Sort Many Files in Subdirectories into Directories by File Type

garyozzy · 02-15-2011, 07:54 AM

I have a script that uses wget to download certain types of files. Mostly mp3's and pdf's.

I want to sort them into two directories, one for mp3, one for pdf.

The problem is, the files are downloaded by wget into directories many layers (4-5) deep.

In the past, I've used

Code:

mv **/* .

until all the files are in one directory then

Code:

mv *.mp3 mp3/ && mv *.pdf pdf/

How can I either put this in a script to do it for me or have a one-liner to do it? Any help would be appreciated.

druuna · 02-15-2011, 08:03 AM

Hi,

Have a look at the find command.

Something like this should do what you want/need: find . -type f -iname "*.mp3" -exec mv {} /path/to/mp3dir/ \;

The above find command starts looking for files in the directory you are standing (the . [dot]), it will only look for normal files (the -type f part) and then look for files ending with mp3 (-iname "*.mp3"), the iname makes sure that it is case insensitive. The -exec ........ part moves files that are found to the /path/to/mp3dir/ directory.

Hope this helps.

edit: You need a second find for the pdf files.

grail · 02-15-2011, 08:05 AM

I would probably go with a find into a loop and then use your moves in an if.

Let me know if you need more information?

garyozzy · 02-15-2011, 08:16 AM

This was very helpful! Thanks!

Just a few questions, though, so I can understand

.

What does the {} after mv do, as well as the \;?

Quote:

Originally Posted by druuna

Hi,

Have a look at the find command.

Something like this should do what you want/need: find . -type f -iname "*.mp3" -exec mv {} /path/to/mp3dir/ \;

The above find command starts looking for files in the directory you are standing (the . [dot]), it will only look for normal files (the -type f part) and then look for files ending with mp3 (-iname "*.mp3"), the iname makes sure that it is case insensitive. The -exec ........ part moves files that are found to the /path/to/mp3dir/ directory.

Hope this helps.

edit: You need a second find for the pdf files.

druuna · 02-15-2011, 08:22 AM

Hi,

The {} is replaced by the file(s) that are found by the find command. The \; tells find where the commands used with the -exec option ends.

The find command is very powerful and can do a lot of things, have a look at these url's for some more examples and explanations:

A Unix/Linux "find" Command Tutorial
Linux find command examples

Hope this helps.

theNbomr · 02-15-2011, 03:18 PM

Quote:

The {} is replaced by the file(s) that are found by the find command

Perhaps more accurately, the {} is a placeholder for each individual filename that matches the rest of the find arguments. So, it will be given iteratively as an argument to the mv command. The difference is somewhat subtle, and in this example it may not matter, however in many other similar applications it does matter.

man find for details.

--- rod.