idstring "UKTP162N" get FILES long get ARCSIZE long endian guess FILES for i = 0 < FILES get SIZE long findloc NAMESZ long "\x2E\x74" savepos OFFSET getdstring NAME NAMESZ clog NAME OFFSET ARCSIZE SIZE next i
Not sure what you're trying to get with that script, it's totally not compatible with your attached file and doesn't make much sense (for example, you're passing NAMESZ offset, obtained from findloc, as getdstring's size), also considering that the file doesn't contain clear offsets or sizes. Here is an example, how it can be parsed in a "blind" way, though it's also not exactly correct, since it's tested only on single file. I suppose resulted files are headerless tga or something, which can be converted with rawtex or something to readable form (converted files in the attachment are created with r8g8b8a8 format). Besides, it looks more like container, where 4 parts of a single symmetric texture are stored in separated files, but I'm probably wrong.
You can start with reading some tutorials, there are some in Tutorials section on zenhax and xentax - like this or that. Also, there are tutorials on other sources, just google for them (like this one).