Ruby refinements subtleties - ruby

There is a pretty good documentation of the current implementation of refinements in ruby here:
http://ruby-doc.org//core-2.2.0/doc/syntax/refinements_rdoc.html,
but there are some strange corner cases.
First, include module is orthogonal to using module (one include the instance method of module while the other activates the refinement). But there is a trick to include a refinement module itself, see
Better way to turn a ruby class into a module than using refinements?.
def to_module(klass)
Module.new do
#note that we return the refinement module itself here
return refine(klass) {
yield if block_given?
}
end
end
class Base
def foo
"foo"
end
end
class Receiver
include to_module(Base) {
def foo
"refined " + super
end
}
end
Receiver.new.foo #=> "refined foo"
Strangely this refinement module can't be used with using!
m=to_module(Base) {}
m.class #=> Module
using m
#=>TypeError: wrong argument type Class (expected Module)
So using only work on the enclosing module of the refinement modules.
Secondly I wanted to use the above yield trick to be able to pass a Proc to refine (even through it only accepts a block), without resorting to converting the Proc back to source as in
https://www.new-bamboo.co.uk/blog/2014/02/05/refinements-under-the-knife/.
But using yield as in the include example does not work:
def ref_module1(klass)
Module.new do
refine(klass) {
yield
}
end
end
class Receiver1
using ref_module1(Base) {
def foo
"refined " + super
end
}
def bar
Base.new.foo
end
end
Receiver1.new.bar #=> NoMethodError: super: no superclass method `foo'
We see that Receiver1 still use Bar#foo and not the refined method.
Howewer we can use module_eval instead:
def ref_module2(klass,&b)
Module.new do
refine(klass) {
module_eval(&b)
}
end
end
class Receiver2
using ref_module2(Base) {
def foo
"refined " + super
end
}
def bar
Base.new.foo
end
end
Receiver2.new.bar #=> "refined foo"
I don't quite understand why module_eval works here and not the yield method. Inside the refinement block, the 'default_definee' is the refinement module, so module_eval which puts the 'default_definee' to self='the refinement module' should not affect it. And indeed in the 'include' example at the beginning, I get the same result when I use module_eval or a direct yield.
Can anyone explain this behavior?

Context (or binding) is the reason why module_eval works and yield doesn't in your last set of examples. It actually has nothing to do with refinements, as demonstrated below.
Starting with module_eval:
class Foo
def run(&block)
self.class.module_eval(&block)
end
end
foo = Foo.new
foo.run {
def hello
"hello"
end
}
puts foo.hello # => "hello"
puts hello => # '<main>': undefined method 'hello' for main:Object (NameError)
In Foo#run we call module_eval on Foo. This switches the context (self) to be Foo. The result is much like we had simple defined hello inside of class Foo originally.
Now let's take a look at yield:
class Foo
def run
yield
end
end
foo = Foo.new
foo.run {
def hello
"hello"
end
}
puts hello # => "hello"
puts foo.hello # => '<main>': private method 'hello' called for ...
yield simply invokes the block in its original context, which in this example would be <main>. When the block is invoked, the end result is exactly the same as if the method were defined at the top level normally:
class Foo
def run
yield
end
end
foo = Foo.new
def hello
"hello"
end
puts hello # => "hello"
puts foo.hello # => '<main>': private method 'hello' called for ...
You might notice that foo seems to have the hello method in the yield examples. This is a side effect of defining hello as a method at the top level. It turns out that <main> is just an instance of Object, and defining top level methods is really just defining private methods on Object which nearly everything else ends up inheriting. You can see this by opening up irb and running the following:
self # => main
self.class # => Object
def some_method
end
"string".method(:some_method) # => #<Method: String(Object)#some_method>
Now back to your examples.
Here's what happens in the yield example:
def ref_module1(klass)
Module.new do
refine(klass) {
yield
}
end
end
class Receiver1
# like my yield example, this block is going to
# end up being invoked in its original context
using ref_module1(Base) {
def foo
"I'm defined on Receiver1"
end
}
def bar
# calling foo here will simply call the original
# Base#foo method
Base.new.foo
end
end
# as expected, if we call Receiver1#bar
# we get the original Base#foo method
Receiver1.new.bar # => "foo"
# since the block is executed in its original context
# the method gets defined in Receiver1 -- its original context
Receiver1.new.foo # => "I'm defined on Receiver1"
As for module_eval, it works in your examples because it causes the block to be run in the context of the new module, rather than on the Receiver1 class.

Related

ruby - override method and then revert

I am trying to find a way that I can override a method, do something, and then revert without leaving any artifacts around.
I have implemented this using mocha but obviously this is not going to fly in a production app. Notice the new method has parameters and the old one does not.
Example as follows
require 'rubygems'
require 'mocha'
class Example
def to_something
self.stubs(:attribs => other(1))
r = attribs_caller
self.unstub(:attribs)
r
end
def other(int)
{"other" => int }
end
def attribs_caller
attribs
end
def attribs
{"this" => 1 }
end
end
a1 = Example.new
puts a1.attribs_caller #=> this1
puts a1.to_something #=> other1
puts a1.attribs_caller #=> this1
class String
alias orig_reverse reverse
def reverse(n)
'fooled you. '*n
end
end
puts "ab".reverse(2)
#=> fooled you fooled you
# clean up:
class String
alias reverse orig_reverse
remove_method(:orig_reverse)
end
puts "ab".reverse #=> ba
Another way to do that, without creating an extra method, is this:
class Foo
def bar
:old_method
end
end
Foo.new.bar # => :old_method
$old_method = Foo.new.method(:bar)
class Foo
def bar
:new_method
end
end
Foo.new.bar # => :new_method
class Foo
define_method($old_method.name, &$old_method)
end
Foo.new.bar # => :old_method
I think that this is better than using an alias method. In Ruby methods are, also, objects. I just take the reference of the object before destructing the association of the object (the method) with the class. After I add the same method. It also works if you use the undef keyword to remove the method from the class. The bad point is that you have to have an object of the class to take the reference of the method.

Ruby - How to use the method parameter as the name of the variable?

How would I use the parameter value as the instance variable name of an object?
This is the object
Class MyClass
def initialize(ex,ey)
#myvar = ex
#myothervar = ey
end
end
I have the following method
def test(element)
instanceofMyClass.element #this obviously doesnt work
end
How can I have the test method return either myvar or myothervar value depending on the element parameter. I don't want to write an if condition though, I want to pass myvar or myother var via element to the object instance if possible.
def test(element)
instanceofMyClass.send(element.to_sym)
end
You'll get a missing method error if instanceofMyClass doesn't respond to element.
def test(element)
instanceofmyclass.instance_variable_get element
end
test :#myvar # => ex
test :#myothervar # => ey
I like the simplicity of send(), though one bad thing with it is that it can be used to access privates. The issue is still remains solution below, but at least then it's explicitly specified, and reader can see which methods are to be forwarded. The first one just uses delegation, while the second one uses more dynamic way to define methods on the fly.
require 'forwardable'
class A
extend Forwardable
def_delegators :#myinstance, :foo, :bar
class B
def foo
puts 'foo called'
end
def bar
puts 'bar called'
end
def quux
puts 'quux called'
end
def bif
puts 'bif called'
end
end
def initialize
#myinstance = B.new
end
%i(quux bif).each do |meth| # note that only A#quux and A#bif are defined dynamically
define_method meth do |*args_but_we_do_not_have_any|
#myinstance.send(meth)
end
end
end
a = A.new
a.foo
a.bar
a.quux
a.bif

Closure doesn't work

If a block is a closure, why does this code does not work, and how to make it work?
def R(arg)
Class.new do
def foo
puts arg
end
end
end
class A < R("Hello!")
end
A.new.foo #throws undefined local variable or method `arg' for #<A:0x2840538>
Blocks are closures and arg is indeed available inside the Class.new block. It's just not available inside the foo method because def starts a new scope. If you replace def with define_method, which takes a block, you'll see the result you want:
def R(arg)
Class.new do
define_method(:foo) do
puts arg
end
end
end
class A < R("Hello!")
end
A.new.foo # Prints: Hello!
If you define the class dynamically, you can alter it as you like:
def R(arg)
c = Class.new
# Send the block through as a closure, not as an inline method
# definition which interprets variables always as local to the block.
c.send(:define_method, :foo) do
arg
end
c
end
class A < R("Hello!")
end
puts A.new.foo.inspect
# => "Hello!"

Including methods into blocks

Anyone know how to get this to work if it's possible?
class Foo
def self.go(&block)
class << block
include Bar
end
puts "Within Foo#go: #{block.methods.include? 'baz'}"
block.call
end
end
module Bar
def baz
puts "I'm happily in my place!"
end
end
Foo.go {
puts "Within actual block: #{methods.include? 'baz'}"
baz
}
This gets the output:
Within Foo#go: true
Within actual block: false
NameError: undefined local variable or method ‘baz’ for main:Object
EDIT: when I print out the block's class in Foo#go, it's Proc, but when I print it out within the Proc, it's Object. Could this be related?
You can't do this. The reason for what you're seeing is that there are two different contexts here. One is the context of the block, which closes over the context where it's defined. The other is the context of the Proc object wrapper, which is just the same as any other object context and completely unrelated to the context of the block itself.
I think the closest you'll get is to instance_eval the block using a context object that has the methods you want, but then the block won't have access to the self that existed where it was defined. It's up to you whether that makes sense for the method you want to write.
The other option is to pass the block an actual receiver for the baz method.
You could use eval with Proc#binding:
module Bar
def baz
puts "hi from baz!"
end
end
def go(&block)
eval('include Bar', block.binding)
block[]
end
baz #=> NameError
go { baz } #=> "hi from baz!"
baz #=> "hi from baz!"
But unless you use a mixin/mixout framework (like mixico or mixology), you'll be placing the methods from the included module into the lexical scope, so they'll still be accessible once the block returns.
require 'rubygems'
require 'mixico'
module Bar
def baz
puts "hi from baz!"
end
end
def go(&block)
Module.mix_eval(Bar, &block)
end
baz #=> NameError
go { baz } #=> "hi from baz!"
baz #=> NameError
Here's a good article on different ways to use a DSL from within a block.
Another alternative, following on from rampion, is to dup the context of the block before mixing into it, this way you're not messing up the context after you've finished.
module Bar
def baz
puts "hi from baz!"
end
end
def go(&block)
dup_context = eval('self', block.binding).dup
dup_context.send(:include, Bar)
dup_context.instance_eval &block
end
Note this will only be useful to you if you're not running any mutator methods in the block

Is 'yield self' the same as instance_eval?

Is there any difference if you define Foo with instance_eval: . . .
class Foo
def initialize(&block)
instance_eval(&block) if block_given?
end
end
. . . or with 'yield self':
class Foo
def initialize
yield self if block_given?
end
end
In either case you can do this:
x = Foo.new { def foo; 'foo'; end }
x.foo
So 'yield self' means that the block after Foo.new is always evaluated in the context of the Foo class.
Is this correct?
Your two pieces of code do very different things. By using instance_eval you're evaluating the block in the context of your object. This means that using def will define methods on that object. It also means that calling a method without a receiver inside the block will call it on your object.
When yielding self you're passing self as an argument to the block, but since your block doesn't take any arguments, it is simply ignored. So in this case yielding self does the same thing as yielding nothing. The def here behaves exactly like a def outside the block would, yielding self does not actually change what you define the method on. What you could do is:
class Foo
def initialize
yield self if block_given?
end
end
x = Foo.new {|obj| def obj.foo() 'foo' end}
x.foo
The difference to instance_eval being that you have to specify the receiver explicitly.
Edit to clarify:
In the version with yield, obj in the block will be the object that is yielded, which in this case is is the newly created Foo instance. While self will have the same value it had outside the block. With the instance_eval version self inside the block will be the newly created Foo instance.
They are different. yield(self) does not change the value of self inside the block, while instance_eval(&block) does.
class Foo
def with_yield
yield(self)
end
def with_instance_eval(&block)
instance_eval(&block)
end
end
f = Foo.new
f.with_yield do |arg|
p self
# => main
p arg
# => #<Foo:0x100124b10>
end
f.with_instance_eval do |arg|
p self
# => #<Foo:0x100124b10>
p arg
# => #<Foo:0x100124b10>
end
You just can drop the self keyword
class Foo
def initialize
yield if block_given?
end
end
Update from comments
Using yield there is a bit new to my taste, specially when used outside irb.
However there is a big and significant difference between instance_eval approach and yield approach, check this snippet:
class Foo
def initialize(&block)
instance_eval(&block) if block_given?
end
end
x = Foo.new { def foo; 'foo'; end }
#=> #<Foo:0xb800f6a0>
x.foo #=> "foo"
z = Foo.new #=> #<Foo:0xb800806c>
z.foo #=>NoMethodError: undefined method `foo' for #<Foo:0xb800806c>
Check this one as well:
class Foo2
def initialize
yield if block_given?
end
end
x = Foo2.new { def foo; 'foo'; end } #=> #<Foo:0xb7ff1bb4>
x.foo #=> private method `foo' called for #<Foo2:0xb8004930> (NoMethodError)
x.send :foo => "foo"
z = Foo.new #=> #<Foo:0xb800806c>
z.send :foo => "foo"
As you can see the difference is that the former one is adding a singleton method foo to the object being initialized, while the later is adding a private method to all instances of Object class.

Resources