Kooka and UTF-8
 

I tried to use kooka in SuSE 9.1, but I failed, as its ocr function returns much garbage in the text because the ocr engine returns ISO 8859-2 text, while kooka shows that in UTF-8. (The conversion is out of question taking into account how often I must use ocr).

I know that both ocr engines (gocr and ocrad) could return UTF-8 text, too, but kooka calls them without the necessary command line parameter, so they output ISO 8859-2 text.

Is there a way to make kooka run those ocr engines with the correct command line parameter?

In the past I used omnipage under wine for ocr with reasonable results. It's always bugged me that this was the only non-linux software. So recently I've started using ocrad with kooka and the results are on a par with what I achieved from omnipage.

My configuration is straight out of the box.

I'm using kooka 0.44 and ocrad 0.12.

And am now windows free!

Well, I believe that kooka actually works well on slackware, but I have a SuSE with default utf-8 charcoding.

Here, even when I enter some accented characters into the recognized text window of kooka, and copy & paste it into an other editor, I get nothing but garbage.

And I do not know how to configure kooka to use utf-8 by default; or how to convert kooka's ocr results to utf-8. (Doing an iso_8859-1 > utf-8 conversion produced garbage, too)