I've been given this class which includes the Enumerable module:
class Team
include Enumerable # LOTS of functionality
attr_accessor :name, :players
def initialize (name)
#name = name
#players = []
end
def add_players (*players) # splat
#players += players
end
def to_s
"#{#name} team: #{#players.join(", ")}"
end
def each
#players.each { |player| yield player }
end
end
I know that yield is used to be able to call a block inside a method when a block is passed as a method argument. What then does the yield do inside this block?
Proper functioning of the included Enumerable interface's methods requires that the class implement an each method that yields successive values. The implementation in your sample code effectively delegates this logic to that of Array's each method (since #players is an Array).
Replacing yield with return would result in no values being provided, and bypassing any code block that was passed in.
You could experiment with the each implementation to verify this.
Related
I'm trying to wrap my head around delegation vs. inheritance so I'm manually delegating a version of Array. One of the specific reasons I read to do this is because when you use things like enumerables, your returned value on the inherited methods reverts back to the parent class (i.e. Array). So I did this:
module PeepData
# A list of Peeps
class Peeps
include Enumerable
def initialize(list = [])
#list = list
end
def [](index)
#list[index]
end
def each(...)
#list.each(...)
end
def reverse
Peeps.new(#list.reverse)
end
def last
#list.last
end
def join(...)
#list.join(...)
end
def from_csv(csv_table)
#list = []
csv_table.each { |row| #list << Peep.new(row.to_h) }
end
def include(field, value)
Peeps.new(select { |row| row[field] == value })
end
def exclude(field, value)
Peeps.new(select { |row| row[field] != value })
end
def count_by_field(field)
result = {}
#list.each do |row|
result[row[field]] = result[row[field]].to_i + 1
end
result
end
protected
attr_reader :list
end
end
When I instantiate this, my include and exclude function great and return a Peeps class but when using an enumerable like select, it returns Array, which prevents me from chaining further Peeps specific methods after the select. This is exactly what I'm trying to avoid with learning about delegation.
p = Peeps.new
p.from_csv(csv_generated_array_of_hashes)
p.select(&:certified?).class
returns Array
If I override select, wrapping it in Peeps.new(), I get a "SystemStackError: stack level too deep". It seems to be recursively burying the list deeper into the list during the select enumeration.
def select(...)
Peeps.new(#list.select(...))
end
Any help and THANKS!
I would recommend using both Forwardable and Enumerable. Use Forwardable to delegate the each method to your list (to satisfy the Enumerable interface requirement), and also forward any Array methods you might want to include that are not part of the Enumerable module, such as size. I would also suggest not overriding the behavior of select as it is supposed to return an array and would at the very least lead to confusion. I would suggest something like the subset provided below to implement the behavior you are looking for.
require 'forwardable'
class Peeps
include Enumerable
extend Forwardable
def_delegators :#list, :each, :size
def initialize(list = [])
#list = list
end
def subset(&block)
selected = #list.select(&block)
Peeps.new(selected)
end
protected
attr_reader :list
end
Example usage:
peeps = Peeps.new([:a,:b,:c])
subset = peeps.subset {|s| s != :b}
puts subset.class
peeps.each do |peep|
puts peep
end
puts peeps.size
puts subset.size
produces:
Peeps
a
b
c
3
2
I think that if Peeps#select will return an Array, then it is OK to include Enumerable. But, you want Peeps#select to return a Peeps. I don't think you should include Enumerable. It's misleading to claim to be an Enumerable if you don't conform to its interface. This is just my opinion. There is no clear consensus on this in the ecosystem. See "Examples from the ecosystem" below.
If we accept that we cannot include Enumerable, here's the first implementation that comes to my mind.
require 'minitest/autorun'
class Peeps
ARRAY_METHODS = %i[flat_map map reject select]
ELEMENT_METHODS = %i[first include? last]
def initialize(list)
#list = list
end
def inspect
#list.join(', ')
end
def method_missing(mth, *args, &block)
if ARRAY_METHODS.include?(mth)
self.class.new(#list.send(mth, *args, &block))
elsif ELEMENT_METHODS.include?(mth)
#list.send(mth, *args, &block)
else
super
end
end
end
class PeepsTest < Minitest::Test
def test_first
assert_equal('alice', Peeps.new(%w[alice bob charlie]).first)
end
def test_include?
assert Peeps.new(%w[alice bob charlie]).include?('bob')
end
def test_select
peeps = Peeps.new(%w[alice bob charlie]).select { |i| i < 'c' }
assert_instance_of(Peeps, peeps)
assert_equal('alice, bob', peeps.inspect)
end
end
I don't normally use method_missing, but it seemed convenient.
Examples from the ecosystem
There doesn't seem to be a consensus on how strictly to follow interfaces.
ActionController::Parameters used to inherit Hash. Inheritance ceased in Rails 5.1.
ActiveSupport::HashWithIndifferentAccess still inherits Hash.
As mentioned in the other answer, this isn't really proper usage of Enumerable. That said, you could still include Enumerable and use some meta-programming to override the methods that you want to be peep-chainable:
module PeepData
class Peeps
include Enumerable
PEEP_CHAINABLES = [:map, :select]
PEEP_CHAINABLES.each do |method_name|
define_method(method_name) do |&block|
self.class.new(super(&block))
end
end
# solution for select without meta-programming looks like this:
# def select
# Peeps.new(super)
# end
end
end
Just so you know, this really has nothing to do with inheritance vs delegation. If Peeps extended Array, you would have the exact same issue, and the exact solution above would still work.
Example code:
module Creatures
class << self
def to_h
{
squirtle: {full_name: 'Squirtle T. Turtle'},
pikachu: {full_name: 'Pikachu B. Pikachu'}
}
end
def keys
to_h.keys
end
def collect
to_h.keys.collect
end
def each
to_h.keys.each
end
end
end
module CompanionHelper
def get_companion_creature_experience(companion_data)
Creatures.each do |creature|
return companion_data[creature]["#{creature}_experience".to_sym] if companion_data.has_key?(creature)
end
end
end
include CompanionHelper
companion_data = {squirtle: {squirtle_experience: 8000}}
get_companion_creature_experience(companion_data)
Forgive me if the example is contrived. The original code is from the insurance world but I can't copy and paste it :)
The crux of the problem is I want to use Creatures.each in another module, pass it a block, and have it work just like Creatures.keys.each would work (i.e. w/ the given example companion data I get 8000 for get_companion_creature_experience(companion_data).
Currently I get Enumerator instead.
Problem is that to_h.keys.each returns Enumerator which does not expect any arguments. Pass a block inside each since you want to use it:
def each &block
to_h.keys.each &block
end
Or you can yield it:
def each
to_h.keys.each do |k|
yield k
end
end
I have a class Klass, and its constructor accepts an argument. We should be able to call methods on this object that are not defined in Klass.
We can chain multiple methods, but in the end, we have to use Klass#result to get the result like:
Klass.new(5).pred.pred.result
and the output here should be 3. I tried using method_missing in Klass and using send on the object's class, but that would have worked without the result method that I have to use. Can someone explain how this can be done with delegation?
You could do something like this:
class Klass
def initialize(number)
#number = number
end
def result
#number
end
def method_missing(method_name, *arguments, &block)
if #number.respond_to?(method_name)
#number = #number.method(method_name).call(*arguments, &block)
return self
else
super
end
end
def respond_to_missing?(method_name, include_private = false)
# be sure to implement this...
end
end
puts Klass.new(5).pred.pred.result # => 3
But it's problematic. In this particular example, since #pred returns a new object (it doesn't modify the object it was called on), we have to reassign the instance variable to the result. It works for pred and other methods that return new Integers, but some methods on Integer don't return an Integer (e.g. Integer#even). In this case you'd get this sort of behavior:
puts Klass.new(4).even?.result # => true
Depending on your particular situation, that might be what you're after. Or, it might be that in your situation all methods the object being delegated to mutate that object, rather than return new instances of the object, in which case the reassignment isn't needed.
I don't think you can use Ruby's existing Delegator and SimpleDelegator constructs, because the only way you can chain the final #result call onto the end is if every delegated call returns the instance of Klass. Using those existing constructs would cause delegated calls to return their normal return values, and the chaining would then be on whatever objects those return values return. For example, using the above code, you'd see this behavior:
puts Klass.new(5).pred.pred.class # => "Klass"
Using SimpleDelegator, you'd see this behavior
require 'delegate'
class Klass2 < SimpleDelegator
# Klass2 methods...
end
puts Klass2.new(5).pred.pred.class # => "Fixnum"
Hope that helps.
I am trying to create a macro "has_accessor_for", that accepts a symbol which is used as a parameter for an internal object that it uses (the Accessorizer object). The problem I am having is, when multiple modules do the has_accessors_for, the parameter (scope) ends up being stuck on the last value it was assigned to.
I added a puts prior to the define_method, which shows that it's scope1, and then scope2... But inside the define_method, it's scope2 always. I am looking for a way to basically encapsulate that variable, so that when it the first module calls has_accessor_for, anytime my_wut is called, it will be bound to scope1... and anytime my_bleah is called, it will be bound to scope2. But as I said, right now, both my_bleah and my_wut are bound to scope2-- If I change the order of the includes in MyModel, then they will both be bound to scope1.
class Accessorizer
def initialize(record, scope)
#record = record
#scope = scope
end
def value_for(key)
#record.send key
end
end
module Magic
def has_accessors_for(scope)
accessors = {}
puts "initial: #{scope}"
define_method :get_value_for do |key|
puts "inside method #{scope}"
accessor.value_for key
end
define_method :accessor do
accessors[:scope] ||= Accessorizer.new(self, scope)
end
end
end
module SomeAccessor
extend Magic
has_accessors_for :scope1
def my_wut
get_value_for :wut
end
end
module SomeOtherAccessor
extend Magic
has_accessors_for :scope2
def my_bleah
get_value_for :bleah
end
end
class MyModel
include SomeAccessor
include SomeOtherAccessor
attr_accessor :wut, :bleah
end
m = MyModel.new
m.wut = 'wut'
m.bleah = 'bleah'
m.my_bleah
m.my_wut
output:
initial: scope1
initial: scope2
inside method scope2
inside method scope2
Short answer: the problem is not with the closures.
Long answer:
define_method :get_value_for do |key|
puts "inside method #{scope}"
accessor.value_for key
end
On a given class there can only be one method called get_value_for - the second definition will overwrite the first.
It doesn't matter so much because you're call accessor in both cases, however that method suffers from the same problem - you define it twice and so the second definition overwrites the first and you end up with only one Accessorizer object.
I think you'll need to rethink your design here.
I have an object Results that contains an array of result objects along with some cached statistics about the objects in the array. I'd like the Results object to be able to behave like an array. My first cut at this was to add methods like this
def <<(val)
#result_array << val
end
This feels very c-like and I know Ruby has better way.
I'd also like to be able to do this
Results.each do |result|
result.do_stuff
end
but am not sure what the each method is really doing under the hood.
Currently I simply return the underlying array via a method and call each on it which doesn't seem like the most-elegant solution.
Any help would be appreciated.
For the general case of implementing array-like methods, yes, you have to implement them yourself. Vava's answer shows one example of this. In the case you gave, though, what you really want to do is delegate the task of handling each (and maybe some other methods) to the contained array, and that can be automated.
require 'forwardable'
class Results
include Enumerable
extend Forwardable
def_delegators :#result_array, :each, :<<
end
This class will get all of Array's Enumerable behavior as well as the Array << operator and it will all go through the inner array.
Note, that when you switch your code from Array inheritance to this trick, your << methods would start to return not the object intself, like real Array's << did -- this can cost you declaring another variable everytime you use <<.
each just goes through array and call given block with each element, that is simple. Since inside the class you are using array as well, you can just redirect your each method to one from array, that is fast and easy to read/maintain.
class Result
include Enumerable
def initialize
#results_array = []
end
def <<(val)
#results_array << val
end
def each(&block)
#results_array.each(&block)
end
end
r = Result.new
r << 1
r << 2
r.each { |v|
p v
}
#print:
# 1
# 2
Note that I have mixed in Enumerable. That will give you a bunch of array methods like all?, map, etc. for free.
BTW with Ruby you can forget about inheritance. You don't need interface inheritance because duck-typing doesn't really care about actual type, and you don't need code inheritance because mixins are just better for that sort of things.
Your << method is perfectly fine and very Ruby like.
To make a class act like an array, without actually inheriting directly from Array, you can mix-in the Enumerable module and add a few methods.
Here's an example (including Chuck's excellent suggestion to use Forwardable):
# You have to require forwardable to use it
require "forwardable"
class MyArray
include Enumerable
extend Forwardable
def initialize
#values = []
end
# Map some of the common array methods to our internal array
def_delegators :#values, :<<, :[], :[]=, :last
# I want a custom method "add" available for adding values to our internal array
def_delegator :#values, :<<, :add
# You don't need to specify the block variable, yield knows to use a block if passed one
def each
# "each" is the base method called by all the iterators so you only have to define it
#values.each do |value|
# change or manipulate the values in your value array inside this block
yield value
end
end
end
m = MyArray.new
m << "fudge"
m << "icecream"
m.add("cake")
# Notice I didn't create an each_with_index method but since
# I included Enumerable it knows how and uses the proper data.
m.each_with_index{|value, index| puts "m[#{index}] = #{value}"}
puts "What about some nice cabbage?"
m[0] = "cabbage"
puts "m[0] = #{m[0]}"
puts "No! I meant in addition to fudge"
m[0] = "fudge"
m << "cabbage"
puts "m.first = #{m.first}"
puts "m.last = #{m.last}"
Which outputs:
m[0] = fudge
m[1] = icecream
m[2] = cake
What about some nice cabbage?
m[0] = cabbage
No! I meant in addition to fudge
m.first = fudge
m.last = cabbage
This feels very c-like and I know Ruby
has better way.
If you want an object to 'feel' like an array, than overriding << is a good idea and very 'Ruby'-ish.
but am not sure what the each method
is really doing under the hood.
The each method for Array just loops through all the elements (using a for loop, I think). If you want to add your own each method (which is also very 'Ruby'-ish), you could do something like this:
def each
0.upto(#result_array.length - 1) do |x|
yield #result_array[x]
end
end
If you create a class Results that inherit from Array, you will inherit all the functionality.
You can then supplement the methods that need change by redefining them, and you can call super for the old functionality.
For example:
class Results < Array
# Additional functionality
def best
find {|result| result.is_really_good? }
end
# Array functionality that needs change
def compact
delete(ininteresting_result)
super
end
end
Alternatively, you can use the builtin library forwardable. This is particularly useful if you can't inherit from Array because you need to inherit from another class:
require 'forwardable'
class Results
extend Forwardable
def_delegator :#result_array, :<<, :each, :concat # etc...
def best
#result_array.find {|result| result.is_really_good? }
end
# Array functionality that needs change
def compact
#result_array.delete(ininteresting_result)
#result_array.compact
self
end
end
In both of these forms, you can use it as you want:
r = Results.new
r << some_result
r.each do |result|
# ...
end
r.compact
puts "Best result: #{r.best}"
Not sure I'm adding anything new, but decided to show a very short code that I wish I could have found in the answers to quickly show available options. Here it is without the enumerator that #shelvacu talks about.
class Test
def initialize
#data = [1,2,3,4,5,6,7,8,9,0,11,12,12,13,14,15,16,172,28,38]
end
# approach 1
def each_y
#data.each{ |x| yield(x) }
end
#approach 2
def each_b(&block)
#data.each(&block)
end
end
Lets check performance:
require 'benchmark'
test = Test.new
n=1000*1000*100
Benchmark.bm do |b|
b.report { 1000000.times{ test.each_y{|x| #foo=x} } }
b.report { 1000000.times{ test.each_b{|x| #foo=x} } }
end
Here's the result:
user system total real
1.660000 0.000000 1.660000 ( 1.669462)
1.830000 0.000000 1.830000 ( 1.831754)
This means yield is marginally faster than &block what we already know btw.
UPDATE: This is IMO the best way to create an each method which also takes care of returning an enumerator
class Test
def each
if block_given?
#data.each{|x| yield(x)}
else
return #data.each
end
end
end
If you really do want to make your own #each method, and assuming you don't want to forward, you should return an Enumerator if no block is given
class MyArrayLikeClass
include Enumerable
def each(&block)
return enum_for(__method__) if block.nil?
#arr.each do |ob|
block.call(ob)
end
end
end
This will return an Enumerable object if no block is given, allowing Enumerable method chaining