Intelligent recognition of renamed/moved file

[airjrdn]

I recently renamed a large number of 200+MB files. Instead of noticing that
the files existed and just needed renamed, delete & copy operations ensued,
which took a long time.

Just wondering what your thoughts might be on...
1 - for each source file that doesn't have a match at the destination based on
name/date/etc.
2 - create a hash of the first 32k of each unmatched file on both source &
destination
3 - if any matches are found, do a full file hash (on those) to see if they
are actual matches (or whatever other method you use)
4 - rename/move any that are actual matches
5 - delete/copy the rest as normal

This should make moved/renamed files not have to incur a full copy operation.

Thanks

[Zenju]

This scenario will be eventually covered via:
[404, Invalid URL: https://sourceforge.net/tracker/?func=detail&aid=2880632&group_id=234430&atid=1093083]