[mtasc] Optimizations

[Nicolas Cannasse]

> I tried to write a bytecode level optimizer a while ago and also found
> that the dynamic nature stands in the way. However, it is possible to
> do stuff like a peephole optimizer or eliminate redundant jumps - I don't
> know if mtasc does something like this.
>
> e.g. you can safely optimize
> setregister r:xy
> pop
> push r:xy
> ->
> setregister r:xy

Yes it's possible, but only if there is not a jump between your setregister
and push :)
MTASC currently does not implement any special bytecode level optimizations.
(Please note that MMC does not either).

Nicolas