ooc-dev team mailing list archive

[Amos Wenger <amoswenger@xxxxxxxxx>]

Compile-time execution of code is something very nice to have - and which
would be very easy if we had an 'interpreter' backend to rock (which sounds
very interesting to code)

But anyway, meta-programming syntax deserves a much larger attention than
just a small off-topic thread in an 'ooc best practices' thread =) As far as
I'm concerned, I'll do experiments on my own, especially syntax-wise. Once
we have a good syntax, it's not too hard to implement (rock's design is not
has horrible as it could be)

Amos Wenger (nddrylliog)

On Thu, May 13, 2010 at 12:23 PM, Wander Nauta <info@xxxxxxxxxxxxxx> wrote:

> Again, greenhorn here...
>
> Maybe it's a little offtopic, but I was thinking a while back of something
> you could call 'compiler-run functions' or macros if you wish. These
> functions should look like regular ones when used, but are not converted
> into C but instead recognized by the compiler. So...
>
> if (version("windows")) {
>     // do something windows-specific
> }
>
> shouldn't converted to a real if, but is instead recognized by rock to mean
> 'Only parse the following code if compiler is on Windows'. You could also
> think of something like...
>
> if (hasFeature("ACS"))) {
>     // Use closure oneliners
> }
>
> if (!hasModule("lang/types") {
>     This really never happens, so I can put anything I want here (except a
> curly brace of course)! It's not parsed, see.
> }
>
> match statements, negation, 'and', 'or', should all work like they do
> normally.
>
> I think reusing existing 'if's and 'match'es would be cleaner then using
> yet another keyword.
>
> Again, this is just my 2c, and thank y'all reading.
>
> Wander
>
> On Wed, May 12, 2010 at 23:44, Scott Olson <scott@xxxxxxxxxxxxxxx> wrote:
>
>> I think macros should be kept as much "just ooc" as they can be. We don't
>> want
>> a separate preproccessor like C has, but macros more like Lisp, which are
>> within the language itself. I think I've heard nddrylliog mention macros
>> that
>> could do things with AST nodes, etc. They could possibly be defined just
>> like
>> functions:
>>
>>        Vector: macro { ... }
>>
>> On Wednesday 12 May 2010 06:34:15 Amos Wenger wrote:
>> > I'm not sure. #foo macro vs # comment seems to confusing / easy to make
>> a
>> > mistake.
>> >
>> > Ideally one could use # in any context, such as:
>> >
>> > Vector3f := #Vector(3, Float)
>>
>> _______________________________________________
>> Mailing list: https://launchpad.net/~ooc-dev
>> Post to     : ooc-dev@xxxxxxxxxxxxxxxxxxx
>> Unsubscribe : https://launchpad.net/~ooc-dev
>> More help   : https://help.launchpad.net/ListHelp
>>
>
>
> _______________________________________________
> Mailing list: https://launchpad.net/~ooc-dev
> Post to     : ooc-dev@xxxxxxxxxxxxxxxxxxx
> Unsubscribe : https://launchpad.net/~ooc-dev
> More help   : https://help.launchpad.net/ListHelp
>
>