Hmm, a side question, are table keys always strings?
And if not, is the fact that i am asking for a string (lua_tostring),
all the time for the table key a problem?
One thing i noticed is the first element's key in the subtable is "1",
which I wasn't expecting, but I guess it uses numbers if you dont
supply an actual key.
-Raymond
On 1/19/07, Raymond Jacobs <raymondj@gmail.com> wrote:
> Damn, I'll have to keep that in mind, I don't think that is what this
> is, it dies on a step over lua_next, so im not letting it run rampant.
>
>
> On 1/19/07, Rici Lake <lua@ricilake.net> wrote:
> >
> > On 19-Jan-07, at 9:31 PM, Jérôme VUARAND wrote:
> >
> > > 2007/1/19, Raymond Jacobs <raymondj@gmail.com>:
> > >> int idx=LUA_GLOBALSINDEX;
> > >> lua_getfenv(l,idx);
> > >>
> > >> while(lua_next(l,idx)!=0)
> > >> {
> > >> const char* key=lua_tostring(l,-2);
> > >>
> > >> //get value at -1 index and check type of it
> > >>
> > >> //do somthing with key and value based on type of
> > >> value
> > >> //if value is a table call this function again to
> > >> traverse it, then
> > >> continue here
> > >>
> > >>
> > >> lua_pop(l,1);
> > >> }
> > >
> > > You forget to push an initial nil on the stack before entering you
> > > while loop (see lua_next documentation example).
> >
> > He did push a nil, although I suspect it was not by design:
> >
> > >> int idx=LUA_GLOBALSINDEX;
> > >> lua_getfenv(l,idx);
> >
> > Since tables don't have environments, lua_getfenv() will do the job.
> >
> > By the way, be careful with recursively enumerating tables; if there's
> > a circular reference, you'll find yourself running off the end of the C
> > stack. (And, since _G._G == _G, there is such a circular reference in
> > the standard globals table.)
> >
>
