Awk substitution within a find command

Question

I am working on a bash script that should find files such as

/var/www/templates/testdoctype/test_file.html.head

and return something like

cp -f '/var/www/sites/t/test/test_file.html' '/home/user/tmp/test_file.html' 

my script so far looks like this:

#!/bin/bash
DOCPATH='/var/www/templates/testdoctype'
INSTALL_PATH='/var/www/sites/t/test'
PKGBACKPATH='/home/user/tmp'
function find_suffix_head {
 find "$FROM/$DOCTYPE" -type f -iname "*.head" -print \
 | awk -v docpath="$DOCPATH" -v installpath="$INSTALL_PATH" -v pkgbackpath="$PKGBACKPATH" \
 '{ sub( docpath, installpath ) sub(/.head$/, "") } { printf "cp -f ""'\''"0ドル"'\''"" " ; sub( installpath, pkgbackpath ) ; print "'\''"0ドル"'\''" }'
}
find_suffix_head

This returns

cp -f '/var/www/templates/testdoctype/test_file.html' '/var/www/templates/testdoctype/test_file.html'

So, sub(/.head$/, "") works as it should, but sub( docpath, installpath ) and sub( installpath, pkgbackpath ) does not.

Answer 1

No need for awk, you can do it with bash:

function find_suffix_head {
 find "$FROM/$DOCTYPE" -type f -name "*.head" | while read filename; do
 filename=${filename%.head} # strip suffix
 filename=${filename#/var/www/templates/testdoctype} # strip prefix
 echo cp -f "$INSTALL_PATH/$filename" "$PKGBACKPATH/$filename"
 done
}

From there you can just run the cp, too, rather than echoing it.

• Dan-Simon Myrland: Ah! That was indeed a much cleaner solution, thank you very much!