Passing dynamic arrays into C.

[Bernard Helyer]

Ah, thank you!

The problem was I read the way T[].sizeof behaves on static arrays, 
thinking I was reading the section on dynamic array properties.

On 25/04/10 21:52, Daniel Murphy wrote:
> It should work if you treat the dynamic array like a pointer and a length.
> eg:
> glBufferData(GL_ARRAY_BUFFER, (vertices[0]).sizeof * vertices.length,
>                     vertices.ptr, GL_STATIC_DRAW);
>
> It's probably because (type).sizeof gives the number of bytes to hold the
> type.
> For dynamic arrays, this is the size of the (length, pointer) pair and not
> the array itself.
>
> float.sizeof // gives 4
> int.sizeof // gives 4
> int[].sizeof // gives 8
> int[3].sizeof // gives 12
>
> class C { int[100] d; };
> C.sizeof // gives the size of the reference, not the instance.
>
> I haven't checked the sizes, but it generally follows something like that.
>
>
> "Bernard Helyer"<b.helyer at gmail.com>  wrote in message
> news:hr0veq$2269$1 at digitalmars.com...
>> I was having a problem interfacing with OpenGL from D, as demonstrated by
>> this program, written once in D, and again in C:
>>
>> http://gist.github.com/378273
>>
>> Now the 'gist' of it is, doing things like this (D):
>>
>>      glBufferData(GL_ARRAY_BUFFER, vertices.sizeof,
>>                   vertices.ptr, GL_STATIC_DRAW);
>>
>>      glVertexPointer(2, GL_FLOAT, GLfloat.sizeof * 2, cast(void*)0);
>>
>> With this data (D):
>>
>>      const GLfloat[] vertices = [-1.0f, -1.0f, 1.0f, -
>>                                   1.0f, -1.0f, 1.0f, 1.0f, 1.0f];
>>
>> and the same data in C:
>>
>>      const GLfloat vertices[] = {-1.0f, -1.0f, 1.0f,
>>                                  -1.0f, -1.0f, 1.0f, 1.0f, 1.0f};
>>
>>
>> I was getting nothing but a blank screen. I'm sure you smart folks know
>> what comes next.
>>
>>       GLfloat[] vertices = {1.0f, 1.0f};  // vertices is a dynamic array.
>>
>>       GLfloat[2] vertices = {1.0f, 1.0f};  // vertices is a static array.
>>
>> The second version enables the program to work as the C one does.
>>
>>
>> My question is, why doesn't passing the `GLfloat[] vertex ...` declaration
>> work? I've looked through the docs, and can't see anything too obvious.
>>
>>
>> Thanks.
>
>