That makes the discussion more sensible to me, thanks!
Well, the problem is akin to using a pointer as an argument to a function
accepting a const-pointer (in C++), which is an error. I would expect BitC to
be at least as stringent!
chain = S(4, &container.s) // chain-->[4, *]---->[5, null]
~~~~~~~~
Type Error: Can't initialize a const reference with a
mutable reference.
PKE
________________________________________
From: [email protected] [[email protected]] On Behalf Of
Jonathan S. Shapiro [[email protected]]
Sent: Thursday, March 24, 2011 1:46 PM
To: Discussions about the BitC language
Subject: Re: [bitc-dev] Mutability, again
Pal:
It was *critical* to the example that the 's' field in the container was *not*
a reference field and therefore not a nullable field. Your analysis relies on
changing that type in a way that breaks the example.
However, there is a further syntactic problem in the example. /container/ is a
reference, so it might have been clearer if I had written it as
let container = Container(S(5, null)) // container-->[5, null]
chain = S(4, &container.s) // chain-->[4, *]---->[5, null]
len = S_length(chain) // returns 2 :
in
container.s := S(6,S(7,null)) // container-->[6, *]-->[7, null]
// chain-->[4, *]-->[6,*] -->[5, null]
Note the addition of the '&' in the initialization of chain. My intention is
that this is capturing an inner reference, not a pointer.
shap
_______________________________________________
bitc-dev mailing list
[email protected]
http://www.coyotos.org/mailman/listinfo/bitc-dev
