On Tuesday 09 September 2008 09:51:37 Allison Randal wrote: > jerry gay wrote:
> > can they be replaced by roles? > Potentially, yes. In the case of the scalar PMC it would make quite a > bit of sense as a role (composing in behavior common to scalar data > types). For the default PMC it makes less sense. But then, I can see > some argument there that default shouldn't be a PMC at all. When I revised vtable initialization, I reordered PMC class initialization to start at the top of the hierarchy and proceed in inheritance order. This means that PMCs can clone their parent vtables and override only the specific function pointers they provide. Abstract PMCs gave me a little trouble because they don't have vtables (they only exist as names). This isn't really a problem for core PMCs, because they can refer to functions defined in other PMCs as they all get compiled together into libparrot. That's a real violation of encapsulation for dynpmcs though, and it's the reason why we export some 12,000 functions from libparrot. In the case of MMD, we also need to initialize MMD table entries for roles and other abstract PMCs. We need a way of initializing vtable function pointers given a vtable and the name of a PMC, and then we can likely support both features. (We might even use the same mechanism for performing vtable swaps when a Key PMC stops holding an INTVAL and starts holding a STRING or a PMC, for example. Call it the ultimate in PIC.) -- c
