The hash tables seem fine to me, but maybe we need a better example.
#lang racket
(struct cpd (x y))
(define C1 (cpd 1 "bla"))
(define C2 (cpd 1 "bla"))
(equal? C1 C2)
#f
(define ht (make-hash))
(hash-set! ht C1 1)
(hash-set! ht C2 2)
(equal? (hash-ref ht C1)
(hash-ref ht C2))
;; produces #f
On Sat, Dec 28, 2013 at 5:34 PM, Sam Tobin-Hochstadt
<sa...@cs.indiana.edu>wrote:
> While your answer is what Manfred is looking for, I still find the
> original behavior worrying. It lets you expose the internals of opaque data
> structures without having the relevant inspector.
>
> Sam
> On Dec 28, 2013 4:16 PM, "J. Ian Johnson" <i...@ccs.neu.edu> wrote:
>
>> You want to make sure your structs are inspectable. You can define your
>> structs with the #:transparent option or define your own equality relation
>> with the gen:equal+hash generic interface.
>> -Ian
>> ----- Original Message -----
>> From: Manfred Lotz <manfred.l...@arcor.de>
>> To: users@racket-lang.org
>> Sent: Sat, 28 Dec 2013 15:41:28 -0500 (EST)
>> Subject: [racket] hash table confusion
>>
>> I have a hash table (created by using make-hash) where the key is a
>> struct. This seems to work fine although the documentation says
>> The make-hash procedure creates a table where keys are compared with
>> equal?,...
>>
>> and it seems that equal? gives #f if I compare two structs having the
>> same contents:
>>
>>
>> (struct cpd (x y))
>>
>> (define C1 (cpd 1 "bla"))
>> (define C2 (cpd 1 "bla"))
>>
>> (equal? C1 C2)
>> #f
>>
>>
>> Why does my hash table work fine? In my hash handling I'm using
>> make-hash, hash-ref!, hash-set!, hash-for-each.
>>
>> --
>> Manfred
>>
>>
>> ____________________
>> Racket Users list:
>> http://lists.racket-lang.org/users
>>
>> ____________________
>> Racket Users list:
>> http://lists.racket-lang.org/users
>>
>
> ____________________
> Racket Users list:
> http://lists.racket-lang.org/users
>
>
____________________
Racket Users list:
http://lists.racket-lang.org/users
