Help renaming files

[pk_fox]

Hi all, I've just downloaded a lot of Mp3 files from Amazon and they have names like this: B00EXWBVRM_(disc_1)_09_-_Why'd_You_Only_Call_Me_When_You're_.mp3, anyone know a quick and dirty ( or slow and elegant ) way to strip all the crap off the name leaving e.g. Why'd_You_Only_Call_Me_When_You're_.mp3, from what I can see the name is always preceded by two underscores, I could code something but I'm sure one of you knows a command line solution. Many thanks

[nelz]

You can use the coreutils rename, the more powerful rename from http://rename.berlios.de/ or a combination of mv and sed (there's also mmv but I've not tried that). But the easiest to get to grips with is probably Krename.

[johnlane]

One way to sort your file names out is to use EasyTag. See my piece in LXF171.

[pk_fox]

One way to sort your file names out is to use EasyTag. See my piece in LXF171.

Hi, yes I'm a big fan of easytag, but I can't see a way of renaming thousands of files using easytag, the situation I'm in is Amazon has kept a record of every CD I've bought from them and has made them downloadable in mp3 format with horribly convoluted names and I need to work out a way of renaming them via a batch process - I will probably end up writing a c# programme to achieve it - thank you for your time.

[MartyBartfast]

This would do it, assuming that the delimiter between the bits you do want and the bits you don't want is always a hyphen "-".

Code: Select all

IFS='
'

for x in *.mp3 ; do y=`echo $x | cut -d"-" -f 2- | sed "s/^_//g" `; mv -v "$x" "$y" ; done



The IFS and the following line are in case some files have a space in their name, it will set the field separator to newline only (not required if there are no files with spaces).

[pk_fox]

This would do it, assuming that the delimiter between the bits you do want and the bits you don't want is always a hyphen "-".

Code: Select all

IFS='
'

for x in *.mp3 ; do y=`echo $x | cut -d"-" -f 2- | sed "s/^_//g" `; mv -v "$x" "$y" ; done



The IFS and the following line are in case some files have a space in their name, it will set the field separator to newline only (not required if there are no files with spaces).

Hi there, thanks for your help, I tried your code and it works fine, my problem now is I need to traverse all directories, the directory structure is /amazonmp3/artist/album/ and so on for every artist plus the artist could have more than one album. Hope you don't mind me asking.

[MartyBartfast]

Off the top of my head try

Code: Select all


for dirname in find /amazonmp3 -type d ; do cd $dirname ; for x in *.mp3 ; do y=`echo $x | cut -d"-" -f 2- | sed "s/^_//g" `; echo mv -v "$x" "$y" ; done ; done



note I've changed the command to 'echo mv -v "$x" "$y" ' that way you can run it and it will just display what it's going to do, if everything looks good just remove the "echo" and run it and it will do the job.

It would probably be neater and more maintainable to stick this in a shell script and indent it properly, particularly if it's going to get any more complex.

[pk_fox]

Off the top of my head try

Code: Select all


for dirname in find /amazonmp3 -type d ; do cd $dirname ; for x in *.mp3 ; do y=`echo $x | cut -d"-" -f 2- | sed "s/^_//g" `; echo mv -v "$x" "$y" ; done ; done



note I've changed the command to 'echo mv -v "$x" "$y" ' that way you can run it and it will just display what it's going to do, if everything looks good just remove the "echo" and run it and it will do the job.

It would probably be neater and more maintainable to stick this in a shell script and indent it properly, particularly if it's going to get any more complex.

Thanks again, Im new to all this scripting game, if I place this in a script, do I need the semi colons at the end of every command ?

[nelz]

You can replace the semi-colons with line breaks - you only need the semi-colons when you run it as a one-liner in a shell.

[MartyBartfast]

something like this is easier to read and will be easier to maintain when you've forgotten what the code was supposed to be doing:

Code: Select all

for dirname in find /amazonmp3 -type d
do
   cd $dirname 
   for x in *.mp3    
   do
      y=`echo $x | cut -d"-" -f 2- | sed "s/^_//g" `
      echo mv -v "$x" "$y"    
   done
done


[nelz]

If you want easier to read, replace the backticks with $().

But the whole thing is much easier with a dedicated command, like rename. If you want to remove everything up to and including the first double underscore

Code: Select all

rename -R -t -s/.\*__//r /amazommp3

The -t option shows the renames without performing them. Remove this, or replace it with -v, to do it for real.

[pk_fox]

If you want easier to read, replace the backticks with $().

But the whole thing is much easier with a dedicated command, like rename. If you want to remove everything up to and including the first double underscore

Code: Select all

rename -R -t -s/.\*__//r /amazommp3

The -t option shows the renames without performing them. Remove this, or replace it with -v, to do it for real.

Whoa, that's impressive - I will check it out thanks.

[nelz]

That's the rename command I mentioned earlier in this thread, not the one from coreutils.

[pk_fox]

That's the rename command I mentioned earlier in this thread, not the one from coreutils.

Hi Nelz, I downloaded the source for the rename tool but the "make" fails with this error

Code: Select all

rename.c:68:6: error: #elif with no expression
 #elif
      ^


I checked the file rename.c and line 68 is a standard #include statement, any ideas ? or can I get a binary from somewhere ?

edit
Apparently gcc compiler has changed since rename was written and if you use #elif it must be followed by an expression (it didn't used to be like this ) so I changed it to #else and all was fine. Thanks for the heads up - really useful tool.

Code: Select all

// Original code.
#if HAVE_REGEX_H
  #include <regex.h>
#elif
  #include "regex.h"
#endif



changed to

Code: Select all


#if HAVE_REGEX_H
  #include <regex.h>
#else
  #include "regex.h"
#endif