File::DirCompare (3pm)

  use File::DirCompare;

  # Simple diff -r --brief replacement
  use File::Basename;
  File::DirCompare->compare($dir1, $dir2, sub {
    my ($a, $b) = @_;
    if (! $b) {
      printf "Only in %s: %s\n", dirname($a), basename($a);
    } elsif (! $a) {
      printf "Only in %s: %s\n", dirname($b), basename($b);
    } else {
      print "Files $a and $b differ\n";
    }
  });

  # Version-control like Deleted/Added/Modified listing
  my (@listing, @modified);     # use closure to collect results
  File::DirCompare->compare('old_tree', 'new_tree', sub {
    my ($a, $b) = @_;
    if (! $b) {
      push @listing, "D   $a";
    } elsif (! $a) {
      push @listing, "A   $b";
    } else {
      if (-f $a && -f $b) {
        push @listing, "M   $b";
        push @modified, $b;
      } else {
        # One file, one directory - treat as delete + add
        push @listing, "D   $a";
        push @listing, "A   $b";
      }
    }
  });

File::DirCompare is a perl module to compare two directories using a callback, invoked for all files that are 'different' between the two directories, and for any files that exist only in one or other directory ('unique' files).

File::DirCompare has a single public compare() method, with the following signature:

File::DirCompare->compare($dir1, $dir2, $sub, $opts);

The first three arguments are required - $dir1 and $dir2 are paths to the two directories to be compared, and $sub is the subroutine reference called for all unique or different files. $opts is an optional hashref of options - see \s-1OPTIONS\s0 below.

The provided subroutine is called for all unique files, and for every pair of 'different' files encountered, with the following signature:

$sub->($file1, $file2)

where $file1 and $file2 are the paths to the two files. For 'unique' files i.e. where a file exists in only one directory, the subroutine is called with the other argument 'undef' i.e. for:

$sub->($file1, undef) $sub->(undef, $file2)

the first indicates $file1 exists only in the first directory given ($dir1), and the second indicates $file2 exists only in the second directory given ($dir2).

\s-1OPTIONS\s0

The following optional arguments are supported, passed in using a hash reference after the three required arguments to compare() e.g.

File::DirCompare->compare($dir1, $dir2, $sub, { cmp => $cmp_sub, ignore_cmp => 1, ignore_unique => 1, matches => $matches_sub, });

cmp

By default, two files are regarded as different if their contents do not match (tested with File::Compare::compare). That default behaviour can be overridden by providing a 'cmp' subroutine to do the file comparison, returning zero if the two files are equal, and non-zero if not. E.g. to compare using modification times instead of file contents: File::DirCompare->compare($dir1, $dir2, $sub, { cmp => sub { -M $_[0] <=> -M $_[1] }, });

ignore_cmp

If you want to see all corresponding files, not just 'different' ones, set the 'ignore_cmp' flag to tell File::DirCompare to skip its file comparison checks i.e. File::DirCompare->compare($dir1, $dir2, $sub, { ignore_cmp => 1 });

ignore_unique

If you want to ignore files that only exist in one of the two directories, set the 'ignore_unique' flag i.e. File::DirCompare->compare($dir1, $dir2, $sub, { ignore_unique => 1 });

matches

Subroutine to be called for file pairs that match, with the following signature: $sub->($file1, $file2) These pairs are ordinarily ignored (unless \*(C`ignore_cmp\*(C' is set).

File::Dircmp, which provides similar functionality (and whose directory walking code I've adapted for this module), but a simpler reporting-only interface, something like the first example in the \s-1SYNOPSIS\s0 above.

Gavin Carr <[email protected]>

Thanks to Robin Barker for a bug report and fix for glob problems with whitespace.

Copyright 2006-2012 by Gavin Carr <[email protected]>.

This library is free software; you can redistribute it and/or modify it under the same terms as Perl itself.