allocate array with new

[Dmitry Olshansky]

On 15.05.2012 17:12, Ondrej Pokorny wrote:
> On Tuesday, 15 May 2012 at 12:36:30 UTC, Dmitry Olshansky wrote:
>> On 15.05.2012 16:27, Ondrej Pokorny wrote:
>>> On Tuesday, 15 May 2012 at 09:44:08 UTC, Jonathan M Davis wrote:
>>>> On Tuesday, May 15, 2012 11:26:49 Namespace wrote:
>>>>> On Tuesday, 15 May 2012 at 09:23:51 UTC, Kagamin wrote:
>>>>> > Difference with what?
>>>>> > new is a safe feature: it allocates in the GC heap
>>>>>
>>>>> That's what i mean. So i have to delete it yourself with "delete
>>>>> arr;", or not?
>>>>
>>>> No. _Never_ use delete. It's going to be deprecated. The GC worries
>>>> about
>>>> freeing memory allocated on the GC heap, and new always allocates on
>>>> the GC
>>>> heap. If you don't want to allocate on the GC heap, then use malloc
>>>> and free,
>>>> in which case you _do_ need worry about freeing the memory.
>>>>
>>>> If you need to force destruction before the GC collects an object, you
>>>> can
>>>> call clear on that object to have its destructor called and its vtbl
>>>> zeroed
>>>> out, but it's memory still isn't freed. That's the GC's job.
>>>>
>>>> If you really have to, you can use core.memory to manipulate the GC
>>>> heap
>>>> (including calling GC.free), but you really shouldn't be messing with
>>>> any of
>>>> that unless you really need to and you know what you're doing.
>>>>
>>>> - Jonathan M Davis
>>>
>>> Hi,
>>> does this hold for structs too?
>>>
>>> struct H
>>> {
>>> this(int a){writeln("ctor");}
>>> ~this(){writeln("dtor");}
>>> };
>>>
>>> ...
>>>
>>> H* h = new H(5);
>>> clear(h);
>>> ...
>>>
>>> output:
>>>
>>> ctor
>>>
>>> seems like destructor is not called.
>>>
>>> if I change declaration of H to class H. output is following:
>>>
>>> ctor
>>> dtor
>>>
>>> I tried to create object according to RAII idiom and in case of struct H
>>> my file handle remained open and I was still able to write to file...
>>>
>>> Ondrej
>>>
>> I thought in C++ RAII is about (i.e. even in C++ no heap allocation):
>> H h = H(5);
>>
>> Same works in D. A call to clear in you code above doesn't call
>> destructor, it only zeros out pointer.
>> If you absolutely need pointers & heap and yet manual memory managment
>> use:
>> clear(*h);
>>
>>
>> Explanation:
>> clear(x) calls x.__dtor if x is struct or class, then assigns
>> x = T.init;
>>
>> A better way might be to just check if x.__dtor is callbale (thus
>> pointer to sstruct will also work).
>
> I don't need this stuff I am playing with language and wonder
> why. Originally I had there auto h = H(5); and after changing H
> to struct I did not realized why that happened.
>
> Thanks a lot for explanation, now it makes perfect sense, but I
> have still little bit unpleasant felling about it because of
> behavior of ref to classes vs pointers to structs with clear.
> Both are dynamically allocated data, so I was expecting same
> response from clear(x).
>

Like I said it's easily fixable at a cost of breaking some code that 
relies on this clear(ptr) zeroing out pointer. I think that it's wrong 
code anyway. So I I'll wrap up a pull request to do just that.


-- 
Dmitry Olshansky