Extract all embedded images from an (XML-format) Word document

wordextract.py

#!/usr/bin/env python3

# Exatrct all embedded images from a Word document in their full resolution.
# In Word: File, Save As, choose XML format
# Then run the resulting Word XML file through this script.

import sys
import base64
from xml.dom.minidom import parse


def main(fname):
    dom = parse(fname)
    for i, part in enumerate(dom.getElementsByTagName('pkg:part')):
        fname = part.getAttribute('pkg:name')
        ctype = part.getAttribute('pkg:contentType')
        print(f"\nPart {i+1}: {fname} ({ctype})")
        if not ctype.startswith('image/'):
            print('Skipping (not an image)')
        elif comp := part.getAttribute('pkg:compression') != 'store':
            print(f'Skipping (compressed: {comp})')
        else:
            fname = fname.lstrip('/').replace('/', '_')
            print(f'Extracting to {fname}')
            data = part.getElementsByTagName('pkg:binaryData')[0].firstChild.data
            with open(fname, 'wb') as f:
                f.write(base64.b64decode(data))

                
if __name__ == '__main__':
    main(*sys.argv[1:])