> > p.s. This doesn't mean i've given up on XML, or anything, just that I need
> > to learn more about it before I even begin. I'm pragmatic - i'll
> > concentrate for now on what works.
>
> So don't waste your time using these functions. Instead, create an interface
> class to do this, and in the meantime, use a replacement -- ie, a virtual class
> that initializes all the information on compile time. Why? because you don't
> need to know the format that is going to be used to write the initialization
> files. You only need the results.
>
Actually, had the 'results' ready to go for a while, i'm just
working on loading em in now :) Though - it's not been very well tested
that the pseudo results that it looks at now to guarentee a good compile
actually are a good representation of data.
I'm actually just storing a map of section names (like effect,
room flags, etc) to [ a map of affect name to values ]. I had a bit of
trouble getting msvc to let me use what is basically a map of maps, but i
just ended up encapsulating the inner map in it's own class, and included
the name of the section in it, so you wouldn't have to backtrace to find
out what goes where.
Is there a normal way to define a map of maps? I had been using
the standard templating structure:
map<string, map<string, int>> variable_name;
but instead had to resort to
map<string, MapEncapsulatingClass> variable name;
and the encap class contained a map of string to int. Just one level of
misdirection, no big deal.
I don't seem to have any luck procuring much information on the
STL, and what has to be the most common stl use (lists of lists, etc).
PjD
--
+---------------------------------------------------------------+
| FAQ: http://qsilver.queensu.ca/~fletchra/Circle/list-faq.html |
| Archives: http://post.queensu.ca/listserv/wwwarch/circle.html |
+---------------------------------------------------------------+
This archive was generated by hypermail 2b30 : 12/03/01 PST