How to remove only the last word from a file

I created the following Perl one-liner in order to remove a word from a file

This Perl also escapes special characters such as @ or $ or *, so every word that contains a special character will removed from the file.

How to change the Perl syntax in order to delete only the last matched word from a file and not all the words?

Example

 more file

 Kuku
 Toto
 Kuku
 kuku

export REPLACE_NAME="Kuku"
export REPLACE_WITH=""
perl -i -pe 'next if /^#/; s/(^|s)Q$ENV{REPLACE_NAME }E(s|$)/$1$ENV{ REPLACE_WITH }$2/'  file

expected results

 more file
 Kuku
 Toto
 Kuku

another example

when – export REPLACE_NAME=”mark@$!”

more file

mark@$!
hgst#@
hhfdd@@

expected results

hgst#@
hhfdd@@

Answer

Use Tie::File to make this easier.

$ perl -MTie::File -E'tie @file, "Tie::File", shift or die $!; $file[-1] =~ s/bQ$ENV{REPLACE_NAME}Eb/$ENV{REPLACE_WITH}/' file

Update: Rewriting as a program in order to explain it.

# Load the Tie::File module
use Tie::File;

# Tie::File allows you to create a link between an array and a file,
# so that any changes you make to the array are reflected in file.
# The "tie()" function connects the file (passed as an argument and
# therefore accessible using shift()) to a new array (called @file).
tie my @file, 'Tie::File', shift
  or die $!;

# The last line of the file will be in $file[-1].
# We use s/.../.../ to make a substitution on that line.
$file[-1] =~ s/bQ$ENV{REPLACE_NAME}Eb/$ENV{REPLACE_WITH}/;

Update: So now you’ve changed you requirements spec. You want to remove the last occurrence of the string, which is not necessarily on the last line of the file.

Honestly, I think you’ve moved past the kind of task that I’d write in command-line switches. It’d write a separate program that looks something like this:

#!/usr/bin/perl

use strict;
use warnings;

use Tie::File;

tie my @file, 'Tie::File', shift
  or die $!;

foreach (reverse @file) {
  if (s/bQ$ENV{REPLACE_NAME}Eb/$ENV{REPLACE_WITH}/) {
    last;
  }
}