ilovechristy has asked for the wisdom of the Perl Monks concerning the following question:

Wise Perl People, I am trying to remove spam from a txt file which looks like the following: 
AUTHOR: bobMonk
TITLE: Monk Software Chart
STATUS: Publish
CATEGORY: Monks
DATE: 05/21/2006 11:36:16 AM
-----
COMMENT:
AUTHOR: Joe Legitimate Man
EMAIL: Joe@legitdomain.com
This is a legitimate comment.
-----
COMMENT:
AUTHOR: more spam casino stuff
EMAIL: marengo@greenfield.com
Blabla, more spam. poker blabla
-----
COMMENT:
AUTHOR: Joe Legitimate Man
EMAIL: Joe@legitdomain.com
This is a legitimate comment.
-----

[download]
My goal is to remove the spam comments using regex filters. Comments start from the term "COMMENT:" and end on "-----". I have tried various approaches to this, but none have been successful. Any assistance would be much appreciated. Thank you all.
Buddha bless you.

Replies are listed 'Best First'.
Re: Delete all "records" which contain a regex match
by ikegami (Patriarch) on Jun 05, 2006 at 16:14 UTC
    How can you tell which commment is spam and which one isn't? I made up a few filters in the following code. 
    local $/ = '-----';

while (<DATA>) {
   # Spam if author contains "casino".
   next if /^AUTHOR:.*casino/msi;
   
   # Spam if body contains "poker".
   next if /^(?!AUTHOR|TITLE|COMMENT|DATE|CATEGORY).*poker/msi;

   # Spam if anything contains "viagra".
   next if /viagra/msi;

   print;
}

    [download]
    Read more... __DATA__ (809 Bytes)
[reply]
[d/l]
[select]
Re: Delete all "records" which contain a regex match
by liverpole (Monsignor) on Jun 05, 2006 at 15:46 UTC
    Hi ilovechristy,

    You say I have tried various approaches to this.  Can you show us the code you've tried, so we'll know where it is you're getting stuck?  That way, if it's a specific bug you've got, it'll be easier to find and fix it, and if it's just a matter of not knowing what to do at a certain point, we can help you get past the particular hurdle.

    s''(q.S:$/9=(T1';s;(..)(..);$..=substr+crypt($1,$2),2,3;eg;print$..$/
[reply]
      My apologies. Here are two methods that I have tried that read in a file from stdin:

      Attempt 1: 
      #!/usr/bin/perl -w

undef $/;

while (<>) {
        s/^COMMENT:\s+AUTHOR: .*?poker.*?^-----$//gsm;
        print;
}

      [download]
      Attempt 2 (completely nonsensical): 
      #!/usr/bin/perl -w

#undef $/;

while (<>) {
        next unless (@foo = /^COMMENT:$/ ... /^-----$/);
        print if ! /poker/;
#       $_ =~ /poker/;
#       print $_, "\n";
        print @foo;
}

      [download]
      Buddha bless you.
[reply]
[d/l]
[select]

        Hi, ilovechristy, here is one way to do it. Also have a look at perlre.

        use strict;
use warnings;
local $/;
my $data = <DATA>;
$data =~  s/COMMENT:(?:(?!-----).)*-----//gs;        
print $data;

output:
-------
AUTHOR: bobMonk
TITLE: Monk Software Chart
STATUS: Publish
CATEGORY: Monks
DATE: 05/21/2006 11:36:16 AM
-----

        [download]

        Prasad

[reply]
[d/l]

Back to Seekers of Perl Wisdom