Hi all,

Here is what I want to do ;

In a directory where I have multiple files, I want to be able to delete only files containing a certain word.

For now, I'm able to pickup those files doing this ;

find . |xargs grep myword

I would like to add rm -f. Tried piping at the end and beginning but gives me errors.

Any suggestions? I'm sure it's pretty easy.

Thanks

find . |xargs grep myword |awk -F: '{print $1}' |sort -u |xargs rm

Every file that matches will give you filename:line that contains myword

The awk -F: says to break the output of the previous xargs command into fields delimited by the colon. The '{print $1}' prints only the first field which will be the filename.

The sort -u says to to only list each filename once (otherwise you could see the same file more than once if it had myword on more than one line).

The xargs rm says to remove the filenames that were listed via the preceding pipes.



Having said all that I'd suggest the above is fairly dangerous unless you're absolutely sure myword appears ONLY in files you want to remove. You can make rm prompt you with the filename and ask if you want to delete it by changing the last part to:

xargs rm -i

I'm not at home, so I can't test it out right now, but I would think you might be able to do what you're trying using wildcards (*). This wouldn't work for hidden files, though (I could be wrong about that, too). I could be wrong about everything, come to think of it. :]

EDIT: The post above me answered your question much better than I, whilst I was typing away this halfway idea. :]

hi there,
if you need it recursive, you can add -r to the grep call.
regards,
slackie1000

Wow. Thanks a lot you guys for all these different ways of doing it.

Regards