Directory contents diff

Question

I'm looking for existing ideas / solutions to the problem of finding differences between two directories. Specifically how to identify files that might have been changed, renamed and moved.

A short list of things I've considered:

• try to pair up files missing in dir A with new files in dir b by using some heuristic such as 75% match in content. This just doesn't seem robust enough (problem cases include: significant changes in content, compression or encryption, possible multiple matches)
• use alternative data streams to add an id to each file. This would work only on NTFS.
• add a header/footer to each file containing and id. There's no way to guarantee header/footer will not corrupt the file.
• ask for user input for each change to determine if file is indeed deleted or simply moved. This is too hard on user.
• require user to rename/move files only by using special commands which will keep track of such changes. This is too hard on user.
• setting up a file system watcher to catch changes on the fly. Several issues (watcher must run at all times, is platform specific...)

Any ideas welcome...

Answer 1

Why don't you simply calculate MD5/SHA-1 or oher hash calculation on the folder content?

http://en.wikipedia.org/wiki/MD5

Build a list of files/folders for A and B. Compare which are present in A but not in B. Compare which are present in B nut not in A. For those which are present in both A and B perform a hash calculation.

Answer 2

A possible, not perfect, solution would be a version control system such as svn or git. that way, all change history is available. But users have to use specific commands.

Answer 3

for content match i recommend using some sort of distributed version control system such as git

it can pretty much detect all file operations such as copies, moves, renames, …