Re: POD doesn't capture code semantics

On 5 Mar 2008, at 18:35, David Golden wrote:
> Also, more generally, consider that formats that begin with a colon
> are treated as POD and parsed as normal (c.f. perlpodspec and the "=
> :biblio" example).  So consider this in combination with the X<> tag
> mentioned above :
>
>    =begin :method
>
>    =head2 X<C<foo>>
>
>    Documentation of foo() here...
>
>    =end :method
>
> That's all legal POD that shouldn't mess up existing parsers.


Ah. I should have RTFM. The :biblio example is exactly the kind of  
semantic tagging I'm talking about. It's a shame it's so verbose but  
already existing is a plus :)

I wonder how you'd tag a subroutine that can be called as a method or  
a regular subroutine. Is

=begin :method|sub

legal?

I can't see a definition of what characters are allowed in an  
identifier in perlpodspec or whether garbage is allowed a the end of a  
=begin line.

-- 
Andy Armstrong, Hexten