I have found that some archived versions of old textbooks use DJVU files. From what I understand, the DJVU file only contains text, but perhaps I am wrong about that and there is some additional image data that is not shown in an e-reader.
Archiving something like a math textbook using this format feels strange, as the equations are converted to seemingly random characters. Presumably this is just a result of an OCR software that converts the symbols to the closest ascii characters. Is there any logic to this?
Does there exist some way to decode the DJVU file to some other format (e.g. PDF but maybe also postscript or latex?) to recover the equations? Or is that information just lost?