UNIX for Dummies Questions & Answers

I'm trying to replace the string %2d from a text file using sed, and I can't seem to find the right key combination. I've tried:

sed 's/%2d/-/' foo

The above doesn't work, presumably because of the %. Interestingly, I don't get any kind of error message at all. It appears to succeed, but when I vi the file, the %2d string is still present throughout the file. Can anyone help me figure out which magic phrase to invoke here?

Thanks!

pipe the output to a new file. or use -i option if your sed supports it.

Did you confirm the above from the output in stdout ?

If so, use solution as suggested by ghostdog

Sorry, I'm a little new at this, and don't know how to use stdout. However, to clarify, it seems to be replacing the first instance of the string in each line it encounters, but if there's more than one ocurrence in the line, the remaining iterations are not removed. For example, I've got a test.txt file with the following line:

AF%20Form%20931%20Performance%20Feedback%20Worksheet(AB%20thru%20TSgt).xfdl

When I run test.txt through the following command: sed 's/%20/ /' test.txt > test2.txt

It becomes this:

AF Form%20931%20Performance%20Feedback%20Worksheet(AB%20thru%20TSgt).xfdl

How do I get the sed command to remove all ocurrences in the file?

And, to answer your question, -i isn't supported in my version. I got an illegal option when I tried to use it:

sed -i 's/%20/ /' test.txt > test2.txt
sed: illegal option -- i

Code:

sed 's/%20/ /g' test.txt > test2.txt

Wahooo! That worked! Thanks so much!

(Such a simple solution, I feel silly now ).