I'm trying to write a simple script that will take a graphical selection, perform optical character recognition on it and put the resulting text into the selection buffer.
In theory there's nothing to it - save the selection as an image file, convert the file to a tiff, run tesseract on that, and select the contents of the resulting output.
However I get stuck at step 1. I can see three potential KDE applications, none of which quite works for me. kolourpaint has no DCOP interface, kview does but file_save does not appear in it, and krita does but doesn't provide edit_paste.
I had a look at the source of kview to see if I could work out why some menu items appear in the DCOP interface and others not. But it's all more investment in time than I feel is warranted by such an unambitious project.
Any suggestions?