Update: A lot to think about, and a lot to throw out.
More features are being tossed by the wayside, as we find simpler
ways to allow a programmer to intuitively infer them.
We're still here, messing around with lex and yacc.
We'll start with a c99-compliant grammar, then implement a GCC Front End,
and continue from that point.
Polymorphism and inheritance
(or just about everything else)
By "popular demand," and in accordance with our weak-typing, Arrays of any dimension
can be typedef'd (as long as proper sizes are specified).
CPrime will most likely not have templates. Unions, weak-typing, member functions,
and function-overloading is more than enough.
To cut down on repetition, CPrime will allow initializations in struct declarations.
It will be the programmer's decision whether or not
to add const to a function pointer, when initialized in a struct declaration.
Douglas Crockford refers to such functionality as "Swiss Inheritance".
Screw what I said below. The #macro and #expression
syntax will perfectly match that of standard regular expressions. Changing the "tried and true"
is just too confusing: everyone is used to the standard, and those who aren't
can get information on it everywhere. in fact, a new pocketbook for that is coming out
sometime around August 1st or 10th, if I remember correctly ... it's definitely on MY wish list :-)
The #macro preprocessor statement is near completion. We're still having trouble
with syntax and getting the most bang for the buck as far as marrying printf/scanf with
regular expression syntax.
We're back up. And we've kicked out
some more features once again in order to keep
CPrime simple, fun, and effective.
I'm sorry to everyone interested in
this project. I got lazy, careless, and
over-busy, and put this task off for a long time.
I'll make this a priority now..