extract-pdf

Overview

Extract content from one or more PDF files and convert to compliant markdown with YAML frontmatter. Supports OCR for scanned documents, password-protected files, template-based field extraction, and batch processing of entire folders.

Usage

python main.py extract-pdf --source SOURCE [OPTIONS]

Options

Option	Description	Default
--source	Path to PDF file or folder of PDFs (required)	--
--output	Output directory	_output
--title	Custom document title	From PDF metadata
--tags	Comma-separated tags for frontmatter	pdf,extracted
--category	Document category	KB Article
--no-images	Disable image extraction	false
--ocr / --no-ocr	Enable/disable OCR	Auto-detect
--ocr-language	OCR language code (e.g., eng, deu)	eng
--password	Password for encrypted PDF	--
--password-prompt	Prompt for password securely	false
--template	Extraction template YAML for field extraction	--
--template-repo	Template repository directory	--
--auto-classify	Auto-classify document and select template	false
--format	Output format: markdown, json	markdown
--aggregate	Aggregate batch results into single JSON	false

Prerequisites

• Repo: content-conductor
• Install: pip install -r requirements.txt from repo root
• OCR (optional): pip install -r requirements-ocr.txt + Tesseract engine

Examples

Extract a single PDF

python main.py extract-pdf --source document.pdf

Batch extract all PDFs in a folder

python main.py extract-pdf --source ./pdf_folder/ --output ./converted/

Force OCR on a scanned document

python main.py extract-pdf --source scanned.pdf --ocr

Template-based extraction with auto-classification

python main.py extract-pdf --source ./invoices/ --template-repo ./templates/ --auto-classify

Output

Creates a timestamped subdirectory with markdown file(s) and extracted images. Each file includes YAML frontmatter with title, tags, category, and extraction metadata.

Related Commands

• cc extract-web -- extract web pages to markdown
• cc extract-youtube-auto -- extract YouTube video data