ├── content.md
│   │   └── tables.json
│   └── markitdown/
│       └── content.md

Script Features

• Handles text-heavy and table-heavy PDFs
• Preserves layout information where possible
• Extracts tables as structured JSON
• Provides multiple format options (md, txt, json)
• Continues on errors (one method failure doesn't stop others)

Method Selection

• markitdown - Best for AI understanding (continuous markdown, no page breaks)
• pdfplumber - Best for documents with complex tables
• pypdf - Fast fallback for simple text extraction
• pdfminer - Best when layout preservation is critical