From: Akinori MUSHA <knu@...>
Date: 2006-11-02T14:35:26+09:00
Subject: Re: Question about the Set class, XOR, and arrays

Hi,

At Thu, 2 Nov 2006 05:32:11 +0900,
Gavin Kistner wrote:
> From: Daniel Berger [mailto:djberg96@gmail.com] 
> > irb(main):003:0> Set[1,2,3] ^ [3,4,5,5]
> > => #<Set: {1, 2, 4}>
> > 
> > What?! I'm confused. Do I need a refresher in Set theory or something?
> 
> Looks like a bug due to Set#^'s implementation that assumes that the
> enumerable passed in has each item only once (or, due to the
> implementation, an odd number of times). 
> 
> irb(main):001:0> require 'set'
> => true
> irb(main):002:0> a = Set[1,2,3]
> => #<Set: {1, 2, 3}>
> irb(main):003:0> a ^ [3,4,5]
> => #<Set: {5, 1, 2, 4}>
> irb(main):004:0> a ^ [3,4,5,5]
> => #<Set: {1, 2, 4}>
> irb(main):005:0> a ^ [3,4,5,5,5]
> => #<Set: {5, 1, 2, 4}>
> 
> The first time it sees "5", it adds it to the resulting set. The second
> time it sees 5, it says "Oh, I already have a 5 in the result, better
> remove it."

I'll commit the following fix which I think is smaller and faster.
Thanks!

Index: set.rb
===================================================================
RCS file: /src/ruby/lib/set.rb,v
retrieving revision 1.30
diff -u -u -r1.30 set.rb
--- set.rb	2 Nov 2006 00:21:26 -0000	1.30
+++ set.rb	2 Nov 2006 05:32:23 -0000
@@ -294,8 +294,8 @@
   # and the given enumerable object.  (set ^ enum) is equivalent to
   # ((set | enum) - (set & enum)).
   def ^(enum)
-    n = dup
-    enum.each { |o| if n.include?(o) then n.delete(o) else n.add(o) end }
+    n = Set.new(enum)
+    each { |o| if n.include?(o) then n.delete(o) else n.add(o) end }
     n
   end
 

-- 
                     /
                    /__  __            Akinori.org / MUSHA.org
                   / )  )  ) )  /     FreeBSD.org / Ruby-lang.org
Akinori MUSHA aka / (_ /  ( (__(  @ iDaemons.org / and.or.jp

"Different eyes see different things,
    Different hearts beat on different strings --
       But there are times for you and me when all such things agree"