[antlr-interest] LLVM example: wiki or book?

Hugo Picado hugo.pcd at gmail.com
Wed Aug 5 14:38:55 PDT 2009


Hi Terence,

You don't actually need to put a big example, as long as it is a good
example. Independently of the specific grammar, I think it should address
common issues such as parsing, tree walk, input/output parameters in rules,
lists, use of the most important predicates such as tokens, header, member,
etc.

Additionally, one think that I've seen that is commonly asked here is how to
implement unary minus, or how to construct tree nodes with different root
tokens, etc.

You can use a small example to show all this, I think a "too much big"
example will somewhere/sometime/somehow confuse the reader.

Back to the question, perhaps your C subset is a nice choice :)

Good luck with that.

Regards,
Hugo.

On Wed, Aug 5, 2009 at 8:25 PM, Terence Parr <parrt at cs.usfca.edu> wrote:

> Hi Guys,
>
> Ok, I'm rapidly running into my deadline for the book. by the end of
> the day I hope to finish the translation and then the generation
> chapters. you will see a beta of at least one of those chapters this
> week. That leaves the sort of " wrapping up" chapter, which doesn't
> really scream "Put a big LLVM example in".
>
> To satisfy you folks, would be sufficient if I put it up on the wiki
> with some good but not perfect description as well as the full source
> code for my C subset to LLVM example?
>
> Ter
> On Jun 24, 2009, at 10:30 AM, J. Stephen Riley Silber wrote:
>
> >
> > I think Waverly's point wasn't the complexity of the grammar of the
> > language.  No one is arguing the simplicity of the Smalltalk
> > grammar.  His point was that compiling a simple imperative language
> > to LLVM will demonstrate the basics of LLVM IL generation.
> > "Everyone knows C", so it should make for a relatively simple
> > tutorial.
> >
> > I know nothing about LLVM besides what's on its project homepage,
> > but I am confident that a C subset example will allow me to ramp up
> > quickly to the point where I could implement "advanced" features
> > such as dynamic dispatch in short order.
> >
> > Ter, consider this a hearty seconding of Waverly's suggestion.  It's
> > a great idea.
> >
> > --- On Wed, 6/24/09, Damien Cassou <damien.cassou at gmail.com> wrote:
> >
> > From: Damien Cassou <damien.cassou at gmail.com>
> > Subject: Re: [antlr-interest] ok, will build an LLVM example for book
> > To: "Edwards, Waverly" <Waverly.Edwards at genesys.com>
> > Cc: "antlr-interest at antlr.org ANTLR" <antlr-interest at antlr.org>
> > Date: Wednesday, June 24, 2009, 9:47 AM
> >
> > On Wed, Jun 24, 2009 at 6:40 PM, Edwards,
> > Waverly<Waverly.Edwards at genesys.com> wrote:
> > > A Smalltalk or Java type language subset might be nice but overly
> > complex, possibly
> > > loosing your reader.
> >
> > The entire Smalltalk language represents no more than 17 AST nodes. I
> > don't consider that overly complex.
> >
> > --
> > Damien Cassou
> > http://damiencassou.seasidehosting.st
> >
> > "Lambdas are relegated to relative obscurity until Java makes them
> > popular by not having them." James Iry
> >
> > List: http://www.antlr.org/mailman/listinfo/antlr-interest
> > Unsubscribe:
> http://www.antlr.org/mailman/options/antlr-interest/your-email-address
> >
> > List: http://www.antlr.org/mailman/listinfo/antlr-interest
> > Unsubscribe:
> http://www.antlr.org/mailman/options/antlr-interest/your-email-address
>
>
> List: http://www.antlr.org/mailman/listinfo/antlr-interest
> Unsubscribe:
> http://www.antlr.org/mailman/options/antlr-interest/your-email-address
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.antlr.org/pipermail/antlr-interest/attachments/20090805/b3594e05/attachment.html 


More information about the antlr-interest mailing list