Oneline filename batch rename (specific sed instance pattern matching)
 

Hi,

I apologise for creating yet another batch rename thread, I know that there is a lot of information on it and I have attempted for a couple hours to find a solution using google and other threads. If I missed a suitable answer I apologise ahead of time.

I would like to use only one liners from my bash command window.

I have a bunch of files that were named with periods as separators. I used the following command to remove the periods and tweak a few other things.

Original filename:
Current filename:
Desired filename:
I have been trying to use this command so that I can replace the fourth instance of '_' with '-' but a lot of commands give me errors while they seem to be working for other people. I have bash, sed, awk, gawk and perl installed and am using Cygwin on Windows XP. I cannot use simple matching as I did previously because the number preceding the dash will be different for each file.

This previews what the command will do if you remove the |/bin/sh and the preview shows the correct operation. However when I append the |/bin/sh to actually execute it, I get this error:
Finally, along the same lines, when I am ready to remove the '_' and replace it with spaces, I can use the command
as I did to make the first set of changes above. I noticed that this command seems to only change the first instance and I have to run it multiple times to remove all of them. How can I make this apply to all pattern instances?

Thanks in advance.

Brute force
 

I managed to continue using these commands to do the batch rename:

I manually replaced the '.' with '-' after I ran this command 3 times to change the fourth instance.

Original sample filename:
Final sample filename using variations of above commands:
I would still like to know if there is a more elegant form for this though so that I could just do a global search and replace of all '.' to '_' for instance and then say change the fourth instance of '_' to '-'.

Thanks.