Re: [dylan] An experiment with IntelliJ

Bruce Mitchener <[email protected]> Fri, 27 Sep 2013 18:34:18 +0700
Newsgroups gmane.comp.lang.dylan.gwydion.devel
Message-ID <CA+esKjPZbSKNBi7YbhiCaY-F8+j+-GE6L-ri01OSs+9oLG3fvA@mail.gmail.com>
On Fri, Sep 27, 2013 at 6:19 PM, Francesco Ceccon <[email protected]>wrote:

> > I plan to start lexing Dylan and then syntax highlighting for Dylan will
> work (without a full blown parser).
>
> Why not a parser? You linked me a yacc file with Dylan's grammar, I think
> it should be easy to port it to the Java equivalent. Correct me if I'm
> wrong.
>

Because it is a lot of work and requires dealing with macros to do
correctly... and that requires loading and understanding the libraries /
modules used by the current source code.


> >       • Accessing some of the data from the Dylan compiler to provide
> feedback.
>
> Can we already do this or not? If I understand it correctly you want to
> make something like libclang for Dylan.
>

I'll either tie into dswank (dime's backend) or use an HTTP-based protocol
to talk to a server in Dylan.


> >       • Starting to parse Dylan into IntelliJ's PSI (like an AST) for
> doing completion and so on.
> >       • Improving the LID file lexer and parser so that we can highlight
> files that don't exist, improper usage of the Dylan extension (a warning),
> and other IDE-like things.
>
> What do you mean by "highlighting file that don't exist"?
>

If you have a LID file, it lists the source files. You could highlight a
file that was listed that didn't exist (due to a typo, say).


> Good work so far by the way.
>

Thanks!

 - Bruce

_______________________________________________
hackers mailing list
[email protected]
https://lists.opendylan.org/mailman/listinfo/hackers