[racket-dev] [racket] match in Advanced Student?

On Tue, Nov 1, 2011 at 7:47 AM, Sam Tobin-Hochstadt <samth at ccs.neu.edu> wrote:
> [switching to dev@]
>
> Probably the right way to do this is for me to add a
> `prop:match-expander' which can be added to arbitrary structure, and
> then to use that in lang/private/firstorder on the `fo' struct.

I've now added `prop:match-expander'. The way to fix this problem (for
`posn' as well as `color') is to create a new sub-structure of `fo',
with an extra field that holds the match expander and the
`prop:match-expander' property pointing to that field.  Then `posn'
and `color' can use that new sub-structure.

Does that make sense?

> On Mon, Oct 31, 2011 at 11:41 PM, Jay McCarthy <jay.mccarthy at gmail.com> wrote:
>> I think extending define-primitive the right thing to do... just
>> saying it isn't what I did.
>>
>> On Mon, Oct 31, 2011 at 9:02 PM, Robby Findler
>> <robby at eecs.northwestern.edu> wrote:
>>> Are you referring to the special case on line 2730 of teach.rkt when
>>> you say posns?
>>>
>>> I think that means that I'd have to make the teaching language depend
>>> on 2htdp/image, which doesn't seem right.
>>>
>>> I would have thought that the right thing would be to extend the
>>> define-primitive protocol so that I can declare that 'color' is both a
>>> function and a struct constructor and then match would not have to
>>> special case 'color'. No?
>>>
>>> Robby
>>>
>>> On Mon, Oct 31, 2011 at 9:48 PM, Jay McCarthy <jay.mccarthy at gmail.com> wrote:
>>>> The HtDP match implementation has a list of such things and goes and
>>>> finds them to turn them back into the non-primitive bound things.
>>>> That's why it works with posns.
>>>>
>>>> Something better could be done, but for now you could add it to the white list.
>>>>
>>>> Jay
>>>>
>>>> On Mon, Oct 31, 2011 at 8:45 PM, Robby Findler
>>>> <robby at eecs.northwestern.edu> wrote:
>>>>> Oh, I see why. 'color' is bound using define-primitive in order to
>>>>> cooperate with the beginner language (and so that 'color' can be a
>>>>> constructor). But doing that interferes with the usual information
>>>>> bound by a struct that match picks up.
>>>>>
>>>>> I'm not sure what the right way to resolve this is.
>>>>>
>>>>> Robby
>>>>>
>>>>> On Mon, Oct 31, 2011 at 9:39 PM, Robby Findler
>>>>> <robby at eecs.northwestern.edu> wrote:
>>>>>> Is make-color coming from 2htdp/image? I see that's not working (not
>>>>>> sure why yet).
>>>>>>
>>>>>> Robby
>>>>>>
>>>>>> On Mon, Oct 31, 2011 at 9:34 PM, Adam Shaw <adamshaw at cs.uchicago.edu> wrote:
>>>>>>> Thanks! That still doesn't work for color:
>>>>>>>   match: color does not refer to a structure definition
>>>>>>> although it does work with my own custom color struct:
>>>>>>> (define-struct clr (r g b))
>>>>>>> (match (make-clr 1 2 3)
>>>>>>>   [(struct clr (r g b)) (* r g b)])
>>>>>>>   - Adam
>>>>>>> On Oct 31, 2011, at 9:26 PM, Jay McCarthy wrote:
>>>>>>>
>>>>>>> One of the branches of the pattern language is
>>>>>>>
>>>>>>> | (struct id (pattern ...))
>>>>>>>
>>>>>>> Your pattern should be
>>>>>>>
>>>>>>> (struct color (r g b))
>>>>>>>
>>>>>>>
>>>>>>> _________________________________________________
>>>>>>>  For list-related administrative tasks:
>>>>>>>  http://lists.racket-lang.org/listinfo/users
>>>>>>>
>>>>>>
>>>>>
>>>>
>>>>
>>>>
>>>> --
>>>> Jay McCarthy <jay at cs.byu.edu>
>>>> Assistant Professor / Brigham Young University
>>>> http://faculty.cs.byu.edu/~jay
>>>>
>>>> "The glory of God is Intelligence" - D&C 93
>>>>
>>>
>>
>>
>>
>> --
>> Jay McCarthy <jay at cs.byu.edu>
>> Assistant Professor / Brigham Young University
>> http://faculty.cs.byu.edu/~jay
>>
>> "The glory of God is Intelligence" - D&C 93
>>
>> _________________________________________________
>>  For list-related administrative tasks:
>>  http://lists.racket-lang.org/listinfo/users
>
>
>
> --
> sam th
> samth at ccs.neu.edu
>



-- 
sam th
samth at ccs.neu.edu