sendmail logging combine two hashes

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

I'm trying to grab each line from sendmail syslog file and combine the different lines where the message id's are the same and combine them to show source and destination.

 
p3D8Bj1f007915 user1@domain.com
p3D8EF8n007918 user2@domain.com

p3D8Bj1f007915 server101@domain.com
p3D8EF8n007918 server111@domain.com
[download]

My goal is the have one hash with output like:

p3D8Bj1f007915 user1@domain.com sent message from server101@domain.com
p3D8EF8n007918 user2@domain.com sent message from server111@domain.com
[download]

So far, the code below only produces two simple hashes that are only uniq by msgid:

#!/bin/perl
 

open (FH, "/var/log/syslog") ;


    foreach $lin (<FH>) {

    chomp($lin) ;

    next unless $lin !~ /^\s*$/ ;           # Gets rid of blank lines

    if ($lin =~ /sendmail-listen/){ 
        $lin      =~ s/<// ;        #removes < from the line
        $lin      =~ s/>,// ;        #removes > from the line


    
        ($a,$b,$c,$d,$e,$msgid,$g)=split(' ',$lin) ;
            $msgid =~ s/:// ;

            #print $g,"\n"; #make sure i'm grabbing the to= and the fr
+om= for the follow $destname
            
            #TO
            if ($g =~ /^to/){
            ($to,$destname)=split('=',$g) ;
            $destname =~ s/,// ;
            chomp($destname) ;
            $to{$msgid} = $destname ;  #Two scalar values going into a
+ hash called $to.
                
            }

            #FROM
            if ($g =~ /^from/){
            ($fr,$fromname)=split('=',$g) ;
            $fromname =~ s/,// ;
            chomp($fromname) ;
            $from{$msgid} = $fromname ;  #Two scalar values going into
+ a hash called $from.
            }


        } #ENDS sendmail-listen

                                                                      
+                                                                     
+                          

    }

close ( FH ) ;
    #print the $to hash
    foreach $too (sort keys %to) {
       print "$too $to{$too}\n";
    }


    #print the $from hash
    foreach $fromm (sort keys %from) {
       print "$fromm $from{$fromm}\n";
    }
[download]

Thanks for looking!

Comment on sendmail logging combine two hashes Select or Download Code

Replies are listed 'Best First'.
Re: sendmail logging combine two hashes by hdb (Monsignor) on Feb 18, 2015 at 11:23 UTC
You could combine your data into one hash `%msg` when you read the file instead of creating two hashes: `$msg{$msgid}{'to'} = $destname; $msg{$msgid}{'from'} = $fromname;` [download]	[reply] [d/l] [select]
Re^2: sendmail logging combine two hashes by chanslor (Acolyte) on Feb 18, 2015 at 12:32 UTC
Thank you! I'm looking at Data::Dumper to figure how to print!	[reply]
Re^3: sendmail logging combine two hashes by hdb (Monsignor) on Feb 18, 2015 at 12:38 UTC
Try `for my $m (keys %msg) { print "$m $msg{$m}{to} send message from $msg{$m}{from}\n"; }` [download]	[reply] [d/l]
Re^4: sendmail logging combine two hashes by chanslor (Acolyte) on Feb 18, 2015 at 14:17 UTC
Re: sendmail logging combine two hashes by johngg (Canon) on Feb 18, 2015 at 14:10 UTC
`next unless $lin !~ /^\s$/ ;`* I wonder why you are using a sort-of coding double negative here. Wouldn't `next if $lin =~ /^\s*$/;` [download] be easier for human readers to parse? Just curious as to your thinking :-s Cheers, JohnGG	[reply] [d/l] [select]
Re^2: sendmail logging combine two hashes by chanslor (Acolyte) on Feb 18, 2015 at 14:24 UTC
Hello, Thanks for speaking out! I'm eager to learn and listening! So my thought should be - next if the line is blank! Thanks, I'm working to become a better perl scripter! :) `#next unless $lin !~ /^\s$/ ; # Gets rid of blank lines next if $lin =~ /^\s$/ ; # Gets rid of blank lines, Thanks JohnGG@pe +rlmonks` [download]	[reply] [d/l]