nc-watermark-pdf

Simple Scripting to Add Watermarks to PDFs in a NextCloud Instance. Contains two versions

PHP Script

To be run as cron job on the NextCloud server. Works by scanning a hard-coded directory for PDF files, compares against a persisted list of previously scanned files, and adds watermarks to new files. Checks only the file name, not content or hash.

Adds watermarks using markpdf. This is a GO binary, which requires the binary to be installed on the system.

Run the script as

php watermark.php

or as

php watermark.php populate

to populate the watermark cache initially with existing, already watermarked PDFs.

When running this as cron job, there are some limitations:

• the directory to scan is hard-coded in the script, so when renaming it, the script needs to be adjusted
• when renaming a file in the cloud, it is detected as a new file and gets a new watermark
• if between two runs of the cron job, a watermarked version of file A is deleted and a watermark-free version of file A is uploaded, the script does not detect it as a "new" file and does not add a watermark again

Python Script

To be run via NextCloud Workflows, if shell_exec is enabled for PHP in the webserver running NextCloud.

It implements two ways of adding watermarks to PDFs:

1. Using PyPDF. This is a pure Python solution, which has its benefits, but is also rather slow.
2. Using markpdf. This is a GO binary, which is much faster, but requires the binary to be installed on the system.

The script runs on Python >= 3.9. Call it as

python script.py --event_type %e --file_id %i --actor_user_id %a --owner_user_id %o --nextcloud_relative_path %n --locally_available_file %f --old_nextcloud_relative_file_path %x

in the workflow specification.

Paths are currently hard-coded and may need to be adjusted for other setups.