[racket] Understanding lexical context

From: Greg Hendershott (greghendershott at gmail.com)
Date: Tue Oct 1 14:42:30 EDT 2013

Thanks for the quick reply!

On Tue, Oct 1, 2013 at 2:13 PM, Sam Tobin-Hochstadt
<samth at cs.indiana.edu> wrote:
> On Tue, Oct 1, 2013 at 2:01 PM, Greg Hendershott
> <greghendershott at gmail.com> wrote:
>>
>> tl;dr: Can anyone suggest how I could improve my
>> understanding of this?  Enough that I could improve Fear of
>> Macros and help others, too? Thanks in advance.
>
> If you've read the resources you described, then you're definitely
> prepared to read [Macros that work together], which lays out all the
> underpinnings of the Racket macro system (and comes with a redex
> model).

Oh, I've read that, too. I should have thought to to re-read that one,
as well. Thanks.

(I'm not sure a redex model will increase its explanatory value for
_me_, but...)

> For this specific macro, here's the answer. In your situation, we can
> look at all of the syntax objects you tried, and where they're from
> (note that the macro stepper would help you here -- look at the
> foreground colors of the identifiers):

Ah. I hadn't grokked the meaning of the colors before.

> `stx`: that's the whole thing, so you're taking the lexical context of
> the () around `inner` in the definition of `outer`.  Not what you
> want.
> `k`: that's the occurrence of `inner` in the definition of `outer`.
> Still not what we want.
> `a`: that's the occurrence of `get` in the definition of `outer`.
> Still not what we want.
>
> `body0`: that's `"hi"`, which is in the _input_ to `outer`. So now we
> have lexical context from the use site, which is what we want.  Yay!
>
> However, that's not really the right solution.  What if we abstracted
> over `outer`:
>
>     (define-syntax-rule (define-as-hi id) (outer id "hi"))
>
> If we use that, it still won't work, because `"hi"` is now in the
> definition of `define-as-hi`, which isn't in the input when you use
> the macro.  So the right choice for the lexical context is probably
> `b`.

Well, two things about that:

1. For me, using #'body0 _does_ still work in this example.

2. Plus it seems like it would be really unfortunate if carefully
choosing one of #'a or #'b were required -- because that would mean
`inner` couldn't be written to be usable by other macros independent
of knowing the origin of what they give to `inner`, right?  I mean,
without knowing if a comes from `outer` or from the usage of `outer`,
and likewise for b, and therefore which of a or b to choose carefully.

Anyway I'll re-read the paper.

Thanks again!

Posted on the users mailing list.