Open davidzbiral opened 4 months ago
@davidzbiral unfortunately, this is a bit complicated. We need to use a monospaced font to be able to calculate coordinates for each letter. If you agree, I will move this to 1.4.2 so we can revisit this issue once the annotator is more stable.
@adammertel Not sure whether counting coordinates is a stable and good approach, I experienced errors where e.g. the last letter failed to be selected while it seemed selected. And I have seen annotation tools where text selection was more secure. But I agree to move this to 1.4.2 - you did already.
Nevertheless, in a maximalistic future, also annotation of image polygons, manuscript lines recognized from image, and PDF lines and words in lines could be requested, and then, I concede, coordinates would be indispensible.