Bug in Range#end with exclusive ranges?

ruby-talk
1

Hi all,

Ruby 1.8.2 preview 2

If I have r = Range.new(0,50,true), shouldn't r.end return 49 and not
50? I am confused by this.

If this has come up on the ML in the past, please direct me (and
others) to the thread - I couldn't find it.

Regards,

Dan

PS - I disagree with this behavior.

2

In mathematical range notation, the difference between 0..50 and
0...50 is [0, 50] and [0, 50). Thus:

  irb(main):001:0> a = 0..50
  => 0..50
  irb(main):002:0> b = 0...50
  => 0...50
  irb(main):003:0> a.include?(49)
  => true
  irb(main):004:0> b.include?(49)
  => true
  irb(main):005:0> a.include?(49.9)
  => true
  irb(main):006:0> b.include?(49.9)
  => true
  irb(main):007:0> a.include?(50)
  => true
  irb(main):008:0> b.include?(50)
  => false

-austin

···

On Wed, 11 Aug 2004 04:26:15 +0900, Daniel Berger <djberg96@hotmail.com> wrote:

Ruby 1.8.2 preview 2

If I have r = Range.new(0,50,true), shouldn't r.end return 49 and
not 50? I am confused by this.

If this has come up on the ML in the past, please direct me (and
others) to the thread - I couldn't find it.

--
Austin Ziegler * halostatue@gmail.com
               * Alternate: austin@halostatue.ca

3

It doesn't seem quite right that Ruby should do this though

irb(main):001:0> a = (0..50)
a = (0..50)
=> 0..50
irb(main):002:0> b = (0.0..50.0)
b = (0.0..50.0)
=> 0.0..50.0
irb(main):004:0> a.last
a.last
=> 50
irb(main):005:0> a.last.class
a.last.class
=> Fixnum
irb(main):006:0> b.last.class
b.last.class
=> Float
irb(main):007:0> a.include?(49.9)
a.include?(49.9)
=> true
irb(main):008:0> b.include?(49.9)
b.include?(49.9)
=> true

Shouldn't Range#include work differently for integers?

martin

···

Austin Ziegler <halostatue@gmail.com> wrote:

In mathematical range notation, the difference between 0..50 and
0...50 is [0, 50] and [0, 50). Thus:

4

No. You want Range#member?

----------------------------------------------------- Range#member?
   rng.member?(val) => true or false

···

On Thu, 12 Aug 2004 16:01:11 +0900, Martin DeMello <martindemello@yahoo.com> wrote:

Austin Ziegler <halostatue@gmail.com> wrote:

In mathematical range notation, the difference between 0..50 and 0...50 is
[0, 50] and [0, 50). Thus:

It doesn't seem quite right that Ruby should do this though
irb(main):001:0> a = (0..50)
a = (0..50)
=> 0..50
irb(main):002:0> b = (0.0..50.0)
b = (0.0..50.0)
=> 0.0..50.0
irb(main):004:0> a.last
a.last
=> 50
irb(main):005:0> a.last.class
a.last.class
=> Fixnum
irb(main):006:0> b.last.class
b.last.class
=> Float
irb(main):007:0> a.include?(49.9)
a.include?(49.9)
=> true
irb(main):008:0> b.include?(49.9)
b.include?(49.9)
=> true

Shouldn't Range#include work differently for integers?

-------------------------------------------------------------------
   Return +true+ if _val_ is one of the values in _rng_ (that is if
   +Range#each+ would return _val_ at some point).

-austin
--
Austin Ziegler * halostatue@gmail.com
               * Alternate: austin@halostatue.ca

5

I consider it poor form that in Enumerable, include? and member? are
the same thing, but not so in Range (which includes Enumerable).
Aliases should be a matter of personal taste

Cheers,
Gavin

···

On Thursday, August 12, 2004, 11:19:59 PM, Austin wrote:

Shouldn't Range#include work differently for integers?

No. You want Range#member?