Maciej Budyś

Hi, manga-ocr author here. I'm glad to see the interest in this project, and would love to see it integrated with jidoujisho! One alternative to Chaquopy would be to export...

You likely have mokuro installed on another Python instance, which doesn't have CUDA-enabled PyTorch. Run: `where python` `where mokuro` to see their locations.

Hey, thanks. Seems mostly correct, though steps 2 and 3 are unnecessary.

You don't need to download any zip, when you run `pip3 install mokuro` this command downloads the package by itself.

Yeah, sorry about that. The fix is now included in 0.1.4 release.

I decided to use the natsort library after all, but thank you for your contribution, and sorry for the delayed repsonse.

ShareX doesn't need to explicitly call manga-ocr, the idea is that you capture a screenshot with ShareX, while manga-ocr is running in the background waiting for an image to appear...

See https://github.com/kha-white/manga-ocr/tree/master/manga_ocr_dev

Recent updates in [comic-text-detector ](https://github.com/dmMaze/comic-text-detector) seem to improve on this issue. I included them in 0.1.4 release.  

Yeah I've tried that too, export seems fine, but inference is not as straightforward since you need to do a beam search or something similar. Definitely doable, but requires some...