On 4 Aug 2000, Russ Allbery wrote:
> Tim Jenness <[EMAIL PROTECTED]> writes:
>
> > It's clear that a generic translator class should also be able to handle
> > the automatic podifying of variables and functions etc.
>
> It can't. The markup for such things doesn't correspond to any standard
> POD sequences. Consider foo(1) in Pod::Man, for example.
>
Hang on. I still haven't worked out why. From the pod docs:
I<text> Italicize text, used for emphasis or variables
B<text> Embolden text, used for switches and programs
S<text> Text contains non-breaking spaces
C<code> Render code in a typewriter font, or give some other
indication that this represents program text
L<name> A link (cross reference) to name
L<name> manual page
and
In particular, you can leave things like this verbatim in
your text:
Perl
FILEHANDLE
$variable
function()
manpage(3r)
This suggests to me that $variable should be written as if the author had
written I<$variable>, manpage(3r) should be written as L<manpage(3r)> etc.
[with the L<> construct distinguishing between perl man pages and generic
manpages]
Are you saying that is not the case? That writing $variable and
I<$variable> imply two different states of translation? If that is true
then where is it documented? If the former is the case then you simply
have to write a generic pre-processor that looks through the pod and
inserts the correct L<> C<> or whatever before Pod::Parser sees it.
I really really really want to know what the issue is here :-)
--
Tim Jenness
JCMT software engineer/Support scientist
http://www.jach.hawaii.edu/~timj
