This idea comes directly from Windows’ snipping tool. But before trying to adapt it (I don’t know if it will be difficult), I wonder if it is a useful feature and if anyone would actually use it.
having a light weight OCR implementation for spectacle would useful for those extracting data from screen shots of text / tables on the screen
screen text is neatly organized and not skewed or angled like it could be on a scanned image, so the OCR capability would not need to be very robust.
this would save the trouble of having to pass the screenshot thru another tool like gImageReader for extracting the text.
this also means a rotate feature would need to be included for the images where tables / text are printed for landscape mode.