http://www.templeos.org/Wb/Kernel/Compress.html
http://www.templeos.org/Wb/Linux/TOSZ.CPP
Files are compressed individually.
The source code is like a PDF because it has binary graphics.
Saving disk space by compressing and decompressing files on the fly by appending a .Z file extension works, but the compression efficiency for individual files must be less than if you were using a compressed volume from the beginning. Compressing each file individually can't take advantage of common tokens between similar files.
Anyway, thanks for commenting, but I fail to see the correlation and the point you're trying to make. Again, FILE_ID.DIZ is about providing a standardized way of finding out about the contents of an archive, (which I suppose could just have one file, aside from the description file,) and so it isn't really relevant to how TempleOS compresses individual files.