Using awk to iterate through list of files?

I'm pretty unfamiliar with awk, but I am looking f开发者_开发知识库or a way to iterate through all files in a directory and execute a simple awk command. My command is a simple substitution that replaces all tabs with two spaces.

awk '{gsub("\t","  ");print}'

How can this be extended to loop through a directory and execute the command on all files?

Pass the files to awk on the command line of course:

$ awk 'program' *

But probably it is easier to use

$ perl -pe 's/\t/  /g' *

Or, if you would rather have an in place edit, simply:

$ perl -i.orig -pe 's/\t/  /g' *

an improved version of Heisenbug's idea

find . -type f -exec awk '{gsub("\t"," ");print}' {} \;

it avoids calling another program like xargs. It seems to be a bit faster with some tests I did.

use find and redirect the output to awk:

find . -type f | xargs awk '{gsub("\t"," ");print}'

you can do it with one awk command

awk -F"\t" '{$1=$1;print $0 >FILENAME}' OFS=" " file

A useful snippet, that shows awk processing a list of files as if it is one stream (notice how the line number is used)

find $HOME/.ssh -name "*.pub" -type f \
  -exec awk '{printf "%c) %12s %s...%s %s\n", 64 + NR, $1, substr($2,1,7), substr($2,length($2)-7), $3}' \
  {} +

example output

A)      ssh-dss AAAAB3N...WhdJHA== bob@your.uncle
B)  ssh-ed25519 AAAAC3N...9a5kCpbT bob@your.uncle
C)      ssh-rsa AAAAB3N...fzh3oUGZ bob@your.uncle