I'm tried to use the duplicate file script to find the duplicate file on my system, but it does not seem to like the spaces in the name:

 

Apr 30 07:08:20 Tower shfs: duplicate object: /mnt/disk4/My Movies/The Perfect Storm/folder.jpg (Minor Issues)

Apr 30 07:09:04 Tower shfs: duplicate object: /mnt/disk4/My Movies/The Perfect Storm.properties (Minor Issues)

Apr 30 07:44:39 Tower last message repeated 2 times

Apr 30 07:45:34 Tower last message repeated 2 times

Apr 30 07:45:44 Tower unmenu[2227]: ls: cannot access /mnt/*/*My: No such file or directory

Apr 30 07:45:44 Tower unmenu[2227]: ls: cannot access Movies/The: No such file or directory

Apr 30 07:45:44 Tower unmenu[2227]: ls: cannot access Perfect: No such file or directory

Apr 30 07:45:44 Tower unmenu[2227]: ls: cannot access Storm.properties*: No such file or directory

Apr 30 07:45:44 Tower unmenu[2227]: ls: cannot access /mnt/*/*My: No such file or directory

Apr 30 07:45:44 Tower unmenu[2227]: ls: cannot access Movies/The: No such file or directory

Apr 30 07:45:44 Tower unmenu[2227]: ls: cannot access Perfect: No such file or directory

Apr 30 07:45:44 Tower unmenu[2227]: ls: cannot access Storm/folder.jpg*: No such file or directory

 

 

Using Beta 5.06a Any one else see this before?

 

thanks,

dave

I know in regular 'ol Linux I use to just use a * right after the name.  eg. Perfect*  Only problem is it will mark files with the same first word as duplicate.

 

 

 

Kryspy

Yes the problem is that the line that searches for the duplicate files cannot handle filenames (or folders in this case) with spaces in them.

 

I'm looking at it right now but my knowledge of regular expressions isn't good enough. Even the comment before the code says:

 

    # I just kept adding backslashes until it worked... honest... Joe L.

    # or... as it says in the subway advertisements... "If U cn rd this, U can get a gud job"

 

 

The offending bit of code from line 43 of 25-unmenu-dupe_files.awk

    cmd="tail -10000 /var/log/syslog | grep \"duplicate object\" | sed \"s/  / /g\" | cut -d\" \" -f8- |"
    cmd = cmd "sed -e \"s/^\\\\/[^\\\\/]*\\\\/[^\\\\/]*\\\\/\\\\(.*\\\\)/ls -lad \\\\/mnt\\\\/*\\\\/'\\\\1'/\""
    cmd = cmd " | tr \"'\" \"*\"  | sort -u |  sh - "

 

Essentially we need quotes around argument list to the ls -lad command.

 

I'm not sure "dupe Files" is working in unMENU Version 1.5 Revision: 241. I tried disabling cache_dir, and let the browser sit for ~5hours and it looks the same screen as when I press the button (see attached). There are no syslog entries about "dupe" or "duplicate".

 

I had a few lines that said there were duplicate files when I was moving large ISO's from one disk to another, but I just think that was from the move process having the file in two places until the move finished. That is what prompted me to see if I have any other duplicate files. My server is 90% full of 7TB (10 disks). Shouldn't take 5+ hours I would think.

 

Does a message show up after the refresh?

I know this is a old post but....

 

has anyone figured out a way to get this script to work with with spaces in the folder/filenames?

 

Otherwise I got a TON of renaming to do

 

 

Myk

 

am getting the following lines at the end of my syslog when I press the dup files button:

 

Apr 10 00:35:12 Tower unmenu[15636]: sh: line 5: syntax error near unexpected token `('
Apr 10 00:35:12 Tower unmenu[15636]: sh: line 5: `ls -lad /mnt/*/*TV/Star_Trek_The_Next_Generation/Season 5/Star Trek- The Next Generation.S05E07.Unification (1).mkv*'

 

It dont like () as well, which is coming from theRenamer for my Tv shows....

 

Myk

 

can anyone tell me what it is supposed to output so I an try and fix the CMD for my setup?

 

Thanks

Myk

 

I don't know how to fix the script, but I think I can help with the root problem of dupes. I use http://clonespy.de/ from a windows box and point it at the disk shares instead of the user shares. Clonespy is pretty powerful, you can use it to root out duplicate names, or duplicate files with different names, or a combination. It uses checksums to verify content duplication, so theoretically it could false positive, but a checksum collision on a non duplicate file with the same bytecount would almost have to be done on purpose. I've used it for years to decrapify random downloads from different sources.