Question

I want to write a bash command that greps all *.txt file with a pattern in current folder to another folder. Should I use find or for loop? I tried using find but it seems to complicate things.

Edit: I want to copy files with a specific pattern to a different folder. For example:

A.txt
B.txt
C.txt

all have the word “foo” in them. I want grep to remove “foo” and send it to a different folder with the same name. I don’t want to change the original file in any way.

Answer 1

Using for would probably be a lot easier for this than find. Something like this:

otherdir='your_other_directory'
for file in *.txt; do
    grep -q 'foo' $file && grep -v 'foo' < $file > $otherdir/$file
done

If your grep doesn’t understand -q then:

otherdir='your_other_directory'
for file in *.txt; do
    grep 'foo' $file > /dev/null && grep -v 'foo' < $file > $otherdir/$file
done

In any case, grep returns a true value to the shell if it finds a match and the X && Y construct executes the Y command if X returns a true value.

UPDATE: The above solution assumes (as noted by Johnsyweb) that you want to remove any lines that contain “foo”. If you just want to remove “foo” without removing whole lines, then sed is your friend:

otherdir='your_other_directory'
for file in *.txt; do
    grep -q 'foo' $file && sed 's/foo//g' < $file > $otherdir/$file
done

Or:

otherdir='your_other_directory'
for file in *.txt; do
    grep 'foo' $file > /dev/null && sed 's/foo//g' < $file > $otherdir/$file
done