Retaining the most recent file in a FS

hiradhu has asked for the wisdom of the Perl Monks concerning the following question:

I'm trying to optimize my project's file server. the folder structure is like this

--- Parent A
      |---Sub B (Sub folder of Parent A)
            |------Sub C (Sub folder of Sub B)
                     |--- File A (File within Sub C)
                     |--- File B (File within Sub C)

            |------Sub D (Sub folder of Sub B)
                    | --- File A (File within Sub D)
                    | --- File C (File within Sub D)
-- Parent B ....likewise many parent folders
[download]

Over the time the file server has huge number of files and duplicates.. Here Sub C ->File A and Sub D --- File A are same to some extent but one of the them is newer. I'm trying to consolidate it as

1. If I decide to retain only Folder Sub C

a. Check for duplicate files (like File A) in both Sub D & Sub C.

b. Check which File A (other duplicate files too) is newer comparing Sub D & Sub C

c. Copy the newer file to Sub C

d. Copy rest of the files to Sub C from Sub D
My Problems are:
1. Comparing time stamps - use of stat() ..dont know how to use "WIN32_SLOPPY_STAT"

2. Recursive search which one to use - grep?

- Use of XCopy?!!
I found File::Xcopy update|UD - copy files only if 1) the file exists in the target and 2) newer in time stamp http://search.cpan.org/~geotiger/File-Xcopy-0.12/Xcopy.pm#xmove($from,_$to,_$pat,_$par) Any help pls...?!!!!!

Comment on Retaining the most recent file in a FS Download Code

Replies are listed 'Best First'.
Re: Retaining the most recent file in a FS by Fletch (Bishop) on Dec 22, 2008 at 13:48 UTC
Sounds like you're trying to kludge something together rather than just biting the bullet and using a proper VCS system (e.g. git, mercurial, svn). The cake is a lie. The cake is a lie. The cake is a lie.	[reply]
Re: Retaining the most recent file in a FS by Corion (Patriarch) on Dec 22, 2008 at 12:50 UTC
`WIN32_SLOPPY_STAT` is an optimization which does not come into play in your case. Use File::Find to recursively find files. Use File::Copy to copy files and File::Path to create directories. Where do you have problems exactly? What code have you written already?	[reply] [d/l]
Re: Retaining the most recent file in a FS by bruno (Friar) on Dec 22, 2008 at 13:51 UTC
To find duplicate files, you can use the module File::Find::Duplicates. Here's a script that uses it, it could serve you as an example or you can use it as-is). Alternatively, you can use an external application, like fdupes (it's in Debian repositories).	[reply]