Anonymize pdf annotations and metadata

anonymize-pdf.py

#!/usr/bin/env python
try:
    import fitz
except ImportError as e:
    import subprocess
    print('Trying to install pip install PyMuPDF')
    subprocess.call("pip install PyMuPDF".split())
    print('Try pip install PyMuPDF')
    raise
import sys
if len(sys.argv) < 2:
    raise ValueError("Please provide a pdf to anonymize")
if len(sys.argv) < 3:
    outfilename = filename.replace('.pdf', '.anon.pdf')
else:
    outfilename = sys.argv[2]
filename = sys.argv[1]
doc = fitz.open(filename)
metadata = doc.metadata
for k, v in metadata.items():
    if k not in ['format']: # retain some metadata
        metadata[k] = ''
doc.set_metadata(metadata)

FIELDS_TO_KEEP = ['content']
FIELD_DEFAULTS = { 'modDate' : "D:20000000000000-00'00'",
                  'other' : 'X'}
for page in doc:
    for annot in page.annots():
        info = annot.info
        # print("Before", info)
        for k, v in info.items():
            # remove all annotation info 
            # other than the content
            if k not in FIELDS_TO_KEEP:
                info[k] = FIELD_DEFAULTS.get(
                        k, FIELD_DEFAULTS['other'])
        # print("After", info)
        annot.set_info(info)
        annot.update()
print("Printing annotations. Check for retained info in the pdf.")
for page in doc:
    for annot in page.annots():
        print(annot.info)


doc.save(outfilename)

LICENSE

MIT License

Copyright (c) 2024 Vikas Dhiman

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.

On Linux (Fedora) script is not working as expected.

• both the packages frontend and fritz had to be installed manually.
• the code does remove the names from annotations made inside a document viewer pdf

I stepped across this gist, since I had the identical problem as the op in https://softwarerecs.stackexchange.com/questions/31971/pdf-anonymizer-remove-potentially-identifying-metadata and indeed I wanted to remove author information from all annotations, so I ended in this gist.

I can reproduce the first problem of @jellepoland and found more bugs in the code. However, I don't know what @jellepoland means with his second bullet. Could you elaborate on this problem?

• the code does remove the names from annotations made inside a document viewer pdf

I am able to review the code and fix the problems. I am running Ubuntu 22.04.5 LTS (Ubuntu Budgie), with Python 3.10.12. This should also work for @jellepoland with Fedora.

@wecacuee could you please add a license to this gist? I would be glad to fork and send pull requests, provided it is FOSS licensed.

@vanschnapen I added MIT License. Please feel free to fork and update.

Updated to remove all info other than the content from annotations.

Thanks @wecacuee! With your commit 5d06efff181f5b99dea6234d01df617c21ecc14b the most urgent fix is done. Empty strings like before title="" were not possible and resulted in no changes in the annotations.
I will continue with my changes in the fork and inform you.

@wecacuee, feel free to add my gist fork as remote and to review my suggested changes.