From: itsme213 <itsme213@...>
Date: 2005-01-27T03:55:47+09:00
Subject: Re: Array::uniq { block } ?

I agree, #uniq is certainly not the only candidate. I think it would be very
useful to compile a list of all such methods that could use a block to
override some default internal method (#== in the case of uniq), and
consider these for 2.0.

"Daniel Berger" <djberg96@hotmail.com> wrote in message
news:1106756045.223279.94600@f14g2000cwb.googlegroups.com...
> Belorion wrote:
> > I have an array of arrays.  I want to be able to do a uniq operation
> > on my array, and have it accept a block so I can specify what to
> > examine for uniqueness.  For example, in IRB
> >
> > a = [ [1,2], [3,4], [1,3] ]
> > a.sort { |x,y| x[0] <=> y[0] }
> >
> > returns => [ [1,2], [1,3], [3,4] ]
> >
> > However, it does not appear uniq accepts blocks:
> >
> > a = [ [1,2], [3,4], [1,3] ]
> > a.uniq { |x,y| x[0] == y[0] }
> >
> > returns => [ [1,2], [3,4], [1,3] ]
> > and a.uniq! with the above block returns nil.
> >
> > Am I stuck writing my own deep_uniq?  Granted, it would be hard, but
> > certainly not as clean as the above.
>
> Not that this is a bad idea, but why do I get this feeling that we're
> eventually going to find an excuse to allow blocks to virtually every
> method in the core library?
>
> Where does it all end?!
>
> Dan
>