evo-pdf-redaction-engine

Performs physical PDF redaction using PyMuPDF (fitz). Applies redaction annotations, removes hyperlinks, purges metadata, and saves securely with garbage collection.

Key Functions

• redact_exact_strings(page, strings) - Add redaction annotations for exact string matches
• redact_regex_patterns(page, patterns) - Add redaction annotations for regex matches
• redact_page_region(page, rect) - Redact an entire rectangular region
• scrub_hyperlinks(page, link_patterns) - Remove identifying hyperlinks and redact their text
• scrub_metadata(doc) - Purge standard and XMP metadata
• anonymize_pdf(input_path, output_path, exact_strings, regex_patterns, link_patterns) - Full pipeline
• save_redacted_document(doc, output_path) - Save with garbage=4 and deflate=True

Usage

import sys
sys.path.insert(0, '/app/environment/skills/evo-pdf-redaction-engine/scripts')
from redaction import anonymize_pdf

anonymize_pdf(
    input_path='/root/paper1.pdf',
    output_path='/root/redacted/paper1.pdf',
    exact_strings=['Author Name', 'University Name'],
    regex_patterns=[r'arXiv:\d{4}\.\d{4,5}'],
    link_patterns=['arxiv.org', 'github.com']
)