Ranges and Enumerable problems

ruby-talk
1

Okay so when I play with "!".."~" wrong things happen.

("!".."~").include?("W")
true

("!.."~").include?(" ")
false

so far so good, however.

("!".."~").each do |c|
puts c
end

!
"

···

#
$
%
&
'
(
)
*
+
,
-
.
/
0
1
2
3
4
5
6
7
8
9
=> "!".."~"

This is wrong!

("!".."~").last
=> "~"

This is also right.

It looks to me like a problem with the Enumerable mixin but I guess the
question is why?

--
"Hey brother Christian with your high and mighty errand, Your actions speak
so loud, I can't hear a word you're saying."

-Greg Graffin (Bad Religion)

2

Note this

irb(main):003:0> "9".succ
=> "10"

Try this instead:

(?!..?~).each {|c| p c.chr}

irb(main):008:0> (?!..?~).include? ?\
=> false

Kind regards

robert

···

2008/7/2 Glen Holcomb <damnbigman@gmail.com>:

Okay so when I play with "!".."~" wrong things happen.

("!".."~").include?("W")
true

("!.."~").include?(" ")
false

so far so good, however.

("!".."~").each do |c|
puts c
end

!
"
#
$
%
&
'
(
)
*
+
,
-
.
/
0
1
2
3
4
5
6
7
8
9
=> "!".."~"

This is wrong!

("!".."~").last
=> "~"

This is also right.

It looks to me like a problem with the Enumerable mixin but I guess the
question is why?

--
use.inject do |as, often| as.you_can - without end

3

Hi,

As David wrote before, have a look at the Ascii-Table.

Okay so when I play with "!".."~" wrong things happen.

irb(main):002:0> "!"[0]
=> 33
irb(main):003:0> "~"[0]
=> 126

("!".."~").include?("W")
true

irb(main):001:0> "W"[0]
=> 87

("!.."~").include?(" ")
false

$ irb
irb(main):001:0> " "[0]
=> 32

32 is not between 33 and 126, so include?() does it right.

hth. regards, Sandor Szücs

···

On 02.07.2008, at 16:48, Glen Holcomb wrote:
--

4

Thanks Robert,

That is interesting. I wonder why it doesn't work as you would expect. The
range object seems to be correct (at least through limited include? testing)
however it returns incomplete info whenever you use an Enumerable method or
a method that appears to use an Enumerable method.

···

On Wed, Jul 2, 2008 at 9:12 AM, Robert Klemme <shortcutter@googlemail.com> wrote:

2008/7/2 Glen Holcomb <damnbigman@gmail.com>:
> Okay so when I play with "!".."~" wrong things happen.
>
> ("!".."~").include?("W")
> true
>
> ("!.."~").include?(" ")
> false
>
> so far so good, however.
>
> ("!".."~").each do |c|
> puts c
> end
>
> !
> "
> #
> $
> %
> &
> '
> (
> )
> *
> +
> ,
> -
> .
> /
> 0
> 1
> 2
> 3
> 4
> 5
> 6
> 7
> 8
> 9
> => "!".."~"
>
> This is wrong!
>
> ("!".."~").last
> => "~"
>
> This is also right.
>
> It looks to me like a problem with the Enumerable mixin but I guess the
> question is why?

Note this

irb(main):003:0> "9".succ
=> "10"

Try this instead:

(?!..?~).each {|c| p c.chr}

irb(main):008:0> (?!..?~).include? ?\
=> false

Kind regards

robert

--
use.inject do |as, often| as.you_can - without end

--
"Hey brother Christian with your high and mighty errand, Your actions speak
so loud, I can't hear a word you're saying."

-Greg Graffin (Bad Religion)

5

Hi --

Okay so when I play with "!".."~" wrong things happen.

("!".."~").include?("W")
true

("!.."~").include?(" ")
false

so far so good, however.

("!".."~").each do |c|
puts c
end

!
"
#
$
%
&
'
(
)
*
+
,
-
.
/
0
1
2
3
4
5
6
7
8
9
=> "!".."~"

This is wrong!

("!".."~").last
=> "~"

This is also right.

It looks to me like a problem with the Enumerable mixin but I guess the
question is why?

Note this

irb(main):003:0> "9".succ
=> "10"

Try this instead:

(?!..?~).each {|c| p c.chr}

irb(main):008:0> (?!..?~).include? ?\
=> false

Thanks Robert,

That is interesting. I wonder why it doesn't work as you would expect. The
range object seems to be correct (at least through limited include? testing)
however it returns incomplete info whenever you use an Enumerable method or
a method that appears to use an Enumerable method.

In 1.9 you get this:

('!'..'~').to_a

=> ["!", "\"", "#", "$", "%", "&", "'", "(", ")", "*", "+", ",", "-",
".", "/", "0", "1", "2", "3", "4", "5", "6", "7", "8", "9", ":", ";",
"<", "=", ">", "?", "@", "A", "B", "C", "D", "E", "F", "G", "H", "I",
"J", "K", "L", "M", "N", "O", "P", "Q", "R", "S", "T", "U", "V", "W",
"X", "Y", "Z", "[", "\\", "]", "^", "_", "`", "a", "b", "c", "d", "e",
"f", "g", "h", "i", "j", "k", "l", "m", "n", "o", "p", "q", "r", "s",
"t", "u", "v", "w", "x", "y", "z", "{", "|", "}", "~"]

David

···

On Thu, 3 Jul 2008, Glen Holcomb wrote:

On Wed, Jul 2, 2008 at 9:12 AM, Robert Klemme <shortcutter@googlemail.com> > wrote:

2008/7/2 Glen Holcomb <damnbigman@gmail.com>:

--
Rails training from David A. Black and Ruby Power and Light:
   Intro to Ruby on Rails July 21-24 Edison, NJ
   Advancing With Rails August 18-21 Edison, NJ
See http://www.rubypal.com for details and updates!

6

It's not Enumerable, it's the Range class's #each method.
In general, Ranges use the <=> operator to test against the first and
last element to determine inclusion, but they use #succ to generate
the next element for #each. Some odd behavior arises because
(s<=>s.succ) != -1 for all s.
I thought that was the whole issue, but it turns out '10' <=> '~' is -1.
It turns out the Range class has special code to deal with Strings -
it also checks that s.succ.length < end.length. Without that, your
example range would enumerate forever.

There has been discussion off and on about changing Range to use some
new alternate method to #succ which would guarantee 'sane' behavoir,
but nothing has ever come of it.

-Adam

···

On 7/2/08, Glen Holcomb <damnbigman@gmail.com> wrote:

That is interesting. I wonder why it doesn't work as you would expect. The
range object seems to be correct (at least through limited include? testing)
however it returns incomplete info whenever you use an Enumerable method or
a method that appears to use an Enumerable method.

7

Thanks David,

That is good to know. Maybe I should upgrade.

···

On Wed, Jul 2, 2008 at 10:34 AM, David A. Black <dblack@rubypal.com> wrote:

Hi --

On Thu, 3 Jul 2008, Glen Holcomb wrote:

On Wed, Jul 2, 2008 at 9:12 AM, Robert Klemme <shortcutter@googlemail.com >> > >> wrote:

2008/7/2 Glen Holcomb <damnbigman@gmail.com>:

Okay so when I play with "!".."~" wrong things happen.

("!".."~").include?("W")
true

("!.."~").include?(" ")
false

so far so good, however.

("!".."~").each do |c|
puts c
end

!
"
#
$
%
&
'
(
)
*
+
,
-
.
/
0
1
2
3
4
5
6
7
8
9
=> "!".."~"

This is wrong!

("!".."~").last
=> "~"

This is also right.

It looks to me like a problem with the Enumerable mixin but I guess the
question is why?

Note this

irb(main):003:0> "9".succ
=> "10"

Try this instead:

(?!..?~).each {|c| p c.chr}

irb(main):008:0> (?!..?~).include? ?\
=> false

Thanks Robert,

That is interesting. I wonder why it doesn't work as you would expect.
The
range object seems to be correct (at least through limited include?
testing)
however it returns incomplete info whenever you use an Enumerable method
or
a method that appears to use an Enumerable method.

In 1.9 you get this:

('!'..'~').to_a

=> ["!", "\"", "#", "$", "%", "&", "'", "(", ")", "*", "+", ",", "-",

".", "/", "0", "1", "2", "3", "4", "5", "6", "7", "8", "9", ":", ";",
"<", "=", ">", "?", "@", "A", "B", "C", "D", "E", "F", "G", "H", "I",
"J", "K", "L", "M", "N", "O", "P", "Q", "R", "S", "T", "U", "V", "W",
"X", "Y", "Z", "[", "\\", "]", "^", "_", "`", "a", "b", "c", "d", "e",
"f", "g", "h", "i", "j", "k", "l", "m", "n", "o", "p", "q", "r", "s",
"t", "u", "v", "w", "x", "y", "z", "{", "|", "}", "~"]

David

--
Rails training from David A. Black and Ruby Power and Light:
Intro to Ruby on Rails July 21-24 Edison, NJ
Advancing With Rails August 18-21 Edison, NJ
See http://www.rubypal.com for details and updates!

--
"Hey brother Christian with your high and mighty errand, Your actions speak
so loud, I can't hear a word you're saying."

-Greg Graffin (Bad Religion)

8

Looks like I spoke too soon. Did Range#each change, or did String#succ?

···

On 7/2/08, David A. Black <dblack@rubypal.com> wrote:

In 1.9 you get this:

> > ('!'..'~').to_a
> >
=> ["!", "\"", "#", "$", "%", "&", "'", "(", ")", "*", "+", ",", "-",
".", "/", "0", "1", "2", "3", "4", "5", "6", "7", "8", "9", ":", ";",
"<", "=", ">", "?", "@", "A", "B", "C", "D", "E", "F", "G", "H", "I",
"J", "K", "L", "M", "N", "O", "P", "Q", "R", "S", "T", "U", "V", "W",
"X", "Y", "Z", "[", "\\", "]", "^", "_", "`", "a", "b", "c", "d", "e",
"f", "g", "h", "i", "j", "k", "l", "m", "n", "o", "p", "q", "r", "s",
"t", "u", "v", "w", "x", "y", "z", "{", "|", "}", "~"]

On 7/2/08, Adam Shelly <adam.shelly@gmail.com> wrote:

There has been discussion off and on about changing Range to use some
new alternate method to #succ which would guarantee 'sane' behavoir,
but nothing has ever come of it.

9

<snip>

Some odd behavior arises because
(s<=>s.succ) != -1 for all s.

Is that so???

I guess you mean there exists an s for which ...

Cheers
Robert

···

On Wed, Jul 2, 2008 at 7:13 PM, Adam Shelly <adam.shelly@gmail.com> wrote:
--
http://ruby-smalltalk.blogspot.com/

---
AALST (n.) One who changes his name to be further to the front
D.Adams; The Meaning of LIFF

10

I'm not sure what you mean

irb(main):027:0> 1<=>1.succ
=> -1
irb(main):028:0> 'a'<=>'a'.succ
=> -1
irb(main):029:0> 'z'<=>'z'.succ
=> 1

···

On 7/2/08, Robert Dober <robert.dober@gmail.com> wrote:

On Wed, Jul 2, 2008 at 7:13 PM, Adam Shelly <adam.shelly@gmail.com> wrote:
<snip>
> Some odd behavior arises because
> (s<=>s.succ) != -1 for all s.

Is that so???

I guess you mean there exists an s for which ...

11

He meant

You said: for all s: (s<=>s.succ) != -1

He said: there exists at least one s: (s<=>s.succ) != -1

It's probably Robert's mathematical training which made him detect that the two statements are not equivalent - and he is absolutely right. :slight_smile:

Kind regards

  robert

···

On 02.07.2008 19:48, Adam Shelly wrote:

On 7/2/08, Robert Dober <robert.dober@gmail.com> wrote:

On Wed, Jul 2, 2008 at 7:13 PM, Adam Shelly <adam.shelly@gmail.com> wrote:
<snip>

Some odd behavior arises because
(s<=>s.succ) != -1 for all s.

Is that so???

I guess you mean there exists an s for which ...

I'm not sure what you mean

12

You see Adam it's not my fault I just had bad education;), sorry I
thought it was just a misstype of yours.

As there exist some s such as ( s <=> s.succ) == -1 (most of them by
the way) your statement
was wrong, I guess that you got carried away by the not, could easily
have happened to me.

! for all s: (s<=> s.succ) == -1

is true. Do we read each other now?

Cheers
Robert

···

On Wed, Jul 2, 2008 at 8:27 PM, Robert Klemme <shortcutter@googlemail.com> wrote:

On 02.07.2008 19:48, Adam Shelly wrote:

On 7/2/08, Robert Dober <robert.dober@gmail.com> wrote:

On Wed, Jul 2, 2008 at 7:13 PM, Adam Shelly <adam.shelly@gmail.com> >>> wrote:
<snip>

Some odd behavior arises because
(s<=>s.succ) != -1 for all s.

Is that so???

I guess you mean there exists an s for which ...

I'm not sure what you mean

He meant

You said: for all s: (s<=>s.succ) != -1

He said: there exists at least one s: (s<=>s.succ) != -1

It's probably Robert's mathematical training which made him detect that the
two statements are not equivalent - and he is absolutely right. :slight_smile:

13

Or another way of saying the same thing (as Robert D.):

   !(for all s: (s <=>s.succ) == -1)

···

On Wed, Jul 2, 2008 at 2:27 PM, Robert Klemme <shortcutter@googlemail.com> wrote:

On 02.07.2008 19:48, Adam Shelly wrote:

On 7/2/08, Robert Dober <robert.dober@gmail.com> wrote:

On Wed, Jul 2, 2008 at 7:13 PM, Adam Shelly <adam.shelly@gmail.com> >>> wrote:
<snip>

Some odd behavior arises because
(s<=>s.succ) != -1 for all s.

Is that so???

I guess you mean there exists an s for which ...

I'm not sure what you mean

He meant

You said: for all s: (s<=>s.succ) != -1

He said: there exists at least one s: (s<=>s.succ) != -1

--
Rick DeNatale

My blog on Ruby
http://talklikeaduck.denhaven2.com/

14

Well, not really. That gives you no information as to whether such a
thing does exist. That statement would be true, but not definitive,
meaning, that it would be true if there is _no_ s that satisfies.

I think Robert was going for "at least 1" in his statement.

But, I guess that's just making a mountain out of a mole hill, because
I think Rick was just pointing out something, and not trying to be
being rigorously complete.

Todd

···

On Wed, Jul 2, 2008 at 3:50 PM, Rick DeNatale <rick.denatale@gmail.com> wrote:

  !(for all s: (s <=>s.succ) == -1)

15

Okay, that may have been a major faux pas on my part :slight_smile:

Todd

···

On Wed, Jul 2, 2008 at 5:07 PM, Todd Benson <caduceass@gmail.com> wrote:

On Wed, Jul 2, 2008 at 3:50 PM, Rick DeNatale <rick.denatale@gmail.com> wrote:

  !(for all s: (s <=>s.succ) == -1)

Well, not really. That gives you no information as to whether such a
thing does exist. That statement would be true, but not definitive,
meaning, that it would be true if there is _no_ s that satisfies.

I think Robert was going for "at least 1" in his statement.

But, I guess that's just making a mountain out of a mole hill, because
I think Rick was just pointing out something, and not trying to be
being rigorously complete.

16

It is. :slight_smile: not (for all e: X(e)) <=> (there is at least one e: not X(e))

Kind regards

robert

···

2008/7/3 Todd Benson <caduceass@gmail.com>:

On Wed, Jul 2, 2008 at 5:07 PM, Todd Benson <caduceass@gmail.com> wrote:

On Wed, Jul 2, 2008 at 3:50 PM, Rick DeNatale <rick.denatale@gmail.com> wrote:

  !(for all s: (s <=>s.succ) == -1)

Well, not really. That gives you no information as to whether such a
thing does exist. That statement would be true, but not definitive,
meaning, that it would be true if there is _no_ s that satisfies.

Okay, that may have been a major faux pas on my part :slight_smile:

--
use.inject do |as, often| as.you_can - without end

17

as being related to De Morgan's laws in Boolean algebra, and Wikipedia seems
to agree De Morgan's laws - Wikipedia

···

On Thu, Jul 3, 2008 at 5:23 AM, Robert Klemme <shortcutter@googlemail.com> wrote:

2008/7/3 Todd Benson <caduceass@gmail.com>:
> On Wed, Jul 2, 2008 at 5:07 PM, Todd Benson <caduceass@gmail.com> wrote:
>> On Wed, Jul 2, 2008 at 3:50 PM, Rick DeNatale <rick.denatale@gmail.com> > wrote:
>>> !(for all s: (s <=>s.succ) == -1)
>>
>> Well, not really. That gives you no information as to whether such a
>> thing does exist. That statement would be true, but not definitive,
>> meaning, that it would be true if there is _no_ s that satisfies.
>
> Okay, that may have been a major faux pas on my part :slight_smile:

It is. :slight_smile: not (for all e: X(e)) <=> (there is at least one e: not X(e))

Yep, I remember getting A's in my logic courses. I vaguely remembered this

--
Rick DeNatale

My blog on Ruby
http://talklikeaduck.denhaven2.com/

18

Wow, I seem to have started quite a tangent. That's exactly what I
was thinking, but when I translated my thoughts into pseudo-ruby, I
put the ! in the wrong place for boolean logic, which I wasn't even
considering.

I never did get an answer to my question about v1.9 - it's probably
time for me to install it - but until I do:
did String#succ change? Does '9'.succ return ':' now?
Is the above statement - which was true in 1.85 - no longer true in 1.9?

-Adam

···

On 7/3/08, Rick DeNatale <rick.denatale@gmail.com> wrote:

On Thu, Jul 3, 2008 at 5:23 AM, Robert Klemme <shortcutter@googlemail.com> > wrote:

> 2008/7/3 Todd Benson <caduceass@gmail.com>:
> > On Wed, Jul 2, 2008 at 5:07 PM, Todd Benson <caduceass@gmail.com> wrote:
> >> On Wed, Jul 2, 2008 at 3:50 PM, Rick DeNatale <rick.denatale@gmail.com> > > wrote:
> >>> !(for all s: (s <=>s.succ) == -1)
> >>

19

Okay, you made your point, but you don't have to rub it in... (crosses
arms and says "hmppph" [picture a little girl here] :slight_smile:

I don't know why, but when I first looked at it, I had a gut feeling
there was something wrong with Rick's statement. Then, I realized my
mistake. Darn it!

Todd

···

On Thu, Jul 3, 2008 at 6:48 AM, Rick DeNatale <rick.denatale@gmail.com> wrote:

On Thu, Jul 3, 2008 at 5:23 AM, Robert Klemme <shortcutter@googlemail.com> > wrote:

> Okay, that may have been a major faux pas on my part :slight_smile:

It is. :slight_smile: not (for all e: X(e)) <=> (there is at least one e: not X(e))

Yep, I remember getting A's in my logic courses. I vaguely remembered this

as being related to De Morgan's laws in Boolean algebra, and Wikipedia seems
to agree De Morgan's laws - Wikipedia

--
Rick DeNatale

20

Adam my apologies, instead of logic nitpicking, which was fun of
course, I could have provided an answer to the original question too
;).
But you will not like it :frowning:

robert@roma:~ 20:27:45
437/7 > ruby1.9 -v -e 'p "9".succ'
ruby 1.9.0 (2007-12-25 revision 14709) [i486-linux]
"10"

Cheers
Robert

···

--
http://ruby-smalltalk.blogspot.com/

---
AALST (n.) One who changes his name to be further to the front
D.Adams; The Meaning of LIFF