On Wed, Aug 29, 2001 at 12:14:55PM -0700, Russ Allbery wrote:
> Kingpin <[EMAIL PROTECTED]> writes:
> 
> > I think Pod::Man should NOT read anything after __DATA__.  After
> > __END__, yes, but not after __DATA__.

Thanks for the forwarding this Russ

I can understand the problem you are having. It must be pretty annoying.
Is it Pod::Checker that is complaining, or is it Pod::Main, or is it Pod::Parser (do 
we know which module is issuing the message at least)

When I was writing Pod::Parser, I was designing it to parse
PODs without having to know anything about what kind of
file those PODs were (or weren't) embedded inside of. This is
largely because I needed to use it for several other tasks that
really didn't have anything to do with trying translate PODs
into some output format. I was using PODs in shell scripts,
tool configuration files, and other such things.

For that reason, I did not want it to have to know anything
about Perl syntax, only about POD syntax. At once point I had
to make a special case to look at '->' and '=>' inside X<...>
(but not --> or ==>). I think I managed to remove that when
X<< ... >> was allowed (but I'm not positive).

If it has to look for __DATA__ then does it have to look for
some especially thorny cases where something might look like
'__DATA__' but doesn't really start the __DATA__ section of a
Perl file? (Does it always have to be at the very beginning
of the line? are there ways to fool POD that don't fool Perl
that Pod::Parser would have to jump thru hoops to figure out?).

Is the issue you are having simply the error message? Or is
it not generating proper output? If there was an option to
suppress certain categories of diagnostics, would that be
sufficient for your needs?


> > My distribution I18N::Charset generates an error because I have lots
> > of ===== in my __DATA__ section.
> 
> I've taken the liberty of forwarding this to [EMAIL PROTECTED] because
> this isn't really a Pod::Man issue, more a Pod::Parser issue, and more
> generally a question of what the intended semantics of POD should be.
> Pod::Man just uses Pod::Parser and will follow whatever decision is made
> there.
> 
> (I think it's a lot simpler to just parse the entire file, but I don't
> know how to solve the problem that you identify above.)
> 
> -- 
> Russ Allbery ([EMAIL PROTECTED])             <http://www.eyrie.org/~eagle/>

Reply via email to