A Python script to split volume-based CBZ archives into chapter-based CBZ archives, for Tachiyomi J2K.

$ python ./j2konverter.py <source> <pattern> <dest>

source is the path to either a single archive or a directory containing multiple archives`

pattern is a stream of tokens that get matched to file names from inside the archives to generate metadata. The aim is to fill the following data structure:

scanner: str = ""
volume: int = 0
chapter: int = 0
page: int = 0
name: str = ""

So, if the file names inside the archive are, for example SomeTitle V01 - C21 / 16 [Chapter Name] [Company] then the pattern to fill the metadata will look like

x|volume|x|chapter|x|page|[name]|[scanner]

x is just used to discard words we don't need. The square brackets are automatically skipped and are used to describe multi-word tokens that we need.

If there is a lot of garbage between two tokens that we need, we can discard it as

loopToken where Token should be replaced by the first token after the garbage that we want.

Finally, dest is the path to the directory where we want the outputs to be generated.