On 02/27/2012 01:17 PM, Justin Whear wrote:
> On Mon, 27 Feb 2012 17:44:40 -0300, Pedro Lacerda wrote:
>
>> The expression "[] is null" evaluates to true here using 2.058, but I
>> expected to be false. What am I missing?
>>
>> Pedro Lacerda
>> <div>The expression"[] is null" evaluates to true here using
>> 2.058, but I expected to be false. What am I
>> missing?</div><div><br></div><div>Pedro Lacerda</div><br> The
expression
>> "[] is null" evaluates to true here using 2.058, but I expected to be
>> false. What am I missing?
>>
>> Pedro Lacerda
>> <div>The expression"[] is null" evaluates to true here using
>> 2.058, but I expected to be false. What am I
>> missing?</div><div><br></div><div>Pedro Lacerda</div><br>
>
> null makes sense to me.
But null also means uninitialized (for reference types anyway). Is a
slice that has just become empty is uninitialized?
> If the length is null, where can the ptr member
> point to other than null?
Since .ptr should not be derefenced, it can point to anywhere.
void main()
{
int[] a = [ 1, 2 ];
a = a[0..0];
// Both of these pass
assert(a is null);
assert(a.ptr !is null);
}
And I think that's another problem: If a.ptr is non-null, then the
memory occupied by the initial elements will not be freed by the GC.
> Plus, anyone coming from C uses null to signify
> empty
Not necessarily. A C data structure may contain no elements and be non-null.
> and anyone coming from Lisp uses the empty list to signify Nil.
Ali
