Where is Fixnum#power! coming from?


#1

Hi all,

Ruby 1.8.4

I realize there’s a Fixnum#power! method defined in rational.rb, but it
appears
that Fixnum has that method available to it whether or not I require
rational.rb. I looked through the Ruby source and didn’t see anything.

djberge@~-710>irb
irb(main):001:0> Fixnum.instance_methods(false).grep(/power/)
=> [“rpower”, “power!”]

So, where the heck is it coming from?

Thanks,

Dan


#2

On Mar 29, 2006, at 9:22 PM, Daniel B. wrote:

I realize there’s a Fixnum#power! method defined in rational.rb,
but it appears that Fixnum has that method available to it whether
or not I require rational.rb. I looked through the Ruby source and
didn’t see anything.

djberge@~-710>irb
irb(main):001:0> Fixnum.instance_methods(false).grep(/power/)
=> [“rpower”, “power!”]

So, where the heck is it coming from?

$ irb -rubygems

2.power!(3)
=> 8
exit
$ irb
2.power!(3)
NoMethodError: undefined method `power!’ for 2:Fixnum
from (irb):1
from :0

Are you requiring rubygems?

– Daniel


#3

Daniel B. wrote:

Hi all,

Ruby 1.8.4

I realize there’s a Fixnum#power! method defined in rational.rb, but it
appears that Fixnum has that method available to it whether or not I
require rational.rb. I looked through the Ruby source and didn’t see
anything.

numeric.c:
flo_pow(x, y)…
fix_pow(x, y)…

and then:
rb_define_method(rb_cFixnum, “", fix_pow, 1);
rb_define_method(rb_cFloat, "
”, flo_pow, 1);

and then in rational.rb:

class Fixnum
alias power! **
end

are You sure irb doesnt use rational.rb ?

lopex


#4

Daniel B. wrote:

Daniel H. wrote:

    from :0

Are you requiring rubygems?

– Daniel

Aha! I’m not explicitly require’ing it, but my $RUBYOPT was set. Once
I unset
it, I got the expected error.

FYI:

RubyGems uses time.rb
time.rb uses parsedate.rb
parsedate.rb uses date/format.rb
date/format.rb uses rational.rb


– Jim W.


#5

Daniel H. wrote:

    from :0

Are you requiring rubygems?

– Daniel

Aha! I’m not explicitly require’ing it, but my $RUBYOPT was set. Once
I unset
it, I got the expected error.

Thanks,

Dan