Hi all.

As someone who is inexperienced in programming, I have attempted to code up a little bash script to manipulate a number of files in a set of directories based on certain criteria/conditions etc.

By way of example, I have directory X, and a set of files in it. I want to rename certain types of files (i.e. those with certain extensions) to the name of the directory they are in - preserving their extensions of course. I also want the script to delete files/subdirectories that do not match these criteria.

One catch is that some of the directories contain two or more of the same file types, (for example, file1.mpg, file2.mpg, file3.mpg). In such a case, I want the files to be renamed directoryX-1.mpg, directoryX-2.mpg etc.

So far, I have come up with the below:

I repeat this command, substituting .mpg for the other extensions I want to manipulate. This means I can achieve the renaming-to-inherited-directory aspect of things. This is obviously very inelegant, and only gets me halfway there in terms of what I want the script to do.

After much reading, I am thinking of an if condition that does basically 'if there are more than one of files *.mpg, then append integers beginning with 1 to the end of the directory name; else proceed with standard rename'. After this portion of the script completes, I'm thinking that it is then I insert something that 'recursively delete all files not having extension X Y or Z'.

It is on the above two points I am stuck.

Excuse the long post - I have tried to be as clear as possible.
Any help is appreciated.

Cheers,
zorblart

First, spread the code out over several lines instead of trying to make it a one liner! This will make it easier to read and to work on.
Also, when starting out, substitute the real command with echo statements which tell what the real command would do. This will keep you from ruining a bunch of files. Once you have debugged the loops and other complexities, then you can substitue the ecgo staements with the real commands.