How does the content scanner work?

Looking at the database file, it appears we can identify content by filename, size, CRC, MD5 or SHA1. If I scan a folder of Blu-ray images does it calculate the CRC, MD5 and SHA1 of every file in the folder to see if it matches anything in the database?


I think for optical disk images, it will check known serial locations and if it finds nothing will just skip.

This is what I’m getting at. If were scanning an .img file, it could be a track on an optical disc or an EPROM dump, maybe a tape image or even a floppy image… Say we need to scan a file “track1.img”, which is a 40Mb - how do we decide what is it or how to scan it?

looks like it’s based on filetype, and ambiguous formats like bin/img aren’t supported:

Hi folks, I have been trying to build a custom library organizer (being on a mac sometimes means I can’t access a lot of the ones out there).

For learning purposes I started looking at the libretro dat files, and looking at some of the files that compose the roms.

Most of the home systems such as SNES/Genesis it was easy to figure out as they are based on CRC/SHA1 Digest of the file.

Disc systems are still a bit challenging, and I wanted to share my experience here and perhaps learn more from others:

For Sega CD and to my surprise Saturn I used this reference:

Dreamcast was also easy to figure a GDI file and also read some bytes from the first track.

Gamecube seems that all serials are stored on the very first bytes of the iso file

Playstation was hard :frowning: Only because I had no easy way to extract the cue/ccd bin file contents of the System.CNF, I ended up porting (bchunk : to java (which is what I’m using for my tool), and then I was able to extract PS serial data.

I have not yet figured out turbografx 16, or neogeo cd. I’m hoping PS2/PS3/PSP follow similar pattern of System.CNF files, but have not reached those systems yet.

So could anyone share their knowledge of how those files get parsed by retroarch?