ToDo

Mostly the script does what I need it to do, so I am unlikely to implement these features soon. But it's fun to consider them.

• Check each epub for duplicate files first. Not sure how common this is.
• Currently, if phrases like ".jpeg" or ".gif" appear in your EPUB's plaintext, then they will be changed to ".png". Using something like TXR or some other parser, it might be possible to avoid this.
• Change the syntax to take arguments, instead of blindly operating on all epub files. How do other "x2y" utilities usually structure their input and output?
• Ideally, use epubcheck to validate the result. This is slightly complicated because what I really care about is if I have added any more warnings, ignoring those that existed already.
• Support optimization of more file types inside an EPUB.
• Experiment with different compression levels for zip.
• Currently, all images get converted to a smaller, 16-color, greyscale png image. For whatever reason, a png file tends to be smaller than any other format, but this conversion does also necessitate changing all the references to each image in other files, as it changes the extension. Investigate why this is so - it seems somewhat non-intuitive.