Page 1 of 1

Better support for Manga search

Posted: Thu Oct 22, 2020 12:44 am
by NightKnight
Hey. Is there a way to add more data/information- sources to Mylar? Mylar works excellently for finding and adding American comics, but does not work so well when if I search for Japanese manga's.

Re: Better support for Manga search

Posted: Thu Oct 22, 2020 2:09 pm
by evilhero
Not gonna happen - sorry.

Adding in different sources, means Mylar has to track items a different way than just CV, which means literally everything has to be modified / changed within the existing code to accommodate the new methods. Then you have to tack on ability to retrieve existing / new data from the new source, parsing it, keeping it maintained, etc. The first aspect alone is enough to kill most aspirations of something like this, let alone the rest. It would be easier to just start from an entirely new base if that was the purpose.

In so far as changing the data source, then you need to make sure that Mylar would be able to properly parse the filenames. Everything is built around American comics for a reason - and that main reason is naming conventions. The fact that there isn't any sane naming convention for American comics is hard enough, but adding in Manga naming conventions, which are absolutely atrocious as far as sanity checking goes, would be insane. It would break most of the core functionality within Mylar if it were added in because of all the exceptions that would have to be made during assumptive parsing. The file parser in Mylar works because it's had tonnes of iterations over the years to fine-tune everything to the point where unless it's some weird naming scheme, it'll parse things properly.

If it is something that interests someone, they can fork Mylar and try to do it - but I'm afraid it's not going to happen for me unless something happens to CV and that data-source goes kaput.