This commit creates output that is compatible with the new SabreDAT format, created by me and inspired by Logiqx XML DATs and the SuperDAT format. It uses a file-folder structure instead of rom-game structure, making it more versitile. It still cannot be read in currently, though it would be read in as a SuperDAT since it is the closest equivalent (so all non-file paths will be part of the game). Also, finding which type of XML DAT it is will be a bit difficult, so detection will have to improve.
This change comprises of multiple smaller changes:
- Convert all WriteToDat to WriteToDatFromDict, including MissFile
- Change obsolete Import to use updated ParseDict
- Remove all references to Parse
- Rename all references to "Dict" specific code to shorter names
- Remove more unused methods rendered obsolete by other changes
This change includes a few changes meant to keep the information from each DAT and have it retained in a single object which can then be written out in any format or added to other DATs with very little issue. This new system is much more versitile since all it needs is a way to convert each DAT to XML and then use the built-in output to output from the XML to the target format. This also gives the ability to clean a DAT by "converting" a DAT to the same format, removing any improper tagging or the such.
This change brings about a few changes:
1) A new LogLevel called "User" that is used for non-verbose, non-error output
2) Only User and Error are output to console now, not all of them
3) All programs have log to file enabled by default and all flags for enabling logging have been removed
4) Some former Verbose statements have been converted over to User because of the shift in usage.
The next step is to include the "source DAT name" in the dupe field instead of just "false". It will still be changed to "true" on merging, however. This will allow for multi-dat AB output since each unique source can have the non-dupe roms outputted. This will require a change in WriteToDat2 as well to enable "AB" mode, or at least a field called source that is default empty. If it's not empty, it tries to write out ONLY things that match that source. Maybe merge it in to the "diff" mechanic. Actually, if it's a flag, it will just output every variant: All in A, All in B, All in both, All not in both. Pretty easy.
