Closed dyreschlock closed 1 year ago
The new workflow for library image is based off of the core list in the database, rather than the platform enum in the code. However, if not platform enum exists for the core, the core will be skipped. Additionally, all of the platform information must be written for the core in the database.
This works well as long as there is JSON for the arcade core.
Need to think of a way to do this with non-JSON arcade cores, if possible.
I'm going to close this for now.
As long as there is JSON sorted into Genre folders, this will work. Eventually, we might need the filehash from the ROM rather than the JSON, but we have no idea how the Library and Playlists will work with openFPGA, if at all.
libretro-thumbnails has a MAME category for Arcade games. Boxarts are arcade flyers for the game. These should be used to generate library images for each of the arcade roms.
This will have to be done different then other roms because of how arcade cores are set up.
The filehash should probably be done on the arcade rom itself, rather than the JSON. And each core has a different directory where JSON are stored. The JSON directory should be data on the core object, and Pocket Games should know their Pocket Core.
Workflow:
As ProcessLibraryThumbnails is called, ARCADE will be on the PocketCoreInfo enum. This will return all PocketCores from the database that have a jsonDirectory that is not null. For each core, it will look at the JSON in that directory. The game name will be taken from the JSON filename, but the filehash used for the library image will be based on the arcade rom. So, the program will have to read the JSON to get the ROM name.