UNIX for Advanced & Expert Users

Hi all,

I have this very strange (and tiring) trouble with sed (or any command that works with files for that matter)

I need to do the following

1. I have a string say "UTN555"
2. I need to extract just the last three digits
3. I did this

echo UTN555 | sed "s/UTN//"

4. Seems pretty simple, but here comes the trouble. If "UTN555" is just any random string, I get the correct result ==> 555

5. The problem is when UTN555 is a file. In this case , sed (or even cut for that matter) interprets it as a file and works on the contents of this file, which is
not what I want to do.

Can you guys please help me figure this one out ??

I got a work around for this , but I think there must be a way to suppress this mis-interpretation. my work around

1. Dump UTN555 into a file
2. Give this file as input to sed.

As for googling , I am at loss for keywords to describe this problem.

Hope that explains the problem. Thanks in advance ppl.

cheers

See: problem extracting substring in korn shell

woooow it worked !!! . though I dont fully recognize the syntax (gotta google I guess). Thanks a ton Perderabo