> > Tuesday, February 10, 2009, 4:19:58 PM, Patrick wrote:
> >
> > > The intent of the extra $ is not to introduce a new class of
variables,
> > > but instead to say "perform the substitution as soon as the template
> > > is read" as opposed to "when it is rendered". Beyond that it's
> > > the same set of available substitutions, and intentionally includes
> > > page variables.
I'm dealing with this entirely on a theoretical basis, but there does seem
to be an inherent ambiguity here... If I'm using FmtTemplateVars() with my
own array of keys/values and one or more of those keys happens to coincide
with a PV name then you have a conflict where it seems like there should be
no conflict. Basically it seems like it would be a logical assumption that
2 separate arrays would have 2 separate "name spaces" and yet you use the
identical syntax to access both name spaces...
Interestingly enough the only place FmtTemplateVars() is used in the core is
in IncludeText() where it is used with only 2 arguments so the PVs are not
an issue at all ({$$Name} is *not* substituted).
As I say, I have no vested interest one way or the other, but it does seem
there's an undesirable conflict of namespaces that logically/intuitively
seem like they should be separate...
-Peter
_______________________________________________
pmwiki-devel mailing list
pmwiki-devel@pmichaud.com
http://www.pmichaud.com/mailman/listinfo/pmwiki-devel
