Re: [racket-users] ~literal vs ~datum

[Shu-Hung You]

I think the case here is a little subtle. The syntax object #'(define
x y) has actually gone through expansion, but Racket's new expansion
model will prune this syntax object's local scopes [1]. One could use
#:local [2] in this case to see the difference:

> (syntax-parse (let ([define 'something-else]) (quote-syntax (define x y)))
[((~literal define) var:id e:expr) 'yes]
[_ 'no])
'yes

> (syntax-parse (let ([define 'something-else]) (quote-syntax (define x y) 
> #:local))
[((~literal define) var:id e:expr) 'yes]
[_ 'no])
'no

The original doc predates the new expander and the result was probably
'no in the old expansion model.

Best,
Shu-Hung

1. 
http://www.cs.utah.edu/plt/scope-sets/general-macros.html#%28part._.Local_.Bindings_and_.Syntax_.Quoting%29

2. https://docs.racket-lang.org/reference/Syntax_Quoting__quote-syntax.html


On Thu, Aug 3, 2017 at 1:53 PM, Jordan Johnson  wrote:
> On Aug 1, 2017, at 11:34, Jens Axel Søgaard  wrote:
>
> The binding information is attached a syntax object during expansion.
> In your example the #'(define x y) haven't gone through expansion,
> so you do not get what your expected result.
>
> However we can change your macro a little...
>
>
> Ah, that clarifies everything. Thanks!
>
> So, the example I quoted is actually from the docs on ~datum, and it seems
> pretty clear that it’s trying unsuccessfully to elicit different behaviors
> that show the difference between ~literal and ~datum. Based on the code
> you’ve provided, I think this would be a better illustration:
>
> (define-syntax (is-define? stx)
>   (syntax-parse stx
> [(_is-define? id)
>  (syntax-parse #'id
>[(~literal define) #''yes]
>[(~datum   define) #''not-really]
>[_ #''not-even-close])]))
>
> (is-define? define) ;; 'yes
>
> (let ([define 42])
>   (is-define? define))  ;; 'not-really
>
> (is-define? something-else) ;; 'not-even-close
>
>
> I’ve submitted that as a PR on the doc file linked above. I hope you don’t
> mind my using your example in that way. It was very helpful.
>
> Regards,
> Jordan
>
> --
> You received this message because you are subscribed to the Google Groups
> "Racket Users" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to racket-users+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [racket-users] ~literal vs ~datum

[Jordan Johnson]

On Aug 1, 2017, at 11:34, Jens Axel Søgaard  wrote:
> The binding information is attached a syntax object during expansion.
> In your example the #'(define x y) haven't gone through expansion,
> so you do not get what your expected result.
> 
> However we can change your macro a little...

Ah, that clarifies everything. Thanks!

So, the example I quoted is actually from the docs on ~datum 
,
 and it seems pretty clear that it’s trying unsuccessfully to elicit different 
behaviors that show the difference between ~literal and ~datum. Based on the 
code you’ve provided, I think this would be a better illustration:

> (define-syntax (is-define? stx)
>   (syntax-parse stx
> [(_is-define? id)
>  (syntax-parse #'id
>[(~literal define) #''yes]
>[(~datum   define) #''not-really]
>[_ #''not-even-close])]))
> 
> (is-define? define) ;; 'yes
> 
> (let ([define 42])
>   (is-define? define))  ;; 'not-really
> 
> (is-define? something-else) ;; 'not-even-close

I’ve submitted that as a PR on the doc file linked above. I hope you don’t mind 
my using your example in that way. It was very helpful.

Regards,
Jordan

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [racket-users] ~literal vs ~datum

[Jens Axel Søgaard]

2017-08-01 20:11 GMT+02:00 Jordan Johnson :

> Hi all,
>
> Reading the documentation on ~datum in syntax/parse, I see:
>
> The ~datum form is useful for recognizing identifiers symbolically, in
> contrast to the ~literal form, which recognizes them by binding.
> That makes sense to me. But then I see:
>
> > (syntax-parse (let
> 
>  ([define
> 
>  'something-else]) #'(define
> 
>  x y))
> [((~datum define
> 
> ) var:id e:expr) 'yes]
> [_
> 
>  'no])
>
> 'yes
> > (syntax-parse (let
> 
>  ([define
> 
>  'something-else]) #'(define
> 
>  x y))
> [((~literal define
> 
> ) var:id e:expr) 'yes]
> [_
> 
>  'no])
>
> 'yes
>
> I’m confused by this. It seems, based on the first sentence I quoted, that
> in the second example, in which the syntax being parsed has *define* bound
> to something other than the usual, the use of *~literal* should make the
> first clause fail to match. And I’d have expected the examples to contrast
> *~datum* and *~literal*, rather than exhibit the same behavior.
>

The binding information is attached a syntax object during expansion.
In your example the #'(define x y) haven't gone through expansion,
so you do not get what your expected result.

However we can change your macro a little:


#lang racket
(require (for-syntax syntax/parse))

(define-syntax (is-define? stx)
  (syntax-parse stx
[(_is-define? id)
 (syntax-parse #'id
   [(~literal define) #''yes]
   [_ #''no])]))

(is-define? define)

(let ([define 42])
  (is-define? define))

This will result in

'yes
'no


/Jens Axel

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[racket-users] ~literal vs ~datum

[Jordan Johnson]

Hi all,

Reading the documentation on ~datum in syntax/parse, I see:

The ~datum 

 form is useful for recognizing identifiers symbolically, in contrast to the 
~literal 

 form, which recognizes them by binding.

That makes sense to me. But then I see:

> (syntax-parse 
> 
>  (let 
> 
>  ([define 
> 
>  'something-else]) #'(define 
> 
>  x y))
[((~datum 

 define 
)
 var:id e:expr) 'yes]
[_ 

 'no])
'yes
> (syntax-parse 
> 
>  (let 
> 
>  ([define 
> 
>  'something-else]) #'(define 
> 
>  x y))
[((~literal 

 define 
)
 var:id e:expr) 'yes]
[_ 

 'no])
'yes

I’m confused by this. It seems, based on the first sentence I quoted, that in 
the second example, in which the syntax being parsed has define bound to 
something other than the usual, the use of ~literal should make the first 
clause fail to match. And I’d have expected the examples to contrast ~datum and 
~literal, rather than exhibit the same behavior.

Could someone please explain this to me?

Thanks,
Jordan

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.