Re: Improving merge of tricky conflicts

[Junio C Hamano <gitster@pobox.com>]

Jeff King <peff@peff.net> writes:

> sub flush {
>   print @ours;
>   print "|||||||\n";
>   show_diff(base => \@base, ours => \@ours);
>   print "|||||||\n";
>   show_diff(base => \@base, theirs => \@theirs);
>   print "=======\n";

Before this gets called, "<<<<<<<\n" from the original has been
emitted to the output, so this shows

	<<<<<<<
	version from ours
	|||||||
	output from diff -u base ours
	|||||||
	output from diff -u base theirs
	=======
	version from theirs

   print @theirs;
>   @ours = @base = @theirs = ();
> }

Unfortunately, there is no ">>>>>>>" shown with this code, as $_
seems to get clobbered before the sub returns, so ...

> sub state_theirs {
>   if (/^>{7}/) { flush(); print; $state = \&state_none }
>   else { push @theirs, $_ }
> }

... that "print" does not do what we want it to do.

Localizing the $_ upfront in the show_diff sub should probably be
sufficient.  That is ...

> sub show_diff {
>   my ($pre_name, $pre_data, $post_name, $post_data) = @_;

+  local ($_);

... here.

>
>   my $pre = File::Temp->new;
>   print $pre @$pre_data;

I am debating myself if I want to see the base version between these
two extra diffs.  Perhaps there is no need, as either one of these
two extra diffs should be sufficient to see what was in the base
version.

Thanks.