From: Rick DeNatale <rick.denatale@...>
Date: 2009-02-20T06:22:55+09:00
Subject: Re: Creating InsensitiveHash in C

--0016e645aaccc38ed004634c24a8
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

On Thu, Feb 19, 2009 at 3:57 PM, I=F1aki Baz Castillo <ibc@aliax.net> wrote=
:

> Hi, I'd like a faster Ruby class same as Hash but in which key and value
> are
> insensitive, so:
>
> ----------------------------------
>  params =3D InsensitiveHash.new
>  params["q"] =3D 0.5
>  params["transport"] =3D "TCP"
>
>  params["Q"]
>  =3D> 0.5
>
>  params["Transport"] =3D=3D "tcp"
>  =3D> true
> ----------------------------------
>
> It will be used extensively so I need it being very fast. I'm thinking in
> implementing it as a new Ruby class in C, modifying Hash class.
>
> Do you think it's feasible? Any suggestion?
> Thanks a lot.
>
> PD: Where is the C file defining Hash class? I read "hash.c" in Hash
> documentation:
>  http://www.ruby-doc.org/core/classes/Hash.html
> but I can't find it in Ruby libs directory.
>

You won't find it in a runtime installation of Ruby. The file hash.h is in
the top level directory of the Ruby source code.

That said, If all you want to do is to make the keys case insensitive, a
naive Ruby implementation is likely to be almost as fast as anything you
could do in C, since the time intensitve implementation parts of Hash are
already in C.

So something like

class InsensitiveHash
    def initialize
       @hash =3D {}
    end

   def [](key}
      @hash[key.downcase]
   end

   def []=3D(key, value)
      @hash[key.downcase] =3D value
   end
end

And yes, I'd recommend having the InsensitiveHash have an instance of Hash
rather than being a subclass of Hash.

--=20
Rick DeNatale

Blog: http://talklikeaduck.denhaven2.com/
Twitter: http://twitter.com/RickDeNatale
WWR: http://www.workingwithrails.com/person/9021-rick-denatale
LinkedIn: http://www.linkedin.com/in/rickdenatale

--0016e645aaccc38ed004634c24a8--