#!/usr/bin/perl -w
#
#  Copyright 1998 DevDaily Interactive.  All Rights Reserved.
#
#---------------------------------------------------------------------#
#  PROGRAM:  readlog2.pl
#
#  PURPOSE:  Read an Apache access_log file in standard ECLF format,
#            and print out the number of hits counted for each
#            $clientAddress recorded.
#---------------------------------------------------------------------#

#------------------------------------------------------------------------------#
#  Global variables that control the program action and output.                #
#------------------------------------------------------------------------------#

$PRINT_HITS_BY_ADDRESS = 1;     # set 'true' to see num hits by IP address
$NUM_RECS_TO_PRINT     = 50;    # num output recs to print per section


#------------------------------------------------------------------------------#


sub usage {
   print STDERR "\n\tUsage:  readlog2.pl access_log_file > output_file\n";
}


#--------------------------<<   main   >>-----------------------------#

   #------------------------------------------------------------#
   #  First, make sure the program is invoked with the correct  #
   #  number of command-line arguments.                         #
   #------------------------------------------------------------#

   $numArgs = $#ARGV + 1;

   if ($numArgs != 1) {
      &usage;
      exit 1;
   }

   $logFile = $ARGV[0];

   open (LOGFILE,"$logFile") || die "  Error opening log file $logFile.\n";


   #----------------------------------------------------------------#
   #  Begin processing the LOGFILE, record-by-record, until the     #
   #  end of the file.  Create a hash to store the number of times  #
   #  we've been accessed from each $clientAddress.                 #
   #----------------------------------------------------------------#

   while(<LOGFILE>) {

         chomp;

	 #----------------------------------------------#
	 #  condense one or more whitespace characters  #
	 #  to one single space                         #
	 #----------------------------------------------#

         s/\s+/ /go;

         #----------------------------------------------------------#
	 #  the next line breaks each line of the access_log into   #
	 #  nine variables                                          #
	 #----------------------------------------------------------#

         ($clientAddress,    $rfc1413,      $username, 
	 $localTime,         $httpRequest,  $statusCode, 
	 $bytesSentToClient, $referer,      $clientSoftware) =
         /^(\S+) (\S+) (\S+) \[(.+)\] \"(.+)\" (\S+) (\S+) \"(.*)\" \"(.*)\"/o;

	 #------------------------------------------------------------#
	 #  Count the number of hits to the web site by each          #
	 #  client address.  This way you can tell how many hits to   #
	 #  attribute to each client.                                 #
	 #------------------------------------------------------------#
	 #  @numHits is a 'hash'                                      #
	 #------------------------------------------------------------#

	 $numHits{$clientAddress}++;

   }

   close (LOGFILE);

   #--------------------------------------------#
   #  Output the number of hits per IP address  #
   #--------------------------------------------#

   print "NUMBER OF HITS PER IP ADDRESS:\n";
   print "------------------------------\n\n";
   $count=0;
   foreach $key (sort {$numHits{$b} <=> $numHits{$a}} (keys(%numHits))) {
      last if ($count >= $NUM_RECS_TO_PRINT);
      print "$numHits{$key} \t\t $key\n";
      $count++;
   }
   print "\n\n";