PDF Redactor

A secure PDF redaction tool that permanently removes sensitive content instead of just covering it. Supports text search, regex patterns, OCR for scanned PDFs, and comprehensive metadata sanitization. Perfect for legal documents and confidential files.

What Makes This Different

• True Redaction: Content is permanently removed from the PDF structure, not just visually hidden
• Comprehensive Sanitization: Removes metadata, JavaScript, embedded files, and other potential information leaks
• Smart Detection: Supports exact text search, regex patterns, and manual rectangle specification
• OCR Support: Can handle scanned PDFs with optional OCR text detection
• Verification: Built-in verification to ensure redaction was successful
• No Web Dependencies: Runs completely offline for maximum security

Quick Start

Installation

# Clone or download this repository
cd pdfredact

# Install core dependencies
pip install -r requirements.txt

# Optional: Install OCR dependencies for scanned PDFs
pip install opencv-python pytesseract pillow

Basic Usage

# Make the CLI executable
chmod +x pdfredact/cli.py

# Redact by keywords and regex patterns
python pdfredact/cli.py input.pdf output.pdf \
  --term "John Q. Public" \
  --term "CONFIDENTIAL" \
  --regex "\b\d{3}-\d{2}-\d{4}\b" \
  --regex "[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Za-z]{2,}" \
  --verify --verbose

# Use pre-defined rectangles from JSON
python pdfredact/cli.py input.pdf output.pdf --rects rectangles.json

# Sanitize only (remove metadata without redacting content)
python pdfredact/cli.py sanitize input.pdf output.pdf

Features

Text-Based Redaction

• Exact Term Search: Find and redact specific words or phrases
• Regex Patterns: Use regular expressions for complex pattern matching
• Common PII Patterns: Built-in patterns for SSNs, emails, phone numbers, etc.

Manual Redaction

• Rectangle Specification: Define exact areas to redact using JSON coordinates
• Preview Mode: Generate preview PDFs showing what will be redacted

Scanned PDF Support

• OCR Integration: Detect text in scanned PDFs using Tesseract OCR
• Raster Redaction: Convert pages to images with redacted areas for maximum security
• Hybrid Processing: Automatically detect text vs. scanned pages

Security & Sanitization

• Metadata Removal: Clear document info, XMP metadata, and creation details
• JavaScript Removal: Remove potentially malicious JavaScript and actions
• Embedded File Removal: Remove attached files and embedded content
• Link Neutralization: Remove or disable external links
• Form Flattening: Remove interactive form fields
• Thumbnail Cleanup: Remove page thumbnails and piece info

Verification & Reporting

• Content Verification: Check output PDF to ensure target content was removed
• String Analysis: Optional strings command verification for hidden content
• Detailed Reports: Generate JSON reports of redaction activities
• Impact Analysis: Preview redaction impact before applying

Advanced Usage

Rectangle Specification

Create a JSON file with rectangle coordinates:

{
  "1": [
    {"x0": 72, "y0": 540, "x1": 320, "y1": 565},
    {"x0": 100, "y0": 100, "x1": 400, "y1": 130}
  ],
  "3": [
    {"x0": 100, "y0": 700, "x1": 420, "y1": 740}
  ]
}

Coordinates are in points (72 points = 1 inch) from the bottom-left origin.

Common Regex Patterns

The tool includes built-in patterns for common PII:

from pdfredact.utils import get_common_patterns

patterns = get_common_patterns()
# patterns["ssn"] = r"\b\d{3}-\d{2}-\d{4}\b"
# patterns["email"] = r"\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Za-z]{2,}\b"
# patterns["phone"] = r"\b\(?(\d{3})\)?[-.\s]?(\d{3})[-.\s]?(\d{4})\b"

OCR Configuration

For scanned PDFs, configure OCR settings:

from pdfredact.ocr import OCRConfig, detect_text_in_scanned_pdf

config = OCRConfig()
config.dpi = 300  # Higher DPI for better accuracy
config.confidence_threshold = 50  # Minimum confidence for text detection

boxes = detect_text_in_scanned_pdf("scanned.pdf", terms=["SECRET"], config=config)

Programmatic Usage

Use as a Python library:

from pdfredact import find_boxes, apply_boxes, hard_sanitize

# Find content to redact
boxes = find_boxes("input.pdf", terms=["CONFIDENTIAL"], regex_patterns=[r"\d{3}-\d{2}-\d{4}"])

# Apply redactions
apply_boxes("input.pdf", "temp.pdf", boxes)

# Sanitize and finalize
hard_sanitize("temp.pdf", "output.pdf")

Command Line Options

Main Redaction Command

python pdfredact/cli.py input.pdf output.pdf [OPTIONS]

Options:

• --term TEXT: Text terms to redact (can be used multiple times)
• --regex TEXT: Regex patterns to redact (can be used multiple times)
• --rects PATH: JSON file with rectangle coordinates
• --fill [black|white]: Fill color for redacted areas (default: black)
• --verify: Verify redaction by checking output for remaining terms
• --verbose, -v: Enable verbose output

Sanitization Only

python pdfredact/cli.py sanitize input.pdf output.pdf [OPTIONS]

Options:

• --verbose, -v: Enable verbose output

Project Structure

pdfredact/
├── cli.py          # Command-line interface
├── detect.py       # Text and pattern detection
├── redact.py       # Redaction application  
├── sanitize.py     # Metadata and content sanitization
├── ocr.py          # OCR support for scanned PDFs
├── utils.py        # Utility functions and validation
└── __init__.py     # Package initialization

Testing

Create Test Files

from pdfredact.utils import create_test_rectangles_json
create_test_rectangles_json("test_rects.json")

Analyze PDF Security

from pdfredact.sanitize import analyze_pdf_security

analysis = analyze_pdf_security("document.pdf")
print(f"Found {len(analysis['metadata_found'])} metadata items")
print(f"JavaScript present: {analysis['javascript_found']}")
print(f"Embedded files: {analysis['embedded_files_count']}")

Preview Impact

from pdfredact.utils import estimate_redaction_impact

impact = estimate_redaction_impact("document.pdf", ["SECRET"], [r"\d{3}-\d{2}-\d{4}"])
print(f"Would redact {impact['total_matches']} items across {impact['pages_affected']} pages")

Important Security Notes

1. Test Thoroughly: Always test redaction on sample documents before processing important files
2. Verify Results: Use the --verify flag and manually review output PDFs
3. Backup Originals: Keep secure backups of original documents
4. Multiple Passes: For highly sensitive content, consider multiple redaction passes with different patterns
5. Physical Security: Ensure both input and output files are handled securely
6. Memory Cleaning: The tool processes PDFs in memory - ensure your system memory is secure

Verification Methods

The tool provides multiple verification approaches:

1. Text Re-extraction: Re-extracts text from output PDF and searches for target terms
2. String Analysis: Uses system strings command to find hidden content
3. Visual Inspection: Generate preview PDFs to manually verify redaction areas
4. Binary Analysis: Optional deep inspection of PDF object structure

Supported PDF Types

• Text-based PDFs: Full support for searchable text redaction
• Scanned PDFs: OCR-based text detection (requires additional dependencies)
• Mixed Content: Automatic detection and appropriate handling
• Encrypted PDFs: Supported if password is available
• Form PDFs: Form fields are flattened and removed during sanitization

Processing Flow

1. Analysis: Determine PDF type (text-based vs. scanned)
2. Detection: Find target content using text search, regex, or OCR
3. Redaction: Apply redaction annotations and remove underlying content
4. Sanitization: Remove metadata, JavaScript, embedded files, etc.
5. Verification: Check output to ensure redaction was successful
6. Reporting: Generate detailed reports of actions taken

Contributing

This tool is designed to be secure and reliable. When contributing:

1. Ensure all changes maintain security principles
2. Add appropriate tests for new functionality
3. Update documentation for new features
4. Consider edge cases and error handling

License

This project is provided as-is for educational and security purposes. Users are responsible for compliance with applicable laws and regulations regarding document redaction and privacy.

---

Remember: True document security requires proper handling throughout the entire document lifecycle. This tool provides technical redaction capabilities, but organizational policies and procedures are equally important for maintaining information security.