=head1 inside =over

[Sean M. Burke]

So I was running Pod::Simple on some test documents and noted a document 
where the author forgot to close an =over, and so it contains the whole 
rest of the document, including several =head1's.  Here's a test case:

=head2 BLOOP

Hoopbehwo!

=over

=item Stuff.  Um.

Brop.

=head1 SVUP

Myup.

=cut

And here's how it was dumped by Pod::Simple::DumpAsXML :

<Document start_line="2">
  <head2 start_line="2">
    BLOOP
  </head2>
  <Para start_line="4">
    Hoopbehwo!
  </Para>
  <over-text indent="4" start_line="6">
    <item-text start_line="8">
      Stuff. Um.
    </item-text>
    <Para start_line="10">
      Brop.
    </Para>
    <head1 start_line="12">
      SVUP
    </head1>
    <Para start_line="14">
      Myup.
    </Para>
  </over-text>
</Document>
Now, I'm thinking of saying that =headN's can't be inside =over's, so that 
if such is found, it will DWIMically close the open =over.
But going the other way, I'm thinking "no way, there's nothing wrong with 
having a =headN inside an =over!  Don't OPPRESS ME with your rules, you 
FASCIST!  ATTICA!!! ATTICA!!!!".

Anyone have any preferences either way?  I think I might be leaning toward 
the former approach (where =headN closes =over) just because I think most 
cases of people apparently having a =head1 inside an =over are actually 
mistakes and that they really meant to close the =over.
--
Sean M. Burke    http://search.cpan.org/~sburke/