combine

   1 #!/usr/bin/perl
   2
   3 =head1 NAME
   4
   5 combine - combine sets of lines from two files using boolean operations
   6
   7 =head1 SYNOPSIS
   8
   9 combine file1 and file2
  10
  11 combine file1 not file2
  12
  13 combine file1 or file2
  14
  15 combine file1 xor file2
  16
  17 _ file1 and file2 _
  18
  19 _ file1 not file2 _
  20
  21 _ file1 or file2 _
  22
  23 _ file1 xor file2 _
  24
  25 =head1 DESCRIPTION
  26
  27 B<combine> combines the lines in two files. Depending on the boolean
  28 operation specified, the contents will be combined in different ways:
  29
  30 =over 4
  31
  32 =item and
  33
  34 Outputs lines that are in file1 if they are also present in file2.
  35
  36 =item not
  37
  38 Outputs lines that are in file1 but not in file2.
  39
  40 =item or
  41
  42 Outputs lines that are in file1 or file2.
  43
  44 =item xor
  45
  46 Outputs lines that are in either file1 or file2, but not in both files.
  47
  48 =back
  49
  50 "-" can be specified for either file to read stdin for that file.
  51
  52 The input files need not be sorted, and the lines are output in the order
  53 they occur in file1 (followed by the order they occur in file2 for the two
  54 "or" operations). Bear in mind that this means that the operations are not
  55 commutative; "a and b" will not necessarily be the same as "b and a". To
  56 obtain commutative behavior sort and uniq the result.
  57
  58 Note that this program can be installed as "_" to allow for the syntactic
  59 sugar shown in the latter half of the synopsis (similar to the test/[
  60 command). It is not currently installed as "_" by default, but you can
  61 alias it to that if you like.
  62
  63 =head1 SEE ALSO
  64
  65 join(1)
  66
  67 =head1 AUTHOR
  68
  69 Copyright 2006 by Joey Hess <joey@kitenet.net>
  70
  71 Licensed under the GNU GPL.
  72
  73 =cut
  74
  75 use warnings;
  76 use strict;
  77
  78 sub filemap {
  79         my $file=shift;
  80         my $sub=shift;
  81
  82         open (IN, $file) || die "$file: $!\n";
  83         while (<IN>) {
  84                 chomp;
  85                 $sub->();
  86         }
  87         close IN;
  88 }
  89
  90 sub hashify {
  91         my $file=shift;
  92
  93         my %seen;
  94         filemap $file, sub { $seen{$_}++ };
  95         return \%seen;
  96 }
  97
  98 sub compare_or {
  99         my ($file1, $file2) = @_;
 100
 101         filemap $file1, sub { print "$_\n" };
 102         filemap $file2, sub { print "$_\n" };
 103 }
 104
 105 sub compare_xor {
 106         my ($file1, $file2) = @_;
 107
 108         compare_not($file1, $file2);
 109         compare_not($file2, $file1);
 110 }
 111
 112 sub compare_not {
 113         my ($file1, $file2) = @_;
 114
 115         my $seen=hashify($file2);
 116         filemap $file1, sub { print "$_\n" unless $seen->{$_} };
 117 }
 118
 119 sub compare_and {
 120         my ($file1, $file2) = @_;
 121
 122         my $seen=hashify($file2);
 123         filemap $file1, sub { print "$_\n" if $seen->{$_} };
 124 }
 125
 126 if (@ARGV >= 4 && $ARGV[3] eq "_") {
 127         delete $ARGV[3];
 128 }
 129
 130 if (@ARGV != 3) {
 131         die "Usage: combine file1 OP file2\n";
 132 }
 133
 134 my $file1=shift;
 135 my $op=lc shift;
 136 my $file2=shift;
 137
 138 if ($::{"compare_$op"}) {
 139         no strict 'refs';
 140         "compare_$op"->($file1, $file2);
 141 }
 142 else {
 143         die "unknown operation, $op\n";
 144 }