Differences between POD::Parser and Perl

[Michael Scott]

Apologies first because I'm sure this must already have been discussed.

I've recently been going over the Parrot source with podchecker to iron 
out the pod errors. I was a bit surprised to discover that what is an 
error for podchecker is not necessarily an error for perl.

Whereas perl seems simply to parse out everything from the start of a 
line beginning with = to the end of a line beginning with =cut, 
podchecker uses POD::Parser's parse_from_filehandle() to processes 
paragraphs at a time. This can lead to podchecker errors like "Spurious 
text after =cut" which perl seems to ignore.

Also, POD::Parser's approach fails to distinguish between pod in 
strings and here documents, and pod at file level. I ended up 
interpolating pod commands into strings etc in order to get podchecker 
to ignore them.

The question I have in relation to this is does perl parse pod 
correctly? The perlpod doc seems to imply that podchecker's para-based 
approach is correct. Though it doesn't seem to mention how pod in 
multiline strings will be treated.

Is perl just being lenient? I'm wondering which standard to enforce.

I'm using perl 5.8.1 on OS X.

Mike