Use end FAD +1 of track 2 as start of lead-out for single density area.
gdi: set track end FAD based on track size, not start of next track
chd: ignore pad frames when calculating end FAD
cdi: end FAD was one off
Issue #1386
introduce Storage abstraction interfaces
use Storage i/f to load zip and 7z archives, cdi/cue/gdi/chd media,
cheats and to find custom textures.
Use android file picker to add storage. Request persisted access perms.
jni helper classes.
Issue #813
Verify md5 of flash, eeprom, vmus if no savestate used
Use chd header SHA1 as digest
Load .state.net state file at start
Force real BIOS for some games
Same problem as https://github.com/libretro/reicast-emulator/pull/24
in unix that iteration would not work if the path of the original .gdi file (where we are finding the 'parent directory') was:
1. not existent, ie a gdi on the same directory as the current. It was iterating until the index 3 and stop there. Disaster strikes later ofc, if the first characters weren't C:\ or other thing like that.
2. in a relative subdirectory that is 'small enough' ie: a/crazy-taxi.gdi. In this case the last '/' wouldn't be found, the derived string would be 'crazy-taxi.gdi' and the file not found ofc.
Anyway, this can be solved simply by searching the whole string and using a signed integer. It will go up to -1, stop iteration and increase to 0 on the len++.
Dunno if using relative subdirectories for the tracks on *other* platforms works because of the path separator being different from what's inside the gdi but with this code fixed it works on linux with / at least.
- Fix coreio fsize
- Fix coreio fopen to seek to 0
- Fix coreio/http to smart-escape urls
- For urls that include \ or % in the filenames, you have to pass them
escaped
- Update gdi parser to use (mostly) streamstream. That code is horrible
and should be rewritten
- coreio core is hacky and horrible at places as well
- Update imgreader to use coreio
- Update cdi parser + driver to use coreio
Flyinghead