05-29-2006, 02:14 PM
Yes, "file" is the mother of all file identification programs, at least historically! 
But I think there's some things that can be done better, and that's why some times ago I started playing with the idea of TrID. For example, .DOC, .XLS, .PPS, etc. for file are the same file type, i.e. a Microsoft OLE 2 compound object. Witch is true, in fact: they share the same container. TrID instead can usually discern from them, or tell that a particular zipped file is in fact an Open Document spreadsheet, or a Skin for the program xy, etc.
Plus TrID was designed with the ability to auto-generate new patterns for new files types simply scanning a bunch of files of that type (with the TrIDScan tool); that enabled me and some people around the world to quickly build a good db very quickly.
But the good news is: I'm just finishing porting TrID to FreeBASIC! So there will be also a native Linux port.
Bye!
But I think there's some things that can be done better, and that's why some times ago I started playing with the idea of TrID. For example, .DOC, .XLS, .PPS, etc. for file are the same file type, i.e. a Microsoft OLE 2 compound object. Witch is true, in fact: they share the same container. TrID instead can usually discern from them, or tell that a particular zipped file is in fact an Open Document spreadsheet, or a Skin for the program xy, etc.
Plus TrID was designed with the ability to auto-generate new patterns for new files types simply scanning a bunch of files of that type (with the TrIDScan tool); that enabled me and some people around the world to quickly build a good db very quickly.
But the good news is: I'm just finishing porting TrID to FreeBASIC! So there will be also a native Linux port.
Bye!
nline TrID file identifier - recognize over 2.300 file types