I don't think such a program exists. But I'll check here just in case. I suspect I need to create this program.

I have 2 file trees. There are a LOT of common files between these trees, and a lot of different files. Differences may exist only in metadata (e.g. time created, owner, permissions) or only in content. Usually if one is different the other is different, but since exceptions can exist, this cannot be depended on.

What I want to do is scan both of these trees and create as output 2 new trees, without disturbing the original trees. Output tree 1 shall contain all files that are unique to input tree 1. Output tree 2 shall contain all files that are unique to input tree 2. Output tree 3 shall contain all files that are shared in common between both input tree 1 and input tree 2.

If output tree 1 and output tree 3 are merged, the result will be identical to input tree 1. If output tree 2 and output tree 3 are merged, the result will be identical to input tree 2.

Any ideas?

Hi,

I don't know of a turnkey solution, but you can use fdupes to find duplicate files.
Combine that with a little bit of scripting and your should be able to achieve what you want without too much pain.

Regarding "similar files": recently there was an interesting blog post by Axel Beckert on that very topic.

http://noone.org/blog/English/Comput...ty-tester.html

HTH,

Evo2.