[].all?{} and [].any?{} Behavior

[captaffy]

I find the following behavior interesting (so interesting that I
modified it), and I would like to hear others’ thoughts on the subject:

[3].all? {|element| element == 3 } # => true
[3].all? {|element| element != 3 } #Â => false (sanity checks)

[].all? {|element| element == 3 } # => true
[].all? {|element| element != 3 } # => true

[].any? {|element| element == 3 } # => false
[].any? {|element| element != 3 } # => false

Ruby 1.8.6 and 1.9.1 both give these results.

The first interesting thing is that both the == and the != checks give
the same logical result (always true for all?, always false for any?).
After thinking about it a little, I decided that this is the desired
behavior.

I also understand why it happens. For example, in the case of all?, the
documentation says that true will be the result if the block never
returns false or nil. In the case of [], the block never gets called, so
the result is true. I know the block never gets called because the
following does not print anything:

[].all? {|dummy| puts ‘print something’}, while

[3].all? {|dummy| puts ‘print something’} does.

The second interesting thing is that a result of this behavior is that
for the same check, all? will give a result of true, while any? will
give a result of false. This seems contradictory.

I would prefer that for [], both all? and any? would give a result of
false for any check. So I have over-ridden Array#all?, returning false
if self == []. My main motivation for doing so is in situautions such
as:

obj_array.find_all{|obj| obj.attr_1 == x}.all?{|obj| obj.attr_2 == y}

If the find_all returns [], I want the all? result to be false, not
true.

I assume that others have run across this but some quick searches did
not turn up anything. I am wondering how others deal with this such as
over-riding as I do, checking for [] each time (which does not seem very
Ruby-like), or even leaving the operation as is because for some, it may
be the desired behavior.

Finally, are there any potential detrimental effects that might occur
due to the behavior modification that I made. I am not a Rails user (if
that matters), I mainly use Ruby for scripting and hardware control
applications (and I am interested in learning as much as I can about
Ruby because I like it so much).

js

[captaffy]

The docs pretty clear…
If the block is not given, Ruby adds an implicit block of{|obj|
obj} (that
is all? <Enumerable.html#M003131> will return true only if none of the
collection members are false or nil.)

Its an assumption is always true unless the block tells it something
different…

Assuimng Happy unless its told not to be happy!

irb(main):019:0> [].all? {|e| puts e; e==3}
=> true
irb(main):020:0> [].all? {|e| puts e; e==2}
=> true
irb(main):021:0> [].all? {|e| puts e; e==1}
=> true
irb(main):022:0> [].all? {|e| puts e; e==100}
=> true
irb(main):023:0>

[].all? Is true…

Its interesting.

irb(main):019:0> [].all? {|e| e==3}
=> true
irb(main):020:0> [].all? {|e| e==2}
=> true
irb(main):021:0> [].all? {|e| e==1}
=> true
irb(main):022:0> [].all? {|e| e==100}
=> true
irb(main):023:0>

[].all? Is true…

Its interesting.

[captaffy]

Hi –

On Fri, 30 Jul 2010, John S. wrote:

[].any? {|element| element != 3 } # => false
returns false or nil. In the case of [], the block never gets called, so
the result is true. I know the block never gets called because the
following does not print anything:

[].all? {|dummy| puts ‘print something’}, while

[3].all? {|dummy| puts ‘print something’} does.

The second interesting thing is that a result of this behavior is that
for the same check, all? will give a result of true, while any? will
give a result of false. This seems contradictory.

As I see it, all? and any? make sense in terms of each other. If
array.all? is true for a condition, that means that array.any? is false
for the opposite of the condition:

array.all? {|e| cond(e) } == !(array.any? {|e| !cond(e) })

I think that always holds. If [].all? were false, it would not; it would
flip from true to false depending on how many elements were in the
array.

To put it in more human terms, array.all? for a condition means: there
is no element in this array which violates the condition. I know that it
feels more natural to say “Every element in this array meets this
condition” – but the problem with that is precisely that there’s no
guarantee that an array has elements, so it’s more accurate to phrase it
in the negative way.

I would prefer that for [], both all? and any? would give a result of
false for any check. So I have over-ridden Array#all?, returning false
if self == []. My main motivation for doing so is in situautions such
as:

obj_array.find_all{|obj| obj.attr_1 == x}.all?{|obj| obj.attr_2 == y}

If the find_all returns [], I want the all? result to be false, not
true.

It feels to me like you’re expecting all? to do too many things. I would
either just use all? as it stands, or do something else entirely, like:

obj_array.find {|obj| obj.attr_1 == x &&! obj.attr_2 == y }

which I’ve probably garbled but you get the idea

Ruby because I like it so much).
Overriding core methods always has the potential to cause problems,
because other code (in the interpreter and/or the standard library
and/or gems and other third-party libraries) may be depending on the
documented behavior.

David

–
David A. Black, Senior Developer, Cyrus Innovation Inc.

The Ruby training with Black/Brown/McAnally
Compleat Philadelphia, PA, October 1-2, 2010
Rubyist http://www.compleatrubyist.com

[captaffy]

Afternoon,

On Thu, Jul 29, 2010 at 2:27 PM, John S. [email protected]
wrote:

I assume that others have run across this but some quick searches did
not turn up anything. I am wondering how others deal with this such as
over-riding as I do, checking for [] each time (which does not seem very
Ruby-like), or even leaving the operation as is because for some, it may
be the desired behavior.

I might suggest your better course would be to create a new method
rather
than override the existing method. Maybe something like js_any? or
js_all?
would be better in that it shows both what it’s doing and that it’s not
the
standard method at the same time. Then you have no issues with any other
code that depends on the standard methodology in terms of the methods. I
appreciate the aesthetics are a little lacking but sometimes that a
small
price to pay.

John

[captaffy]

John W Higgins wrote:

I might suggest your better course would be to create a new method
rather
than override the existing method. Maybe something like js_any? or
js_all?
would be better in that it shows both what it’s doing and that it’s not
the
standard method at the same time. Then you have no issues with any other
code that depends on the standard methodology in terms of the methods. I
appreciate the aesthetics are a little lacking but sometimes that a
small
price to pay.

After reading your post and the previous post from David, I agree.

Thanks,
js

[captaffy]

Hello!

I’d have to agree with John on this one. It is really bizarre that
[].any? returns false and [].all? returns true. It is just not what to
expect without reading documentation and thinking about every word
written in there and trying things out in irb.

I love Ruby because most of the time everything just works as you’d
expect without reading any documentation at all and this is not the
case. It just doesn’t make sense that asking Ruby to answer the
question “does this empty collection have all elements as 3?” to
return true. Read again the question and you’ll understand what i
mean. Again, if i think about Ruby being a very much like a version of
an English language, then this doesn’t make sense also.

And since i’d expect it to do what i have been expected then i haven’t
thought about this situation ever also. I’d consider this to be
changed in future versions of Ruby, but what am i to decide on such
things

It would make sense if [].all? and [].any? would return false. Always.

Jarmo P.

IT does really matter - http://www.itreallymatters.net

[captaffy]

David A. Black wrote:

As I see it, all? and any? make sense in terms of each other. If
array.all? is true for a condition, that means that array.any? is false
for the opposite of the condition:

array.all? {|e| cond(e) } == !(array.any? {|e| !cond(e) })

I think that always holds. If [].all? were false, it would not; it would
flip from true to false depending on how many elements were in the
array.

To put it in more human terms, array.all? for a condition means: there
is no element in this array which violates the condition. I know that it
feels more natural to say “Every element in this array meets this
condition” – but the problem with that is precisely that there’s no
guarantee that an array has elements, so it’s more accurate to phrase it
in the negative way.

I must admit that I would never have thought of the behavior in this
manner. But the way that you explain it, it makes sense. When I first
started using Ruby, which was my first (and so far only) OO language, I
found it surprisingly difficult to make the switch from procedural to OO
programming. I really had to change my way of thinking. And this was
just for OO programming, not to mention the Ruby way of thinking as
illustated here. So I ask questions and thankfully,
usually someone can set me straight.

It feels to me like you’re expecting all? to do too many things. I would
either just use all? as it stands, or do something else entirely, like:

obj_array.find {|obj| obj.attr_1 == x &&! obj.attr_2 == y }

which I’ve probably garbled but you get the idea

Yes, I know what you mean and I actually have thought of doing something
similar to your code. But a couple of the things that I really like
about Ruby are 1) the Enumerable module itself, and 2) the ability to
string together methods. Well, that and the fact that Ruby is dynamic.
So I probably am asking all? to do too much in this case, but I love to
string methods containing code blocks together as in the example I gave.
So much power in so little space.

Overriding core methods always has the potential to cause problems,
because other code (in the interpreter and/or the standard library
and/or gems and other third-party libraries) may be depending on the
documented behavior.

Yes, I was afraid someone was going to say this (in fact the next person
to reply said the same thing). I think I knew deep down that this was
the case, and that I should leave all? as is.

Thanks,
js

[captaffy]

On Thu, Jul 29, 2010 at 5:27 PM, John S. [email protected]
wrote:

The second interesting thing is that a result of this behavior is that
for the same check, all? will give a result of true, while any? will
give a result of false. This seems contradictory.

Well, there’s a theoretical basis for this. Enumeration#all? is an
implementation of the universal quantifier from predicate logic (that
upside down A) symbol.

By convention the universal quantifier evaluates to true for an empty
set:

Universal quantification | The empty set

Vacuous truth

Ruby because I like it so much).
For your own usage as long as it doesn’t mess up some other code you
are using, feel free.

For library code, such as in a gem I think it would be better to think
up some other method name rather than changing the standard, e.g.

module Enumerable
def non_vacuous_all?(&b)
!empty? && all?(&b)
end
end

[3].all? {|element| element == 3 } # => true
[3].all? {|element| element != 3 } # => false

[].all? {|element| element == 3 } # => true
[].all? {|element| element != 3 } # => true

[3].non_vacuous_all? {|element| element == 3 } # => true
[3].non_vacuous_all? {|element| element != 3 } # => false

[].non_vacuous_all? {|element| element == 3 } # => false
[].non_vacuous_all? {|element| element != 3 } # => false

[].any? {|element| element == 3 } # => false
[].any? {|element| element != 3 } # => false

There may be a better name than non_vacuous_all? but I can’t think of
one.

–
Rick DeNatale

Blog: http://talklikeaduck.denhaven2.com/
Github: rubyredrick (Rick DeNatale) · GitHub
Twitter: @RickDeNatale
WWR: http://www.workingwithrails.com/person/9021-rick-denatale
LinkedIn: http://www.linkedin.com/in/rickdenatale

[captaffy]

2010/7/30, Jarmo P. [email protected]:

It just doesn’t make sense that asking Ruby to answer the
question “does this empty collection have all elements as 3?” to
return true. Read again the question and you’ll understand what i
mean. Again, if i think about Ruby being a very much like a version of
an English language, then this doesn’t make sense also.

There’s no easy answer to this question, as there is nothing easy
dealing with an empty list. Can you prove that not all elements in []
equal 3 ? You can’t, because no element has another value. Thus, this
is true.

At first glance, I would intuitively agree with you, but after a
while, I can’t find a mean to prove what may sound obvious to you (but
is not).

[captaffy]

Xavier Noëlle wrote:

is true.

At first glance, I would intuitively agree with you, but after a
while, I can’t find a mean to prove what may sound obvious to you (but
is not).

Here’s one way to think of it, in terms of an example:

pattern = /…/
tests = [ Test.new(…), …, Test.new(…) ]
runnable_tests = tests.select {|test| pattern === test.name}

if runnable_tests.all?{|test|test.pass}
puts “Ok!”
else
puts “Failed!”
end
END

How do we want this program to work if no tests match the pattern? Is it
a failure case? IMO, it is not.

I don’t want to code a special case for runnable_tests.empty?

This seems very English-like to me, but YMMV of course.

[captaffy]

Rick Denatale wrote:

By convention the universal quantifier evaluates to true for an empty
set:
Universal quantification - Wikipedia
Vacuous truth - Wikipedia

Well, that seals it for me. My thinking is wrong if it goes against
convention, not only in Ruby, but in universal quantification.

For library code, such as in a gem I think it would be better to think
up some other method name rather than changing the standard, e.g.

module Enumerable
def non_vacuous_all?(&b)
!empty? && all?(&b)
end
end

[3].all? {|element| element == 3 } # => true
[3].all? {|element| element != 3 } # => false

[].all? {|element| element == 3 } # => true
[].all? {|element| element != 3 } # => true

[3].non_vacuous_all? {|element| element == 3 } # => true
[3].non_vacuous_all? {|element| element != 3 } # => false

[].non_vacuous_all? {|element| element == 3 } # => false
[].non_vacuous_all? {|element| element != 3 } # => false

[].any? {|element| element == 3 } # => false
[].any? {|element| element != 3 } # => false

There may be a better name than non_vacuous_all? but I can’t think of
one.

Agreed, I should use a similar technigue for my specific applications.

js

[captaffy]

On Fri, Jul 30, 2010 at 10:49 AM, Rick DeNatale
[email protected]wrote:

[].all? {|element| element != 3 } # => true
There may be a better name than non_vacuous_all? but I can’t think of one.

How about #appall? to imply that it is a pessimistic implementation of
#all?

[captaffy]

Joel VanderWerf wrote:

pattern = /…/
tests = [ Test.new(…), …, Test.new(…) ]
runnable_tests = tests.select {|test| pattern === test.name}

if runnable_tests.all?{|test|test.pass}
puts “Ok!”
else
puts “Failed!”
end
END

How do we want this program to work if no tests match the pattern? Is it
a failure case? IMO, it is not.

I don’t want to code a special case for runnable_tests.empty?

This seems very English-like to me, but YMMV of course.

Yes, I can see where you would want your result to be according to
convention and opposite of what I stated.

js

[captaffy]

2010/7/30 Rick DeNatale [email protected]:

Universal quantification - Wikipedia
Vacuous truth - Wikipedia

To complement this, De Morgan’s Laws help do the conversion between
all? and and? variants properly:

De Morgan's laws

I assume that others have run across this but some quick searches did

For your own usage as long as it doesn’t mess up some other code you
are using, feel free.

I disagree: IMHO it is a bad idea to change such fundamental behavior
if only for own code. This opens the door widely for all sorts of
bugs and issues. For example, you get used to #all? doing also the
emptyness check and get confused when reading other code which of
course relies on the regular behavior. Or you forget the “require”
for the file that changes semantics of #all? and #any? and receive in
turn subtly bugs which might be hard to track down. Even worse, you
use library code that in turn uses #all? or #any? without you knowing
it and this code suddenly breaks.

For library code, such as in a gem I think it would be better to think
up some other method name rather than changing the standard, e.g.

That’s definitively the way to go if the behavior should be put into a
method.

[].all? {|element| element != 3 } # => true
There may be a better name than non_vacuous_all? but I can’t think of one.
I’d rather stick with two method calls because it makes crystall clear
what’s happening. Also, you may first want to check for emptyness and
if else branch based on that knowledge (or the other way round). In
other words: often you may want to separate both checks.

Kind regards

robert

[captaffy]

2010/7/30 Jarmo P. [email protected]:

I’d have to agree with John on this one. It is really bizarre that
[].any? returns false and [].all? returns true. It is just not what to
expect without reading documentation and thinking about every word
written in there and trying things out in irb.

This behavior is exactly what I expect from these methods.
Otherwise it would break all sorts of boolean logic. These methods
are defined the way they are with good reason - even if this does not
meet your expectation here. It’s the most reasonable definition in
light of boolean logic and De Morgan’s laws which I believe
everybody writing software must know by heart and understand because
it is at the very foundation of our profession.

I love Ruby because most of the time everything just works as you’d
expect without reading any documentation at all and this is not the

Frankly, this is a dangerous thing to do. Generally doing away with
documentation will bring you in trouble sooner or later.

Kind regards

robert

[captaffy]

On Sat, Jul 31, 2010 at 6:07 AM, Robert K.
[email protected] wrote:

turn subtly bugs which might be hard to track down. Even worse, you
use library code that in turn uses #all? or #any? without you knowing
it and this code suddenly breaks.

Whether or not it’s a bad idea, and I tend to agree that it is. I
said what I did for at least two reasons:

1. I tend not to be puritanical, just as I wouldn’t restrict what
   anyone wanted to do in the privacy of their own homes unless it was
   harmful to others, I think you should be able to write whatever code
   you want to under the same philosophy, whether or not it’s harmful to
   you.

2. Since one of the most useful ways to learn anything in such a way
   that you remember and internalize it is to make a mistake and realize
   the consequences.

[].all? {|element| element == 3 } # => true

There may be a better name than non_vacuous_all? but I can’t think of one.

I’d rather stick with two method calls because it makes crystall clear
what’s happening. Also, you may first want to check for emptyness and
if else branch based on that knowledge (or the other way round). In
other words: often you may want to separate both checks

Here I completely disagree. Extracting commonly used code to a well
named method is an essential part of writing and maintaining code.

for example, I see nothing wrong with the sum method which active
support adds to Enumerable

Calculates a sum from the elements. Examples:

payments.sum { |p| p.price * p.tax_rate }

payments.sum(&:price)

The latter is a shortcut for:

payments.inject { |sum, p| sum + p.price }

It can also calculate the sum without the use of a block.

[5, 15, 10].sum # => 30

[“foo”, “bar”].sum # => “foobar”

[[1, 2], [3, 1, 5]].sum => [1, 2, 3, 1, 5]

The default sum of an empty list is zero. You can override this

default:

[].sum(Payment.new(0)) { |i| i.amount } # => Payment.new(0)

def sum(identity = 0, &block)
if block_given?
map(&block).sum(identity)
else
inject { |sum, element| sum + element } || identity
end
end

Following your argument, this is bad because you might want to use
inject and + separately.

But having such methods doesn’t prevent you in the least from using
inject, +, empty?, any? or any other method used to implement a
slightly more abstract extracted method separately. It does help to
keep your code DRY and to make it more understandable overall since
you don’t have to re-understand the effect of the separate invocations
each time you encounter them, as long as you are careful and name the
abstraction in an ‘intention revealing’ way.

And doing this also enables changing the implementation of the
abstraction without holistic changes to the code. Yes I know about de
Morgan’s rules (I have a CS degree granted by a 1970’s era Electrical
Engineering department). Placing the implementation in an abstraction
allows you to do the math proofs/unit testing and refactoring to meet
particular non-functional requirements in one place, which is a good
thing.

I recently was working on a refactoring a large Rails application
taken over from another development shop, which had several nasty bugs
on just this issue of all? returning true for an empty collection. It
turns out that there are definitely cases where you want to test that
a collection has at least 1 element and that all of the elements have
some property. Having said that perhaps a better name for the method
might be

at_least_one_and_all?

That might be a tad long, but I’d rather have a longer but more
intention revealing name, and let one of the several editors I use
deal with keeping my keystroke count down.

–
Rick DeNatale

Blog: http://talklikeaduck.denhaven2.com/
Github: rubyredrick (Rick DeNatale) · GitHub
Twitter: @RickDeNatale
WWR: http://www.workingwithrails.com/person/9021-rick-denatale
LinkedIn: http://www.linkedin.com/in/rickdenatale

[captaffy]

Josh C. wrote:

On Fri, Jul 30, 2010 at 10:49 AM, Rick DeNatale
[email protected]wrote:

[].all? {|element| element != 3 } # => true
There may be a better name than non_vacuous_all? but I can’t think of one.

How about #appall? to imply that it is a pessimistic implementation of
#all?

I think that the best name for that method would be

def all!?(&block)
!empty? && all?(&block)
end

It’s a pity it is not permitted!

so I propose

def strict_all?(&block)
!empty? && all?(&block)
end

[captaffy]

On Sun, Aug 1, 2010 at 12:28 PM, Maurizio De Santis
[email protected] wrote:

I think that the best name for that method would be

def all!?(&block)
!empty? && all?(&block)
end

No that doesn’t make sense to me all not? Not at all!

All not what?

Personally I think this attempts too hard to make the method name
short, at the expense of revealing the intention.

It’s a pity it is not permitted!

so I propose

def strict_all?(&block)
!empty? && all?(&block)
end

I don’t like this either, it doesn’t evoke the right meaning, to me at
least. What are we being strict about? I would tend to read it as
following strictly the conventional meaning of the existential
quantifier, which the existing Enumerable#all? method already does.

–
Rick DeNatale

Blog: http://talklikeaduck.denhaven2.com/
Github: rubyredrick (Rick DeNatale) · GitHub
Twitter: @RickDeNatale
WWR: http://www.workingwithrails.com/person/9021-rick-denatale
LinkedIn: http://www.linkedin.com/in/rickdenatale

[captaffy]

On Sun, Aug 1, 2010 at 12:45 PM, Rick DeNatale [email protected]
wrote:

def all!?(&block)
!empty? && all?(&block)
end

No that doesn’t make sense to me all not? Not at all!

All not what?

And right after I hit send, I realized that I’d probably expect
semantics more like:

#def all!?
def all_not?
if block_given?
all? {|element| ! yield element}
else
all? {|element| ! element}
end
end

–
Rick DeNatale

Blog: http://talklikeaduck.denhaven2.com/
Github: rubyredrick (Rick DeNatale) · GitHub
Twitter: @RickDeNatale
WWR: http://www.workingwithrails.com/person/9021-rick-denatale
LinkedIn: http://www.linkedin.com/in/rickdenatale

[captaffy]

On Sun, Aug 1, 2010 at 11:46 AM, Rick DeNatale
[email protected]wrote:

How about #appall? to imply that it is a pessimistic implementation of
#all?

Ho