Fileformats: constant indetifier sequences

• • Reply

  12 years ago

  by konikula

  Matej Kasper

  Trutnov, Czech Republic

  Joined 15 years ago

  Hi, because I think that my conception of disc structure backup is valid, and as result using dataminer (christophe's checkdisk) and consequently disc structure backup is only valid and also most filling method for file restoring. But as I am working on wide range of smaller or bigger projects, I have no time to study every potential fileformat, so actual version is brute force in sense, that filename and path is 100% recovered only for unchanged files, and ocassionaly for slightly or "somehow fortunatelly" modified files. If we walk through our data-modifiing habits, we shall see, that for example mp3 data layer is dynamical through tag updates compared to exe files, which mainly stay intact from creation at disk. So why I am creating this thread, (which should be rather sticky)

  if you understand concept of path & filename recovery, and if you have some knowledge in particular dataformat, which is of dynamic character, please, discuss (or just leave here) the sequence which with best efficiency identifies such file after any change in its tags (or similar 'dynamic' file part).

  Example of such quick note:
  MP3: check format, skip ID3 tags and following 100 bytes should ever be mp3 data stream, which identifies file for longest file's lifetime.

  Another sample:
  JPEG: see your thread for imagetypes, to skip JFIF and EXIF informations, after you are in data part, take anything, it is static.

  HTML: parse out TITLE, and concat it with META informations (it is discutable, but I guess it could help)

   

  I hope someone will join this thread :)

  Report abuse