docx

The unpack.py script uses rglob to recursively process all XML files in the unpacked directory without any limits: 'xml_files = list(output_path.rglob("*.xml")) + list(output_path.rglob("*.rels"))'. A malicious Office file could contain thousands of XML files or deeply nested directory structures, causing resource exhaustion. The script processes every file found without size or count limits.

ooxml/scripts/unpack.py

xml_files = list(output_path.rglob('*.xml')) + list(output_path.rglob('*.rels'))
for xml_file in xml_files:
    content = xml_file.read_text(encoding='utf-8')
    dom = defusedxml.minidom.parseString(content)
    xml_file.write_bytes(dom.toprettyxml(indent='  ', encoding='ascii'))

The validation/redlining.py file uses subprocess.run with shell=False but constructs commands from file paths that could contain malicious characters. While shell=False prevents shell injection, the code doesn't validate that file paths are safe: 'subprocess.run(["pandoc", "--track-changes=all", str(original_docx), "-o", str(original_md)], capture_output=True, text=True, check=True)'. If file paths contain special characters or are manipulated, this could lead to unexpected behavior.

ooxml/scripts/validation/redlining.py

subprocess.run(["pandoc", "--track-changes=all", str(original_docx), "-o", str(original_md)], capture_output=True, text=True, check=True)