Sorry for yet another regex question, but I have exhausted all my knowledge of regex and need guidance please on solving a problem.

I have inherited some java code which was originally created using Poseidon modelling tool. We no longer use the modelling tool but the code has lots of lovely tags left behind which I was going to remove with a 'quick' perl script (but now I'm stuck).

Each file has a number of tags that look like the following

 * @poseidon-object-id [ Im39098af8m115b2fe9698mm7dca ]
[download]

N.B. The above line starts with a space just for clarication and the part in the bracket is always different

I have tried a couple of routines which traverse the filesystem (which I found on this site)

#!/usr/bin/perl -w

# Update: Script should now work without generating
# any warnings concering the %skip hash. Thanks to
# jmcnamara for helping with that :)

# This script will remove carriage returns (\r) from
# all files located below a specified parent directory.
# Could be expanded to allow any modifications.

use strict;
use File::Find;
use File::Basename;

####################################
# Configuration

# Directory where recursive modification will begin
my $dir = "/home/anthony/Development/Utils/temp1";

# File extensions to skip (ie: images)
# Each entry has the following look:
#      'file extension' => 1
my %skip = (
   'gif'  => 1,
   'jpg'  => 1,
   'jpeg' => 1,
   'png'  => 1
);


####################################
# Main Program - No Need To Edit :)

my $count;  # Just a counting thing
my @dirs;   # All recursed directories

# This does at least 85% of the work!
# It gets a recursive list of all directories
find(\&{
  sub {
    if (-d "$File::Find::dir/$_") {
      if ($count++ == 0) {
        push @dirs, $File::Find::dir;
      } else {
        push @dirs, $File::Find::dir . "/$_";
      }
    }
  }
}, $dir);


# Loop through all directories
foreach my $curDir (@dirs) {

  # Get list of files from the directory
  opendir DIR, $curDir;
    my @files = readdir DIR;
  closedir DIR;

  # Loop through the files
  foreach my $file (@files) {

    # Skip directories and "." and ".." entries
    # as well as excluded file extensions
    my ($nil,$ext) = $file =~ /^(.*?)\.(.*?)$/gs;
    $ext = '' unless defined $ext;
    next if (-d "$curDir/$file" || $file =~ /^\./ || $skip{$ext});

    # Read in the file
    open FILE, "$curDir/$file";
      my @lines = <FILE>;
    close FILE;

    # Modify and print the file back
    open FILE, ">$curDir/$file";
    
      print FILE map { s/.*\@[poseidon]$/ /g; $_; } @lines;
     
    close FILE;

  }

}
[download]

I have also tried a couple of simpler CLI's such as

 perl -p -i -e 'm/^@poseidon//g' AbstractEnum.java
[download]

But I am just either removing only the word 'poseidon' or I delete the entire contents of the file!

All I would like to do is to remove the entire line which starts with

* @poseidon ...

Could someone kindly help please.

Many thanks in advance.

20090501 Janitored by Corion: Changed PRE to code tags, as per Writeup Formatting Tips