[GUI Plugin] OCRthisPDF

[feuille]

[GUI Plugin] OCRthisPDF – Version 1.0.0 – May 31, 2026

A Calibre GUI plugin that adds a text layer to scanned PDFs (direct OCR processing / "in-place OCR" based on OCRmyPDF),
including an optional proofreading function (not yet implemented).

Version History:
----------------
Version 1.0.0 – May 31, 2026
- Initial release.

Installation:
-------------
1. Install OCRmyPDF and its dependencies (Tesseract, Ghostscript, etc.) according to the OCRmyPDF documentation at:
https://ocrmypdf.readthedocs.io/en/l...roduction.html. Additional training files for German and German Fraktur script are provided by the University of Mannheim: https://github.com/UB-Mannheim/tesseract/wiki
2. Install the plugin as usual. No configuration is currently required.
3. Select a suitable menu for the plugin in the Calibre settings.

Usage:
------
- In most cases, the default settings are sufficient.
- Detailed information can be found in the OCRmyPDF documentation.
- Language information for Tesseract is derived from the book metadata; the default is English.
- If the text in the PDF image is not correctly aligned (e.g., rotated by 90 degrees), check the "rotate" box.
- For books in German Fraktur script, check the "fraktur" box.
- The proofreading step has not yet been implemented (coming in version 2).

Bug Reports and Suggestions:
----------------------------
If you encounter any issues or have suggestions, please report them in the MobileRead forum.

[Menuhin]

I tried this plugin while I can successfully run ocrmypdf on a shell.

On selecting a pdf (highlighted) in calibre and clicking the ocrthispdf button, I pressed the "Do OCR for this book button". I did not run and there are some error messages:

calibre, version 8.5.0
ERROR: Unhandled exception: <b>ValueError</b>:__main__.__spec__ is None

calibre 8.5 embedded-python: False
Linux-6.12.107+deb13-amd64-x86_64-with-glibc2.41 Linux ('64bit', 'ELF')
('Linux', '6.12.107+deb13-amd64', '#1 SMP PREEMPT_DYNAMIC Debian 6.12.107-1 (2026-08-29)')
Python 3.13.5
Interface language: None
EXE path: /usr/bin/python3.13
Successfully initialized third party plugins: Find Duplicates (1, 10, 10) && KOReader Sync (0, 8, 2) && OCRthisPDF (1, 0, 0) && Toolbars Button (2, 0, 0)
Traceback (most recent call last):
File "calibre_plugins.ocrthispdf.main", line 1048, in ocr_my_pdf
if importlib.util.find_spec('__main__', package='ocrmypdf') is None:
~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^ ^^^^^^
File "<frozen importlib.util>", line 111, in find_spec
ValueError: __main__.__spec__ is None

[feuille]

Originally Posted by Menuhin

I tried this plugin while I can successfully run ocrmypdf on a shell.

On selecting a pdf (highlighted) in calibre and clicking the ocrthispdf button, I pressed the "Do OCR for this book button". I did not run and there are some error messages:

It looks like the OCRmyPDF installation isn't being found. I'll wrap the module search and loading sequence in a try-catch block for the next release for more clear messages.