From: Tim Pease <tim.pease@...>
Date: 2008-05-06T01:34:38+09:00
Subject: Re: Help with rb_iv_set


On May 5, 2008, at 7:35 AM, Bryan Richardson wrote:

> Hi Rick,
>
> So you're saying that by doing it the way I was doing it (using
> rb_float_new) I am indeed doing it correctly?
>

To be explicit about your specific case ...

rb_iv_set(self, "@voltage", rb_float_new(1.01));


Blessings,
TwP


> --
> Thanks!
> Bryan
>
> On Sun, May 4, 2008 at 9:29 PM, Rick DeNatale  
> <rick.denatale@gmail.com>
> wrote:
>
>> On Sun, May 4, 2008 at 11:13 PM, Bryan Richardson
>> <btrichardson@gmail.com> wrote:
>>> Hello all,
>>>
>>> I'm trying to use rb_iv_set to set a double value in a C extension  
>>> I'm
>>> writing.
>>>
>>> rb_iv_set(self, "@voltage", 1.01);
>>>
>>> This doesn't seem to work, as I get a '0' instead of '1.01'.  If I  
>>> use
>>> rb_float_new(1.01), then it works.  However, the voltage value in  
>>> my C
>>> structure is a double, and rb_double_new doesn't exist.
>>>
>>> Any suggestions?
>>
>> rb_iv_set needs a VALUE for the third parameter.
>> and the value of a Ruby float is a C double:
>>
>> From the 1.8.6 source:
>>
>> in ruby.h
>>
>> VALUE rb_iv_set _((VALUE, const char*, VALUE));
>>
>> and also:
>>
>> struct RFloat {
>>   struct RBasic basic;
>>   double value;
>> };
>>
>> And from numeric.c
>>
>> VALUE
>> rb_float_new(d)
>>   double d;
>> {
>>   NEWOBJ(flt, struct RFloat);
>>   OBJSETUP(flt, rb_cFloat, T_FLOAT);
>>
>>   flt->value = d;
>>   return (VALUE)flt;
>> }
>>
>>
>> --
>> Rick DeNatale
>>
>> My blog on Ruby
>> http://talklikeaduck.denhaven2.com/
>>
>>