Issue Bug Resolved

#92320Metadata extraction (saving them) crash when meta title is too long

Maxime Lafontaine 09/16/20, 16:26 updated 10/20/26, 09:45

Description

The sys_file_metadata title field is a tinytext (255 chars).
If a PDF has metadata for the title longer that 255, the task crash and stop extracting the metadata.
The error is only visible in the log file.

Possible solution: crop the title, add a flag to problematic files and bypass them to extract metadata of next files.

This is a real problem because sometime the metadata are generator by the application (ex: Word) and nobody know what it is.

Thanks.

Reviews 2

History 610

  1. Georg Ringer changed Complexity easy Sprint Focus On Location Sprint <span data-processor="relativetime" data-value="2025-03-27T12:09:29+00:00" > 03/27/25, 12:09 </span>

  2. Georg Ringer commented and changed status Under Review Resolved done 0 % 100 % <span data-processor="relativetime" data-value="2026-10-20T08:50:12+00:00" > 10/20/26, 08:50 </span>

    Applied in changeset commit:7a5f0e49a3b750cb3375dda1d3014a94536992bd.

  3. Gerrit Code Review pushed patch set 2 for 14.3 to change 96902 status Resolved Under Review <span data-processor="relativetime" data-value="2026-10-20T09:24:52+00:00" > 10/20/26, 09:24 </span>

  4. Georg Ringer commented and changed status Under Review Resolved <span data-processor="relativetime" data-value="2026-10-20T09:45:12+00:00" > 10/20/26, 09:45 </span>

    Applied in changeset commit:e66909391c9b101752d9c6ab4176d877a94de3bd.