/jp/ - Otaku Culture

>>24566718
I've made a pretty poor proof-of-concept implementation.
A few things to note:
- Currently, the user cannot preview the image when selecting a file, due to the default file-selector of tkinter.
- The "crop image window" is initially too small and has to be enlarged manually.
- It' supposedly rather slow to "paste" images like this into a canvas. This is very notable in the video, but I never had any problems until I tried to record it. Maybe my computer is too strong to properly test this??
- I cropped the video to a rather small part of my screen to not dox myself and keep a small file-size. The first selected image is used for the central image (not shown), while the second selected image (file selection window not shown) starts filling out the second-largest squares. I've followed the process all the way until it starts filling out the smallest grid, so I know it works.

If the first two points are addressed, I think the script would be usable even in it's current form. It basically does everything specified in
>>24566510
except manually allowing the user to set a source image directory and template and auto-detecting boxes (the first 2 points). The first point is trivial, and the second point is outside of the scope of the project for now.
There are a ton of other improvements that could be made, of course, such as having an undo function in case someone would mess up an image selection or a crop hundreds of images into the process (not that the elite that visits this thread every would do such a silly mistake). Without an undo function, the best things is probably to fix up offending squares manually in GIMP after going through all 500+ squares without stopping.

I'm happy to share code in case anyone wants to work on their own version. Any suggestion on how to address the above shortcomings (in particular, on how to preview images at file-selection, and the undo function) would also be helpful: I'm not overly familiar with tkinter or PILLOW.