Ruby with C

I am new at programming ruby with C. I am trying to pass in a value from
a test ruby program to my C class and trying to print it out. But when i
print it out i am getting the wrong number printerd out.

Here is what i am trying to do.

Here is the ruby test file

class TestTest < Test::Unit::TestCase
  def test_test
    k = PKey.new('4')
    assert_equal(Object, PKey.superclass)
    assert_equal(PKey, k.class)
    trtying = k.bit(5)
  end
end

Here is the C file

static VALUE t_bit(VALUE self, VALUE obj)
{
    VALUE n;

    n = NUM2INT(obj);
    printf("Fun %d", INT2NUM(n));
    return Qnil;
}

Any suggestion would be helpfull. Thanks

···

--
Posted via http://www.ruby-forum.com/.

Hi.

By putting the variable 'n' through the INT2NUM macro as you pass it
into printf you're re-converting it back into a Ruby VALUE which is
not what you want to do. It will give you unpredictable (or at least
more-difficult-to-predict :wink: ) results...

Assuming you just want to print the value of 'obj' as an integer, you
should probably declare 'n' as an integer instead of as a VALUE, and
use the NUM2INT (as you have) to convert the Ruby VALUE 'obj' into a
plain int. Then you can pass it straight into printf (as you would
with any other int) ... no need for INT2NUM there.

hope that helps,
-Harold

···

On 3/25/07, dogwasstar@hotmail.com <dogwasstar@hotmail.com> wrote:

Here is the C file

static VALUE t_bit(VALUE self, VALUE obj)
{
    VALUE n;

    n = NUM2INT(obj);
    printf("Fun %d", INT2NUM(n));
    return Qnil;
}

Your question has already been answered. Alternatively, use inline to learn without all the hassle (no makefile, no builds, no type conversions, no time wasted). Just try out what you want to do straight up in ruby with inlined C, and then go look at what the C code is producing:

#!/usr/local/bin/ruby

require 'rubygems'
require 'inline'

class PKey
   inline do |builder|
     builder.c <<-EOM
       static void bit(int n) {
         printf("Fun %d\\n", n);
       }
     EOM
   end
end

k = PKey.new
k.bit(5)

···

############################################################
# Produces:
#
# #include "ruby.h"
#
# # line 9 "./blah.rb"
# static VALUE bit(VALUE self, VALUE _n) {
# int n = FIX2INT(_n);
#
# printf("Fun %d\n", n);
# return Qnil;
# }
#
# #ifdef __cplusplus
# extern "C" {
# #endif
# void Init_Inline_PKey_f671() {
# VALUE c = rb_cObject;
# c = rb_const_get_at(c,rb_intern("PKey"));
# rb_define_method(c, "bit", (VALUE(*)(ANYARGS))bit, 1);
#
# }
# #ifdef __cplusplus
# }
# #endif

You mean C++ class?
C doesn't have objects or classes.

···

On Mar 25, 2007, at 8:54 AM, Harold Hausman wrote:

On 3/25/07, dogwasstar@hotmail.com <dogwasstar@hotmail.com> wrote:

Here is the C file

static VALUE t_bit(VALUE self, VALUE obj)
{
    VALUE n;

    n = NUM2INT(obj);
    printf("Fun %d", INT2NUM(n));
    return Qnil;
}

Hi.

By putting the variable 'n' through the INT2NUM macro as you pass it
into printf you're re-converting it back into a Ruby VALUE which is
not what you want to do. It will give you unpredictable (or at least
more-difficult-to-predict :wink: ) results...

Assuming you just want to print the value of 'obj' as an integer, you
should probably declare 'n' as an integer instead of as a VALUE, and
use the NUM2INT (as you have) to convert the Ruby VALUE 'obj' into a
plain int. Then you can pass it straight into printf (as you would
with any other int) ... no need for INT2NUM there.

hope that helps,
-Harold

> Here is the C file
>
> static VALUE t_bit(VALUE self, VALUE obj)
> {

    int n = NUM2INT(obj);
    printf("Fun %d", n);

> return Qnil;
> }
hope that helps,
-Harold

Exactly what Harold said, but a little code in case that helps.

pth

···

On 3/24/07, Harold Hausman <hhausman@gmail.com> wrote:

On 3/25/07, dogwasstar@hotmail.com <dogwasstar@hotmail.com> wrote: