new

The constructor takes the following forms:

  my $delt = VCS::Lite::Delta->new( '/my/file.diff',$sep); # File name
  my $delt = VCS::Lite::Delta->new( \*FILE,$sep);       # File handle
  my $delt = VCS::Lite::Delta->new( \$string,$sep); # String as scalar ref
  my $delt = VCS::Lite::Delta->new( \@foo, $id1, $id2) # Array ref

$sep here is a regexp by which to split strings into tokens. The default is to use the natural perl mechanism of $/ (which is emulated when not reading from a file). The arrayref form is assuming an array of hunks such as the output from Algorithm::Diff::diff.

The other forms assume the input is the text form of a diff listing, either in diff format, or in unified format. The input is parsed, and errors are reported.

diff

  print OUTFILE $delt->diff

This generates a standard diff format, for example:

4c4 < Now wherefore stopp'st thou me? --- > Now wherefore stoppest thou me?

udiff

  print OUTFILE $delt->udiff

This generates a unified diff (like diff -u) similar to the form in which patches are submitted.

id

  my ($id1,$id2) = $delt->id;
  $delt2->id('foo.pl@@1','foo.pl@@3')

The id method allows get and set of the names associated with the two elements being diffed. The id is set for delta objects returned by VCS::Lite->diff, to the element IDs of the VCS::Lite objects being diffed.

Diff format omits the file names, hence the IDs will not be populated by new. This is not the case with diff -u format, which includes the file names which are passed in and available as IDs.

hunks

  my @hunklist = $delt->hunks

A hunk is a technical term for a section of input containing a difference. Each hunk is an arrayref, containing the block of lines. Each line is itself an arrayref, for example:

  [
    [ '+', 9, 'use Acme::Foo;'],
    [ '-', 9, 'use Acme::Bar;'],
  ]

See the documentation on Algorithm::Diff for more details of this structure.