masses/logrulediff

#!/usr/bin/perl

use warnings;
use strict;

# Compare hit rule differences for two logfiles from similar mass-check
# runs.  Might be used after tweaking some code and checking if anything
# changed.  Assumed that the exact same corpus is used for both runs.

# $ ./logrulediff ham-hege.log.old ham-hege.log
# ham/0d055650bfee6a5a0a1b43944f73eb7bb7fa7d39 +NEW_RULE -DISAPPEARED_RULE

die "Usage: $0 <logfile1> <logfile2>"
  unless -f $ARGV[0] && -f $ARGV[1];

my %rules1;
open(IN, $ARGV[0]) or die;
while (<IN>) {
  next if /^#/;
  next unless /^[.Y]\s+-?\d+\s+(\S+)\s+(\S+)/;
  $rules1{$1}{$_}++ foreach split(',', $2);
}
close IN;

my %rules2;
open(IN, $ARGV[1]) or die;
while (<IN>) {
  next if /^#/;
  next unless /^[.Y]\s+-?\d+\s+(\S+)\s+(\S+)/;
  $rules2{$1}{$_}++ foreach split(',', $2);
}
close IN;

foreach my $f (sort keys %rules1) {
  if (!defined $rules2{$f}) {
    print STDERR "!!! $f not found in second logfile\n";
    next;
  }
  my @adds;
  my @subs;
  foreach my $rule (keys %{$rules1{$f}}) {
    next if exists $rules2{$f}{$rule};
    push @subs, $rule;
  }
  foreach my $rule (keys %{$rules2{$f}}) {
    next if exists $rules1{$f}{$rule};
    push @adds, $rule;
  }

  if (@adds + @subs > 0) {
    print "$f ";
    print "+$_ " foreach (@adds);
    print "-$_ " foreach (@subs);
    print "\n";
  }
}

foreach my $f (sort keys %rules2) {
  if (!defined $rules1{$f}) {
    print STDERR "!!! $f not found in first logfile\n";
    next;
  }
}