leol1328 / read-it

OCR in a KDE Service Menu

Read-It

Convert scanned documents with a single right click in KDE Dolphin to plain text or a (searchable) "sandwich" PDF.

tl;dr;

Creates SCAN-0001.png.txt (or SCAN-0001.png.pdf) from SCAN-0001.png

Dependency

Read-It utilizes (and checks) tesseract(1). Please use your package management system to install. (the required package might be named 'tesseract-ocr')

If tesseract is not installed, Read-It will either not show up at all or show an error message.

Limitations

Read-It is restricted to reading text from images only. Currently no plans to read from PDF as well.

Read-It uses a very rough method to determine the default language to use with tesseract:

_tesseract_lang_code="$(tesseract --list-langs | grep ${LANG:0:2})"

(This works fine for a variety of languages such as englisch, italian, german, but not for spanish.)

It also roughly overwrites existing extraction files of the same name.

Ideas

Create one PDF with multiple pages from a list of image files.

Changelog

• 2023-10-03: 0.2.1: removed default TryExec=, replaced by own check, being more verbose on not-install error and success on creation
• 2023-10-03: 0.3.0: added option to create a seachable "sandwich" PDF