[ragel-users] Re: Generating Java source code

[Colin Fleming]

Haha, I don't think that limit is going anywhere. But the patch I
provided works pretty well, and I'm hopefully going to improve it some
more soon. But even without the patch, the Mongrel parser should work
fine, it's small.

In my experience, the Java generation works ok if you don't use any of
the more esoteric features of Ragel. I haven't used the last couple of
versions though, I plan to get more involved now I'm back from
vacation.

Cheers,
Colin

On 21/11/06, Adrian Thurston <thurs... at cs.queensu.ca> wrote:
>
> Steve,
>
> I would still consider Java code generation to be experimental. A number of
> tests do pass, however there is a big change which need to be made. Colin
> Fleming reported that Java has a 64K limit on array sizes which causes
> compilation to fail for large machines. He even provided a patch, though it
> hasn't been integrated yet because I've been slowly mulling over the problem
> ... more like sulking actually. I think part of me really wants to believe
> that the limit will magically go away. Anyways, a fix should come soonish.
>
> In the meantime it may well be that Mongrel's parser is small enough that
> this limit is not hit.
>
> -Adrian
>
> Steve Shreeve wrote:
> > Ragel users (and devs),
> >
> > What is the current status of being able to generate java code? I
> > apologize if this has already been discussed at length (I didn't see
> > any previous comments).
> >
> > Does the current Java code generation work? Anyone looking at a Java
> > version of mongrel's HTTP 1.1 parser?
> >
> > Thanks,
> >
> > Steve Shreeve
> >
> >
> >
>
> >
>