Skip to content

Latest commit

 

History

History
10469 lines (6930 loc) · 268 KB

README.md

File metadata and controls

10469 lines (6930 loc) · 268 KB

Book Notes

Just my notes from various books/docs on programming. This is public only because I don't want to waste a github private slot, so don't look.

TODO

  • Underscore.js Docs
  • Exceptional Ruby
  • Algorithms in a Nutshell
  • Practical Vim
  • Code Complete
  • Cracking the Coding Interview
  • Elements of Programming Interviews
  • Programming Interviews Exposed

Practical Object-Oriented Design in Ruby

Writing Code That Embraces Change:

Always wrap instance variables in accessor methods instead of directly referring to variables.

Enforce single responsibility everywhere

Methods that have a single responsibilty:

  • Expose previously hidden qualities
  • Avoid the need for comments
  • Encourage reuse
  • Are easy to move to another class
  • (me: easier to understand)
  • (me: easier to test)

Because you are writing changeable code, you are best served by postponing decisions until you are absolutely forced to make them. Any decision you make in advance of an explicit requirement is just a guess. Don’t decide; preserve your ability to make a decision later.

OpenStruct vs Struct

The difference between the two is that Struct takes position order initialization arguments while OpenStruct takes a hash for its initialization and then derives attributes from the hash.

[OpenStruct, Struct blog posting] (http://blog.steveklabnik.com/posts/2012-09-01-random-ruby-tricks--struct-new)

TODO: [Official Docs] (http://www.ruby-doc.org/)

Operator I keep forgetting to use
@wheel ||= Wheel.new(rim, tire)
specifying defaults using fetch (page 49)
def initialize(args)
  @chainring = args.fetch(:chainring, 40) 
  @cog = args.fetch(:cog, 18) 
  @wheel = args[:wheel]
end

My example

# defaults to blue, nil is transparent
def set_pixel(args)
  pixel = args[:color] || :blue
  return "setting to #{pixel}"
end

pp set_pixel({:color => :red}) #=> "setting to red"
pp set_pixel({}) #=> "setting to blue"
pp set_pixel({:color => false}) #=> "setting to blue"  WRONG, we want false for transparent

# use fetch instead so we can pass nil or false

def set_pixel(args)
  pixel = args.fetch(:color, :blue) # :blue returned if no key :color 
  return "setting to #{pixel}"
end

pp set_pixel({:color => :red}) #=> "setting to red"
pp set_pixel({}) #=> "setting to blue"
pp set_pixel({:color => false}) #=> "setting to false"
specifying defaults by merging a defaults hash
def initialize(args)
  args = defaults.merge(args) 
  @chainring = args[:chainring]
  # ...
end

def defaults
  {:chainring => 40, :cog => 18}
end

Law Of Demeter notes:

Using delegation to hide tight coupling is not the same as decoupling the code.

Writing Inheritable Code

Recognize the Antipatterns
  • An object that uses a variable with a name like type or category to determine what message to send to self contains two highly related but slightly different types.
  • When a sending object checks the class of a receiving object to deter- mine what message to send, you have overlooked a duck type.
Insist on the Abstraction

All of the code in an abstract superclass should apply to every class that inherits it.

Subclasses that override a method to raise an exception like “does not implement” are a symptom of this problem. When subclasses override a method to declare that they do not do that thing they come perilously close to declaring that they are not that thing. Nothing good can come of this.

def default_tire_size
  raise NotImplementedError, "This #{self.class} cannot respond to:"
end
Honor the Contract

Subclasses agree to a contract; they promise to be substitutable for their superclasses.

Use the Template Method Pattern

???

Preemptively Decouple Classes

Avoid writing code that requires its inheritors to send super; instead use hook messages to allow subclasses to participate while absolving them of responsibility for knowing the abstract algorithm.

TODO: Forwardable

Ruby Cookbook

A way strings are made that I keep forgetting: %Q and %q and matched chars

str = %q{This is a string}
w = 'nother'
str = %Q|This is a#{w} string|

I also keep forgetting here doc syntax, ('-' lets you indent the end token)

str = <<-EOS
Line one
Line two
line three
EOS

Printf style string formatting:

'this is a %s' % 'string'

ERB outside of rails

require 'erb'

template = %q{
  Contents:
    <% array.each do |element| -%>
      <%= element %>
    <% end -%>
}

template = ERB.new template, nil, '-'

array = %w{one two three four five}
puts template.run(binding)

using () in regex in split()

'one two three four five'.split(/\s+/) #=> ["one", "two", "three", "four", "five"]
'one two three four five'.split(/(\s+)/) #=> ["one", " ", "two", " ", "three", " ", "four", " ", "five"]

print smiley to terminal

puts "\xe2\x98\xBA"

ring bell

puts "\a"

String#each was removed in 1.9 and replaced with String#each_line and String#each_char

String#scan (can also take a code block)

"one two three four".scan(/\w+/) #=>  ["one", "two", "three", "four"] 

Duck typing note:

The idea to take to heart here is the general rule of duck typing: to see whether provided data implements a certain method, use respond_to? instead of checking the class.This lets a future user (possibly yourself!) create new classes that offer the same capability, without being tied down to the preexisting class structure. All you have to do is make the method names match up.

String#slice and [] work the same

If you pass indexes:

s = 'My kingdom for a string!'
puts s.slice(3,7) # => "kingdom"
puts s[3,7]       # => "kingdom"

If you pass a regex:

puts "abcd".slice(/../) # => 'ab'
puts "abcd"[/../] # => 'ab'

Wrapping long lines

def wrap(s, width=78)
  s.gsub(/(.{1,#{width}})(\s+|\Z)/, "\\1\n")
end

All ways to make regex

/something/
Regexp.new("something")
Regexp.compile("something")
%r{something}

(regex) sub and gsub

re = /this/im
'this'.sub(re, 'that')
'this'.sub!(re, 'that')

'this'.gsub(re, 'that')
'this'.gsub!(re, 'that')

(regex) perlish match

'this' =~ re #=> 0
re =~ 'this' #=> 0

(regex) slice and its other form (works with regex memory too)

'this and that'.slice(/\s.*\s/) #=> ' and '
'this and that'[/\s.*\s/] #=> ' and '

'this and that'[/\s(.*)\s/] # memory
puts $1 #=> 'and'

(regex) sub/gsub with hash param

upcase_hash = ('a'..'e').to_a.inject({}) {|accum, x| accum[x]=x.upcase; accum }
upcase_hash #=> {"a"=>"A", "b"=>"B", "c"=>"C", "d"=>"D", "e"=>"E"}
str = 'here, there, everywhere'
puts str.gsub(/e/, upcase_hash) #=> "hErE, thErE, EvErywhErE" 

(regex) sub/gsub with block param

str = 'here, there, everywhere'
puts str.gsub(/e/) {|m| m.upcase! } #=> hErE, thErE, EvErywhErE

(regex) case statement

string = '123abc'

case string
when /^[a..zA..Z]+$/
  "All Letters"
when /^[0-9]+$/
  "All Numbers"
else
  'Mixed'
end #=> 'Mixed'

Regexp::union (my really uncreative example)

victim = 'this is my string'
re = Regexp.union('this', 'string')
victim.gsub!(re, '_')
puts victim #=> "_ is my _"

Array operators

[1,2,3] << [4,5,6] #=> [1, 2, 3, [4, 5, 6]] 
[1,2,3] + [4,5,6] #=> [1, 2, 3, 4, 5, 6]

You can split out an array into its components

array = [:red, :green, :blue]
c, a, b = array
a # => :green
b # => :blue
c # => :red

You can even use the splat operator to extract items from the front of the array:

a, b, *c = [12, 14, 178, 89, 90]
a # => 12
b # => 14
c # => [178, 89, 90]

To ensure that duplicate values never get into your list, use a Set instead of an array. If you try to add a duplicate element to a Set, nothing will happen.

require 'set'
survey_results = [1, 2, 7, 1, 1, 5, 2, 5, 1]
distinct_answers = survey_results.to_set
# => #<Set: {5, 1, 7, 2}>

Numeric#step

0.step(10, 2) {|x| print x, ' ' } #=> '0 2 4 6 8 10 '

Using a hash with empty array default values

hash = Hash.new { |hash, key| hash[key] = [] }
hash[:a] << 1
hash[:b] << 2
hash[:b] << 3
pp hash #=> {:a=>[1], :b=>[2, 3]}

Reverse lookup for hash

hash = {one:1, two:2, three:3}
pp hash.invert[1] #=> :one

Rails 4 Test Prescriptions

Prescriptions
  1. Use the TDD process to create and adjust your code’s design in small, incremental steps.

  2. In a test-driven process, if it is difficult to write tests for a feature, strongly consider the possibility that the underlying code needs to be changed.

  3. Initializing objects is a good starting place for a TDD process. Another good approach is to use the test to design what you want a successful interaction of the feature to look like.

  4. When possible, write your tests to describe your code’s behavior, not its implementation.

  5. Keeping your code as simple as possible allows you to focus complexity on the areas that really need complexity.

  6. Choose your test data and test-variable names to make it easy to diagnose failures when they happen. Meaningful names and data that doesn’t overlap are helpful.

  7. Test isolation makes it easier to understand test failures by limiting the scope of potential locations where the failure might have occurred.

  8. Your tests are also code. Specifically, your tests are code that does not have tests.

  9. If you find yourself writing tests that already pass given the current state of the code, that often means you’re writing too much code in each pass.

  10. Refactoring is where a lot of design happens in TDD, and it’s easiest to do in small steps. Skip it at your peril.

  11. Try to extract methods when you see compound Booleans, local variables, or inline comments.

  12. Fixtures are particularly useful for global semi-static data stored in the database.

  13. Your go-to build strategy for factory_girl should be build_stubbed unless there is a need for the object to be in the database during the test.

  14. Avoid defining associations automatically in factory_girl definitions. Set them test by test, as needed. You’ll wind up with more manageable test data.

  15. Use partial doubles when you want to ensure most of your real object behavior. Use full doubles when the behavior of the stubbed object doesn't matter - only its public interface does.

  16. The use of the allow_any_instance_of stub modifier often means the underlying code being tested could be refactored with a more useful method to stub.

  17. If you're stubbing methods that do not belong to your program, think about whether the code would be better if restructured to wrap the external behavior.

  18. A stubbed method that returns a stub is usually okay. A stubbed method that returns a stub that itself contains a stub probably means your code is too dependent on the internals of other objects.

  19. Don't mock what you don't own. (Not convinced by this one at all)

  20. A controller test should test controller behavior. A controller test should not fail because of problems in the model.

  21. When testing for view elements, try to test for DOM classes that you control rather than text or element names that might be subject to design changes.

  22. When testing a Boolean condition, make sure to write a test for both halves of the condition.

  23. By far the biggest and easiest trap you can fall into when dealing with integration tests is the temptation to use them like unit tests.


Setting up RSpec (for minitesters)
group :development, :test do 
  gem 'rspec-rails', '~> 3.1'
end

$ rails generate rspec:install
Details (better details in doc notes)

The actual spec is defined with it(), which takes an optional string argument that documents the spec, and then a block that is the body of the spec. The string argument is not used internally to identify the spec—you can have multiple specs with the same description string.

For single-line tests in which a string description is unnecessary, we use specify to make the single line read more clearly, such as this:

specify { expect(user.name).to eq("fred") }

Why not just use it('') instead of specify() , I'm confused, looks better?

expect() takes any object as an argument and returns ExpectationTarget

ExpectationTarget holds on to the object and itself responds to two messages, to() and not_to()

Both to() and not_to() expect as an argument a RSpec matcher.

RSpec matcher responds to a matches?() method.

Look at some objects

RSpec.describe "Something" do
    puts expect(true) #=> RSpec::Expectations::ExpectationTarget
    puts eq(3) #=> RSpec::Matchers::BuiltIn::Eq
    puts be_truthy #=> RSpec::Matchers::BuiltIn::BeTruthy
    puts expect(3).to eq(3) #=> true
    puts expect(3).to be_truthy #=> true
end

be() is weird

RSpec.describe "Something" do
  puts be #=> RSpec::Matchers::BuiltIn::Be
  puts be(3) #=> RSpec::Matchers::BuiltIn::Equal
end

Both to and not_to are ordinary Ruby methods that expect as an argument an RSpec matcher. There's nothing special about an RSpec matcher; at base it's just an object that responds to a matches? method. There are several predefined matchers and you can write your own.

let(:project){ Project.new }
let(:task){ Task.new }

Using let, you can make a variable available within the current describe without having to place it inside the before block and without having to make it an instance variable.

This version of let() will always run, its not lazy

let!(:project){ Project.new }

Any matcher of the form be_whatever or be_a_whatever assumes an associated whatever? method—with a question mark—on the actual object and calls it.

expect(task).not_to be_complete
task.mark_completed
expect(task).to be_complete

All work as pending

it 'something'

it 'something', :pending do
end

it 'something' do
  pending 'not ready'
end

In RSpec 3 all pending specs are actually run if there is code in the block part of the spec. The code is executed, with any failure in the pending spec treated as a pending result, rather than a failure result. However, if the code in the pending spec passes, you'll get an error that effectively means, "You said this was pending, but lo and behold, it works. Maybe it's not actually pending anymore; please remove the pending status."

( NOTE: The following doesn't seem quite right )

If you want the spec to not run, and not test for whether it works, employ the preceding syntax but use skip instead of pending. Alternative, you can prefix the method name with x, as in xit or xdescribe. A skipped test will not run, meaning you won't get any notification if the test suddenly starts to pass.

Rationale for not putting business logic in controller (in this case its put inside a CreatesProject factory:

We’ve been able to cover the controller logic in just these two short tests because we placed the business logic in the action object. If we hadn’t, all those tests we wrote for CreatesProject would be part of the controller test suite. As controller tests, they would run slower. More importantly, the tests would potentially be separated from the code where the expected failure would occur, making them less likely to drive design and less likely to be useful in trou- bleshooting.

We want to make the controller test completely isolated from the action object that it interacts with. The key insight is that the controller test needs to test only the behavior of the controller itself—the fact that the controller calls the action object with the correct parameters and uses the values as expected. Whether the action object works correctly or even if it exists is a problem for the action object test. When testing the controller, the controller’s behavior is what’s important, not the action object.

Debugging is twice as hard as writing the code in the first place. Therefore, if you write the code as cleverly as possible, you are, by definition, not smart enough to debug it.

  1. A test is straightforward if its purpose is immediately understandable.

  2. A test is well defined if running the same test repeatedly gives the same result.

  3. A test is independent if it does not depend on any other tests or external data to run.

  4. A truthful test accurately reflects the underlying code—it passes when the underlying code works, and fails when it does not. This is easier said than done.

And here we have an inadvertent admission, TDD is bullshit. TDD is a little test first, and a lot test after.

When the main cases are done, you try to think of ways to break the existing code. Sometimes you’ll notice something as you’re writing code to pass a previous test, like, “Hey, I wonder what would happen if this argument were nil?” Write a test that describes what the output should be and make it pass. Refactoring gets increasingly important here because special cases and error conditions tend to make code complex, and managing that complexity becomes really important to future versions of the code. The advantage of waiting to do special cases at the end is that you already have tests to cover the normal cases, so you can use those to check your new code each step of the way.

When you break out related attributes into their own class, as in this Name example (below), you'll often find it's much easier to add complexity when you have a dedicated place for that logic. When you need middle names or titles, it's easier to manage that in a separate class than it would be if you had a half implementation of names in multiple classes. You'll also find that these small classes are easy to test because Name no longer has a dependency on the database or any other code. Without dependencies, it's easy to set up and write fast tests for name logic.

class Name

  attr_reader :first_name, :last_name

  def initialize(first_name, last_name)
    @first_name, @last_name = first_name, last_name
  end

  def full_name
    "#{first_name} #{last_name}"
  end

  def sort_name
  "#{last_name}, #{first_name}"
  end

end

class User < ActiveRecord::Base

  delegate :full_name, :sort_name, to: :name #<-- name

  def name #<-- name
    Name.new(first_name, last_name)
  end

end

Two styles writing tests with multiple assertions

it "marks a task complete" do
  task = tasks(:incomplete)
  task.mark_complete
  expect(task).to be_complete
  expect(task).to be_blocked
  expect(task.end_date).to eq(Date.today.to_s(:db))
  expect(task.most_recent_log.end_state).to eq("completed")
end

describe "task completion" do
  let(:task) {tasks(:incomplete)}
  before(:example) { task.mark_complete }
  specify { expect(task).to be_complete }
  specify { expect(task).to be_blocked }
  specify { expect(task.end_date).to eq(Date.today.to_s(:db)) }
  specify { expect(task.most_recent_log.end_state).to eq("completed") }
end

The tradeoff is pretty plain: the one-assertion-per-test style has the advantage that each assertion can fail independently—when all the assertions are in a single test, the test bails on the first failure. In the all-in-one test, if expect(task).to be_complete fails, you won’t even get to the check for expect(task).to be_blocked. If all the assertions are in separate tests, everything runs indepen- dently but it’s harder to determine how tests are related. There are two signif- icant downsides to the one-assertion style: first, there can be a significant speed difference since the single-assertion-per-test version will run the com- mon setup multiple times, and second, the one-assertion style can become difficult to read, especially if the setup and test wind up with some distance between them.

Often I compromise by making my first pass at TDD in the one-assertion-per- test style, which forces me to work in baby steps and gives me a more accurate picture of what tests are failing. When I’m confident in the correctness of the code, I consolidate related assertions, giving me the speed benefit moving forward.

About shoulda matchers (book discourages)

describe Task do
it { should belong_to(:project) }
it { should belong_to(:user) }
it { should ensure_length_of(:name) }
end

Tests like that are not particularly valuable for a TDD process because they are not about the design of new features. If you’re doing the TDD process, you shouldn’t start from the idea that your Task belongs to a Project. Rather, as you describe features the relationship is implied from the feature tests that you’re writing. More operationally, this means that in a good TDD process, any condition in the code that would cause a direct test like those Shoulda matchers to fail would also cause another test to fail. In which case, what’s the point of the Shoulda matcher?

Testing ActiveRecord Finders

Be aggressive about extracting compound finder statements to their own method, in much the same way and for much the same reason as I recommended for compound Boolean logic. The methods are easier to understand and reuse if they are bound together behind a method name that defines the intent of the method. When we talk about mock objects you'll also see that having finders called behind other methods makes it much easier to avoid touching the database when you don't need to.

Shared example

Directory spec/support is loaded by default rails_helper.rb before specs are run. Put shared example definitions in there.

RSpec.shared_examples "sizeable" do
  let(:instance) { described_class.new } #<--- notice described_class.new

  it "knows a one-point story is small" do
    allow(instance).to receive(:size).and_return(1)
    expect(instance).to be_small
  end

  it "knows a five-point story is epic" do
    allow(instance).to receive(:size).and_return(5)
    expect(instance).to be_epic
  end
end

Note be_epic and be_small are predicate matchers for the methods small?() and epic?()

In spec files

RSpec.describe Task do
  it_should_behave_like "sizeable"
end

There are a couple of other ways to invoke a shared example group - RSpec defines synonymous methods include_examples and it_behaves_like.

When invoking the group, you can also have the it_should_behave_like method take a block argument. Inside that block, you can use let statements to define variables, which are then visible to the shared example specs. In other words, an alternative to creating an instance with described_class is to place the burden on the calling spec to create a variable and give it an appropriate name in the it block.

Write Your Own RSpec Matchers
RSpec::Matchers.define :be_of_size do |expected|
  match do |actual|
    actual.total_size == expected
  end
end

Remember: the expected value is the value defined by the test, and the actual value is the value defined by the code. Here is the form in which the matcher gets called:

expect(actual_value).to be_of_size(expected_value)

Adding Data to Tests

Why Fixtures Are a Pain

  • Fixtures are global
  • Fixtures are spread out
  • Fixtures are distant
  • Fixtures are brittle
FactoryGirl basic factory creation
  • build(:project, ...) returns model instance without saving to database

  • create(:project, ...) returns model instance and saves to database

  • attributes_for(:project) returns a hash of all the attributes in the factory that are suitable for passing to AR#new or AR#create. Used most often for creating hash that will be sent as params to a controller test.

  • build_stubbed(:project), which is almost magical. Like build, it returns an unsaved model object. Unlike build, it assigns a fake ActiveRecord ID to the model and stubs out database-interaction methods (like save) such that the test raises an exception if they are called.

This is the strategy by which I determine which of these methods to use:

  • Use create only if the object absolutely must be in the database. Typically, this is because the test code must be able to access it via an ActiveRecord finder. However, create is much slower than any of the other methods, so it’s also worth thinking about whether there’s a way to structure the code so that persistent data is not needed for the test.

  • In all other cases, use build_stubbed, which does everything build does, plus more. Because a build_stubbed object has a Rails ID, you can build up real Rails associations and still not have to take the speed hit of saving to the database.

Associations and Factories

Create a project instance to go with factory, (weird syntax) (and arent those supposed to be methods, why are there colons in there?)

FactoryGirl.define do 
  factory :task do
    title: "To Something" 
    size: 3
    project
  end 
end
task = FactoryGirl.create(:task)

Creates a factory for both task and project.

By default, however, even if you call the parent factory with build, the subordinate factory is still called with create. This is a side effect of how Rails manages associations. The associated object needs an ID so that the parent object can link to it, and in Rails you get an ActiveRecord ID only when an ActiveRecord instance is saved to the database.

As a result, even if you use the build strategy specifically to avoid slow and unnecessary database interaction, if the factory has associations you will still save objects to the database. Since those associated factories may themselves have associations, if you aren’t careful you can end up saving a lot of objects to the database, resulting in prohibitively slow tests.

It’s exactly this characteristic of factory_girl that has made it unwelcome in some circles, particularly if the people in those circles have to maintain large, unwieldy test suites. Factory-association misuse can be a big cause of a slow test suite, as tests create many more objects than they need to because of factory_girl associations.

Since the build_stubbed strategy assigns an ID to the objects being created, using build_stubbed sidesteps the whole issue. If a factory with associations is instantiated using build_stubbed, then by default all the associations are also invoked using build_stubbed. That solves the problem as long as you always use build_stubbed.

My preferred strategy is to not specify attributes in factories at all, and if I need associated objects in a specific test, I explicitly add them to the test at the point they’re needed. Why?

  • The surest way to keep factory_girl from creating large object trees is to not define large object trees.
  • Tests that require multiple degrees of associated objects often indicate improperly factored code. Making it a little harder to write associations in tests nudges me in the direction of code that can be tested without associations.

I'm not sure about this advice, what if you have lots of foreign key constraints, or models that are closely tied together

Stubs, Mocks, Spies

A stub is a fake object that returns a predetermined value for a method call without calling the actual method on an actual object.

allow(thing).to receive(:name).and_return("Fred")

A mock is similar to a stub, but in addition to returning the fake value, a mock object sets a testable expectation that the method being replaced will actually be called in the test. If the method is not called, the mock object triggers a test failure. You can write the following snippet to create a mocked method call instead of a stub, using expect instead of allow:

expect(thing).to receive(:name).and_return("Fred")

There is a third test-double pattern, called a spy. A spy is often declared like a stub, but allows you to specify a testable expectation later in the test. Typically, we would place the body of the test in between.

allow(thing).to receive(:name).and_return("Fred")
# body of test
expect(thing).to have_received(:name)

Three types here:

allow(thing).to receive(:name).and_return("Fred")


expect(thing).to receive(:name).and_return("Fred")


allow(thing).to receive(:name).and_return("Fred")
...
expect(thing).to have_received(:name)

Using spies mitigates a common criticism of mock-object testing, which is that it can be difficult to look at a mock test and see exactly what behavior is being tested for. (I don't see how this is more clear that an ordinary mock, doesn't a spy just turn a stub into a mock?)

In RSpec, as in many Ruby double libraries, there are two kinds of fake objects. You can create entire objects that exist only to be stubs, which we'll call full doubles, or you can stub specific methods of existing objects, which we'll call partial doubles.

A partial double is useful when you want to use a "real" ActiveRecord object but you have one or two dangerous or expensive methods you want to bypass. A full double is useful when you're testing that your code works with a specific API rather than a specific object - by passing in a generic object that responds to only certain methods, you make it hard for the code to assume anything about the structure of the objects being coordinated with.

Creating full double

twin = double(first_name: "Paul", weight: 100)

For when you know your double needs to mimic a specific object, RSpec provides the concept of a verifying double. A verifying double checks to see whether messages passed to the double are actually real methods in the application. RSpec has a few methods to allow you to declare what to verify the double against:

instance_twin = instance_double("User")
instance_twin = instance_double(User)
class_twin = class_double("User")
class_twin = class_double(User)
object_twin = object_double(User.new)

instance_double will not recognize methods defined via method_missing, object_double will

instance_double uses method_defined? and class_double uses responds_to?

In addition to verifying the existence of the method, RSpec double verification ensures that the arguments passed to the method are valid. The doubled method will also have the same public/protected/private visibility as the original method. (How does it do this?)

You might use a full double object to stand in for an entire object that is unavailable or prohibitively expensive to create or call in the test environment.

Partial stub example (or is it a partial double spy?)

# this is a pointless test
it "stubs an object" do
  project = Project.new(name: "Project Greenlight")
  allow(project).to receive(:name) 
  expect(project.name).to be_nil # <--- this is pointless, only testing rspec-mocks
end

Stubbing instances of an Class

allow_any_instance_of(Project).to receive(:save).and_return(false)

The RSpec docs explicitly recommend not using this feature if possible, since it is "the most complicated feature of rspec-mocks, and has historically received the most bug reports."

A very common use of stub objects is to simulate exception conditions. If you want your stubbed method to raise an exception, you can use the and_raise method, which takes an exception class and an optional message:

allow(stubby).to receive(:user_count).and_raise(Exception, "oops")

The return values of the stubbed method walk through the values passed to and_return. Note that the values don't cycle; the last value is repeated over and over again.

allow(project).to receive(:user_count).and_return(1, 2)

The expectation is that some method takes a block argument, and we want to pass through method and send arg to the block.

allow(project).to receive(:method).and_yield("arg")

Prescription 19. Don't mock what you don't own.

One reason to mock only methods you control is, well, that you control them. One danger in mocking methods is that your mock either doesn't receive or doesn't return a reasonable value from the method being replaced. If the method in question belongs to a third-party framework, the chance that it will change without you knowing increases and thus the test becomes more brittle.

I'm not sure about the above paragraph. If a third party lib changes, your stuff is going to break (or fail?) anyway. Is it the difference between failing and being brittle?

More importantly, mocking a method you don't own couples your test to the internal details of the third-party framework. By implication, this means the method being tested is also coupled to those internal details. That is bad, not just if the third-party tool changes, but also if you want to refactor your code; the dependency will make that change more complicated.

The solution, in most cases, is to create a method or class in your application that calls the third-party tool and stubs that method (while also writing tests to ensure that the wrapper does the right thing).

Is the above overkill?

Testing Controllers and Views

Note: controller testing doesn't test routes

Evaluating Controller Results

  • Did it return the expected HTTP status code? RSpec provides the response.status object and the have http_status matcher for this purpose.

  • Did it pass control to the expected template or redirected controller action? Here we have the render_template and redirect_to matchers.

  • Did it set the values that the view will expect? For this we have the special hash objects assigns, cookies, flash, and session.

Rails controller tests do not - I repeat, do not - follow the redirect.

Example of a controller test

it "shows a task" do
  task = Task.create!
  get :show, id: task.id
  expect(response).to have_http_status(:success)
  expect(assigns(:task).id).to eq(task.id)
  expect(session[:previous_page]).to eq("task/show")
end

I'm aggressive about moving controller logic that interacts with the model to some kind of action object that doesn't have Rails dependencies. The controller logic and controller testing then tends to be limited to correctly dispatching successful and failed actions. That said, many Rails developers, notably including David Heinemeier Hansson, find adding an extra layer of objects to be overkill and think that worry about slow tests is misplaced. I recommend you try both ways and see which one best suits you.

Integration testing with Capybara

In a Rails context, the following are fodder for integration tests:

  • The interaction between a controller and the model or other objects that provide data
  • The interaction between multiple controller actions that comprise a common work flow.
  • Certain security issues that involve the interaction between a user state and a particular controller action.

These things, generally speaking, are not integration tests. Use unit tests instead:

  • Special cases of business logic, such as what happens if data is nil or has an unexpected value
  • Error cases, unless an error case genuinely results in a unique user experience
  • Internal implementation details of business logic

OMG THIS BOOK IS SO BORING.

Effective Ruby

Item 1: Understand What Ruby Considers to Be True

EVERY value is true except false and nil.

If you need to differentiate between false and nil, either use the nil? method or use the “==” operator with false as the left operand. If you use x on the left then there's a possiblity that x.== was messed with and you won't get what you expect.

if false == x
  something
end

Item 2: Treat All Objects as If They Could Be nil

There’s a surprisingly large number of ways nil can unexpectedly get introduced into your running program. The best defense is to assume that any object might actually be the nil object. This includes arguments passed to methods and return values from them.

When appropriate, use conversion methods such as to_s and to_i to coerce nil objects into the expected type. Make methods like the following more robust by using to_s instead of resorting to throwing "undefined method" exception

def fix_title (title)
  title.to_s.capitalize
end

def add_ten(x)
  x.to_i + 10
end

use compact() to remove nils from arrays

name = [first, middle, last].compact.join(" ")

Item 3: Avoid Ruby’s Cryptic Perlisms

Avoid using ~= and $1, $2, use match() instead

if m = 'ERROR: bad stuff'.match(/^ERROR:\s+(.+)$/)
  m[1] # 'bad stuff'
end

Note: match() also works in reverse (its both a method of string and Regexp)

/find (me)/.match('find me')[1]  #=> 'me'
'find me'.match(/find (me)/)[1]  #=> 'me'

use $LOAD_PATH instead of $:

For other cryptic perlism

require('English')

Look them up

$ ri English

Avoid methods that implicitly read from, or write to, the $_ global variable (I didn't even know it existed outside perl, so no problem)

Item 4: Be Aware That Constants Are Mutable

  • Always freeze constants to prevent them from being mutated.

  • If a constant references a collection object such as an array or hash, freeze the collection and its elements.

  • To prevent assignment of new values to existing constants, freeze the module they’re defined in.

Item 5: Pay Attention to Run-Time Warnings

  • Use the “-w” command-line option to the Ruby interpreter to enable compile-time and run-time warnings. You can also set the RUBYOPT environment variable to “-w”.

  • If you must disable run-time warnings, do so by temporarily setting the $VERBOSE global variable to nil.

Item 6: Know How Ruby Builds Inheritance Hierarchies

  • To find a method, Ruby only has to search up the class hierarchy. If it doesn’t find the method it’s looking for it starts the search again, trying to find the method_missing method.

  • Including modules silently creates singleton classes that are inserted into the hierarchy above the including class.

  • Singleton methods (class methods and per-object methods) are stored in singleton classes that are also inserted into the hierarchy.

Item 7: Be Aware of the Different Behaviors of super

  • When you override a method from the inheritance hierarchy the super keyword can be used to call the overridden method.

  • Using super with no arguments and no parentheses is equivalent to passing it all of the arguments that were given to the enclosing method.

  • If you want to use super without passing the overridden method any arguments, you must use empty parentheses, i.e., super().

  • Defining a method_missing method will discard helpful information when a super call fails. See Item 30 for alternatives to method_ missing.

Item 8: Invoke super When Initializing Subclasses

  • Ruby doesn’t automatically call the initialize method in a super- class when creating objects from a subclass. Instead, normal method lookup rules apply to initialize and only the first matching copy is invoked.

  • When writing an initialize method for a class that explicitly uses inheritance, use super to initialize the parent class. The same rule applies when you define an initialize_copy method.

Item 9: Be Alert for Ruby’s Most Vexing Parse

  • Setter methods can only be called with an explicit receiver. Without a

  • receiver they will be parsed as variable assignments.

  • When invoking a setter method from within an instance method use self as the

  • receiver.

  • You don’t need to use an explicit receiver when calling nonsetter methods. In

  • other words, don’t litter your code with self.

Item 10: Prefer Struct to Hash for Structured Data

Struct can take a block for defining methods

Reading = Struct.new(:date, :high, :low) do 
  def mean
    (high + low) / 2.0
  end
end
  • When dealing with structured data which doesn’t quite justify a new class prefer using Struct to Hash.

  • Assign the return value of Struct::new to a constant and treat that constant like a class.

Item 11: Create Namespaces by Nesting Code in Modules

Referring to KEY fails

module SuperDumbCrypto
  KEY = "asf"
end

class SuperDumbCrypto::Encrypt
  def initialize (key=KEY)
  end
end

SuperDumbCrypto::Encrypt.new #=> uninitialized constant SuperDumbCrypto::Encrypt::KEY

Fully qualified referring to KEY works (not lexical because we closed the module first)

module SuperDumbCrypto
  KEY = "asdf"
end

class SuperDumbCrypto::Encrypt
  def initialize (key=SuperDumbCrypto::KEY)
    puts key
  end
end

SuperDumbCrypto::Encrypt.new #=> 'asdf'

Lexically scoped call to KEY works (didn't close module)

module SuperDumbCrypto
  KEY = "asdf"

  class SuperDumbCrypto::Encrypt
    def initialize (key=KEY)
      puts key
    end
  end

end # module

SuperDumbCrypto::Encrypt.new  #=> 'asdf'

Item 12: Understand the Different Flavors of Equality

  • Never override the equal? method. It's expected to strictly compare objects and return true only if they're both pointers to the same object in memory (i.e., they both have the same object_id).

  • The Hash class uses the eql? method to compare objects used as keys during collisions. The default implementation probably doesn't do what you want. Follow the advice in Item 13 and then alias eql? to == and write a sensible hash method. (is this a big deal?)

  • Use the == operator to test if two objects represent the same value. Some classes like those representing numbers have a sloppy equality operator that performs type conversion.

  • case expressions use the === operator to test each when clause. The left operand is the argument given to when and the right operand is the argument given to case.

Item 13: Implement Comparison via <=> and the Comparable Module

  • Implement object ordering by defining a <=> operator and including the Comparable module.

  • The <=> operator should return nil if the left operand can't be compared with the right.

  • If you implement <=> for a class you should consider aliasing eql? to ==, especially if you want instances to be usable as hash keys, in which case you should also override the hash method.

Item 14 Share Private State Through Protected Methods

  • Share private state through protected methods.

  • Protected methods can only be called with an explicit receiver from objects of the same class or when they inherit the protected methods from a common superclass.

Item 15 Prefer Class Instance Variables to Class Variables

@var is the same in both places

class MyClass

  @var = :here_in_open

  def self.instance
    @var = :here_in_instance
  end

end

use Module singleton if needed

require 'singleton'

class Klass
  include Singleton
  # ...
end

a,b  = Klass.instance, Klass.instance
p a == b #=> true
  • Prefer class instance variables to class variables.

  • Classes are objects and so have their own private set of instance variables.

Item 16: Duplicate Collections Passed as Arguments before Mutating Them

Can use Marshal to make a deep copy, but won't work with some stuff like closures, filehandles, etc

irb> a = ["Monkey", "Brains"]
irb> b = Marshal.load(Marshal.dump(a))
  • Method arguments in Ruby are passed as references, not values. Notable exceptions to this rule are Fixnum objects. (my note: I think its actually pass by value and those values are references)

  • Duplicate collections passed as arguments before mutating them.

  • The dup and clone methods only create shallow copies.

  • For most objects, Marshal can be used to create deep copies when needed.

Item 17: Use the Array Method to Convert nil and Scalar Objects into Arrays

class Pizza
  def initialize (toppings)
    Array(toppings).each do |topping| 
      add_and_price_topping(topping)
    end 
  end
# ...
end
  • Use the Array method to convert nil and scalar objects into arrays.

  • Don’t pass a Hash to the Array method; it will get converted into a set of nested arrays.

Item 18: Consider Set for Efficient Element Inclusion Checking

Putting structs into a set example. Use hash on date as the 'hash key' (set uses a object.hash to determine if element to be added is unique).

From Set Docs: The equality of each couple of elements is determined according to Object#eql? and Object#hash, since Set uses Hash as storage.

require('set')

class AnnualWeather

  Reading = Struct.new(:date, :high, :low) do
    def eql? (other) date.eql?(other.date); end
    def hash; date.hash; end # prevents duplication of date
  end

  def initialize () 
    @readings = Set.new
  end

  def add(date, high, low)
    @readings << Reading.new(date, high, low)
  end

end

w = AnnualWeather.new
w.add(2001, 50, 40) #=> added
w.add(2002, 55, 45) #=> added
w.add(2002, 60, 50) # won't add to set because of eql?() and hash() on date
  • Consider Set for efficient element inclusion checking.

  • Objects inserted into a Set must also be usable as hash keys.

  • Require the “set” file before using Set.

Item 19: Know How to Fold Collections with reduce

Examples:

(1..10).inject(0, :+) #=> 55

users.reduce([]) do |names, user| 
  names << user.name if user.age >= 21 names
end 
  • Always use a starting value for the accumulator.

  • The block given to reduce should always return an accumulator. It’s fine to mutate the current accumulator, just remember to return it from the block.

Item 20: Consider Using a Default Hash Value

With a default value to Hash.new() we dont need the ||= line

array.reduce(Hash.new(0)) do |hash, element|
  # hash[element] ||= 0 # Make sure the key exists.
  hash[element] += 1 # Increment the value.
  hash # Return the hash to reduce. 
end

Use of fetch as a default value

h = {}
h[:accum] = h.fetch(:accum, 0) + 1
  • Consider using a default Hash value.

  • Use has_key? or one of its aliases to check if a hash contains a key. That is, don’t assume that accessing a nonexistent key will return nil.

  • Don’t use default values if you need to pass the hash to code that assumes invalid keys return nil.

  • Hash#fetch can sometimes be a safer alternative to default values.

Item 21: Prefer Delegation to Inheriting from Collection Classes

Core classes can be spiteful

irb> class LikeArray < Array; end
irb> x = LikeArray.new([1, 2, 3]) ---> [1, 2, 3]
irb> y = x.reverse ---> [3, 2, 1]
irb> y.class ---> Array doh!
  • Prefer delegation to inheriting from collection classes.

  • Don’t forget to write an initialize_copy method that duplicates the delegation target.

  • Write freeze, taint, and untaint methods that send the corresponding message to the delegation target followed by a call to super.

Item 22: Prefer Custom Exceptions to Raising Strings

raise("coffee machine low on water")

same as

raise(RuntimeError, "coffee machine low on water")

Another reason to inherit from StandardError comes from the default behavior of the rescue clause. As you know, you can omit the class name when handling exceptions with rescue. In this case, it will intercept any exception whose class (or superclass) is StandardError. (The same is true when you use rescue as a statement modifier.)

Custom exception

class TemperatureError < StandardError
  attr_reader(:temperature)
  def initialize (temperature)
    @temperature = temperature
    super("invalid temperature: #@temperature")
  end
end

Avoid raising strings as exceptions; they're converted into generic RuntimeError objects. Create a custom exception class instead.

Custom exception classes should inherit from StandardError and use the Error suffix in their class name.

When creating more than one exception class for a project, start by creating a base class that inherits from StandardError. Other exception classes should inherit from this custom base class.

If you write an initialize method for your custom exception class make sure to call super, preferably with an error message.

When setting an error message in initialize, keep in mind that setting an error message with raise will take precedence over the one in initialize.

Item 23: Rescue the Most Specific Exception Possible

begin
 task.perform
rescue NetworkConnectionError => e
 # Retry logic...
rescue InvalidRecordError => e
 # Send record to support staff...
rescue => e # danger, could e not be original exception?
 service.record(e)
 raise
ensure
 ...
end

Store and reraise current exception if you are going to do something fancy with it so it doesn't get overritten by another execption. Also note a def can have a rescue clause.

def send_to_support_staff (e)
 ...
rescue
 raise(e)
end
  • Rescue only those specific exceptions from which you know how to recover.

  • When rescuing an exception, handle the most specific type first. The higher a class is in the exception hierarchy the lower it should be in your chain of rescue clauses.

  • Avoid rescuing generic exception classes such as StandardError. If you find yourself doing this you should consider whether what you really want is an ensure clause instead.

  • Raising an exception from a rescue clause will replace the current exception and immediately leave the current scope, resuming exception processing.

Item 24: Manage Resources with Blocks and ensure

Example of made up Lock class, not that only calling with block 'unlocks' the lock

class Lock
  def self.acquire

    lock = new # Initialize the resource.
    lock.exclusive_lock!

    if block_given? 
      yield(lock) # pass lock to block
    else
      lock # Act more like Lock::new and return lock handle.
    end

  ensure

    if block_given?
      lock.unlock if lock # Make sure it gets unlocked.
    end

  end 
end

Idealized version of above with just method.

def get_resource

  resource = Resource.get

  if block_given?
    yield(resource)
  end

rescue
  # nothing special
ensure

  if block_given?
    resource.release
  end

end
  • Write an ensure clause to release any acquired resources.

  • Use the block and ensure pattern with a class method to abstract away resource management.

  • Make sure variables are initialized before using them in an ensure clause.

Item 25: Exit ensure Clauses by Flowing Off the End

Item 24 makes the point that using an ensure clause is the best way to manage resources in the presence of exceptions. More generally, ensure can be used to perform any sort of housekeeping before leaving the current scope. It’s a fantastic, general-purpose cleaning product. But like any useful tool, ensure comes with a warning label and some sharp edges.

One of the features that both rescue and ensure share is the ability to terminate exception processing. You already know that rescue catches exceptions. From within rescue you can choose to cancel propagation and deal with the error, resuming normal control flow. You can also restart exception processing by raising the original exception or by creating a new one.

You probably don’t expect that an ensure clause can alter control flow and swallow exceptions. That’s certainly not its primary purpose. Nevertheless, it’s possible, fairly simple, and slightly subtle. ALL IT TAKES IS AN EXPLICIT RETURN STATEMENT:

The following return in ensure will eat exceptions if there are any and will always be the return value for tricky(), which is just dumb code

def tricky
  return 'horses'
ensure
  return 'ponies' 
end

next and break can also silently discard exceptions in interators

items.each do |item|
  begin
    raise TooStrongError if item == 'lilac' 
  ensure
    next # Cancels exception, continues iteration.
  end
end
  • Avoid using explicit return statements from within ensure clauses. This suggests there’s something wrong with the logic in the method body.

  • Similarly, don’t use throw directly inside ensure. You probably meant to place throw in the body of the method.

  • When iterating, never use next or break in an ensure clause. Ask yourself if you actually need the begin block inside the iteration. It might make more sense to invert the relationship, placing the itera- tion inside the begin.

  • More generally, don’t alter control flow in an ensure clause. Do that in a rescue clause instead. Your intent will be much clearer.

Item 26: Bound retry Attempts, Vary Their Frequency, and Keep an Audit Trail

retries = 0
begin
  service.update(record)
rescue VendorDeadlockError => e
  raise if retries >= 3
  retries += 1
  logger.warn("API failure: #{e}, retrying...")
  sleep(5 ** retries)
  retry
end
  • Never use retry unconditionally; treat it like an implicit loop in your code. Create a variable outside the scope of a begin block and re-raise the exception if you’ve hit the upper bound on the retry limit.

  • Maintain an audit trail when using retry. If retrying problematic code doesn’t go well you’ll want to know the chain of events that led up to the final error.

  • When using a delay before a retry, consider increasing it within each rescue to avoid exacerbating the problem.

Item 27: Prefer throw to raise for Jumping Out of Scope

match = catch(:jump) do 
  @characters.each do |character|
    @colors.each do |color|
    if player.valid?(character, color)
      throw(:jump, [character, color]) end
    end 
  end
end

As you can see, catch takes a symbol to use as the label and a block to execute. If throw is used in that block with the same label then catch will terminate the block and return the value given to throw. You don’t have to pass a value to throw either; it’s completely optional. If you omit the value it will be nil, or if throw isn’t called during the execu- tion of the block then catch will return the last value of its block. The only mandatory part of the throw invocation is the label symbol. If the throw label doesn’t match the catch label the stack will unwind look- ing for a matching catch, eventually turning into a NameError excep- tion if one can’t be found.

If you find yourself using exceptions purely for control flow you might want to consider using catch and throw instead. But if you find yourself using catch and throw too often, you’re probably doing something wrong.

  • Prefer using throw to raise when you need complicated control flow. An added bonus with throw is that you can send an object up the stack, which becomes the return value of catch.

  • Use the simplest control structure possible. You can often rewrite a catch and throw combination as a pair of method calls that simply use return to jump out of scope.

Item 28: Familiarize Yourself with Module and Class Hooks

My Note: I'll probably always regret getting fancy with this stuff unless I'm writing something like rails or rspec.

  • All of the hook methods should be defined as singleton methods.

  • The hooks that are called when a method is added, removed, or undefined only receive the name of the method, not the class where the change occurred. Use the value of self if you need to know this.

  • Defining a singleton_method_added hook will trigger itself.

  • Don’t override the extend_object, append_features, or prepend_features methods. Use the extended, included, or prepended hooks instead.

Item 29: Invoke super from within Class Hooks

Keep in mind that since modules can insert class hooks, it’s not always obvious when the hook you’re writing might override another one higher up in the inheritance hierarchy. Using super is good way to future-proof your code, but ultimately, you’ll have to use your best judgment.

Item 30: Prefer define_method to method_missing

Method missing screws up duct-typing and makes objects obscure.

Following example: rather than inheriting from Hash, delegates to an internal hash instead (probably should use Forwardable)

class HashProxy
  Hash.public_instance_methods(false).each do |name|
    define_method(name) do |*args, &block|
      @hash.send(name, *args, &block)
    end
  end
  def initialize
    @hash = {}
  end
end
  • Prefer define_method to method_missing.

  • If you absolutely must use method_missing consider defining respond_to_missing?.

Item 31: Know the difference between the Variants of eval

  • Methods defined using instance_eval or instance_exec are singleton methods.

  • The class_eval, module_eval, class_exec, and module_exec methods can only be used with classes and modules. Methods defined with one of these become instance methods.

Item 32: Consider Alternatives to Monkey Patching

  • While refinements might not be experimental anymore, they're still subject to change as the feature matures.

  • A refinement must be activated in each lexical scope in which you want to use it.

Item 33: Invoking modified methods with alias chaining

Add logging to any method

module LogMethod

  def log_method (method)

    # Choose a new, unique name for the method.
    orig = "#{method}_without_logging".to_sym

    # Make sure name is unique.
    if instance_methods.include?(orig)
      raise(NameError, "#{orig} isn't a unique name")
    end

    # Create a new name for the original method.
    alias_method(orig, method)

    # Replace original method.
    define_method(method) do |*args, &block|
      $stdout.puts("calling method '#{method}'")
      result = send(orig, *args, &block)
      $stdout.puts("'#{method}' returned #{result.inspect}")
      result
    end

  end

end

irb> Array.extend(LogMethod)
irb> Array.log_method(:first)

irb> [1, 2, 3].first
calling method 'first'
'first' returned 1
---> 1

irb> %w(a b c).first_without_logging
---> "a"
  • When setting up an alias chain, make sure the aliased named is unique.

  • Consider providing a method that can undo the alias chaining.

Item 34: Consider supporting differences in Proc Arity

  • Unlike weak Proc objects, their strong counterparts will raise an ArgumentError exception if called with the wrong number of arguments.

  • You can use the Proc#arity method to find out how many arguments a Proc object expects. A positive number means it expects that exact number of arguments. A negative number, on the other hand, means there are optional arguments and it is the ones’ com- plement of the number of required arguments.

Item 35: Think Carefully Before Using Module Prepending

module MyModule
  def say
    print " MyModule 1 "
    super
  end
end

class Parent
  def say
    print " Parent "
  end
end

class Child < Parent
  # include MyModule # <-----  include
  prepend MyModule # <-----  prepend
  def say
    print " Child "
    super
    puts
  end
end

# include MyModule
Child.new.say #=> " Child  MyModule  Parent "

# prepend MyModule, runs say() in MyModule first
Child.new.say #=> " MyModule  Child  Parent "
  • Using the prepend method inserts a module before the receiver in the class hierarchy, which is much different than include, which inserts a module between the receiver and its superclass.

  • Similar to the included and extended module hooks, prepending a module triggers the prepended hook.

Item 39: Strive for Effectively Tested Code

  • Use fuzzing and property-based testing tools to help exercise both the happy and exception paths of your code.

  • Test-code coverage can give you a false sense of security since exe- cuted code isn’t necessarily correct code.

  • It’s much easier to test a feature while you’re writing it.

  • Before you start to search for the root cause of a bug, write a test that fails because of it.

  • Automate your tests as much as possible.

Item 41: Be Aware of IRB's Advanced Features

  • Define custom IRB commands in the IRB::ExtendCommandBundle module or a module that is then included into IRB::ExtendCommand Bundle.

  • Use the underscore _ variable to access the result of the last expression.

  • The irb command can be used to start a new session and change the current evaluation context to an arbitrary object.

  • Consider the popular Pry gem as an alternative to IRB.

Item 42: Manage Gem Dependencies with Bundler

  • In exchange for a little bit of flexibility you can load all of the gems specified in your Gemfile by using Bundler.require after loading Bundler.

  • When developing an application, list your gems in the Gemfile and add the Gemfile.lock file to your version-control system

  • When developing a RubyGem, list your gem dependencies in the gem-specification file and do not include the Gemfile.lock file in your version-control system.

Item 43: Specify an Upper Bounds for Gem Dependencies

gem('money', '>= 5.1.0', '< 5.2.0')
gem('money', '~> 5.1.0') # same (somehow)

The pessimistic version operator creates a range of version numbers by manipulating the version string to its right. In this example, the lower bound of the range turns into “>= 5.1.0”. The upper bound is created in two steps. First, the rightmost digit is removed from the version string, so 5.1.0 becomes 5.1. Next, the new rightmost digit is incremented, changing 5.1 into 5.2. The resulting number becomes the upper bound: “< 5.2.0”.

  • Omitting an upper bound on a version requirement is akin to say- ing that your application or library supports all future versions of a dependency.

  • Prefer an explicit range of version numbers over the pessimistic ver- sion operator.

  • When releasing a gem to the public, specify the dependency version requirement as wide as you safely can with an upper bound that extends until the next potentially breaking release.

RSpec Docs

Basic Structure

describe creates an ExampleGroup

RSpec.describe Object do
end

Alais for describe is context

RSpec.context Object do
end

ExampleGroups can be nested with describe or context (the omit the RSpec receiver except in the outer example group)

RSpec.describe Calculator do

  describe :another_describe do
    example 'example test' do
      expect(2).to eq(2)  
    end
  end

  context :another_context do
    example 'example test' do
      expect(2).to eq(2)  
    end
  end

end

You can skip an entire example group by putting an 'x' before describe or context

RSpec.describe Calculator do

  xdescribe :another_describe do
    example 'example test' do
      expect(2).to eq(2)  
    end
  end

  xcontext :another_context do
    example 'example test' do
      expect(2).to eq(2)  
    end
  end

end

Examples are placed inside ExampleGroup

RSpec.describe Object do

  it 'does something' do
  end

end

Alternatives examples to 'it' or skip

  specify 'something' do
  end

  it 'does something' do
  end 

  example 'does something' do
  end

Mark examples as skip with 'x'

xit 'something' do
end

xexample 'something' do
end

xspecify 'something' do
end

Pending tests are expected to fail, if they don't fail its a failed test

# successful pending test
example 'example test' do
  expect(2).to eq(2) #<-- good  
  pending #<-- use before failing test
  expect(2).to eq(1) #<-- bad 
end

# failing pending test
pending 'example test' do
  expect(2).to eq(2)  
end

Why are skipped tests displayed as 'pending', exactly same as pending tests???

Calculator
  example test (PENDING: No reason given)

Build in Matchers

My assumed MVP matchers
expect(actual).to eq(expected)  # passes if actual == expected
expect(actual).to match(/expression/)
expect(actual).to be_truthy   # passes if actual is truthy (not nil or false)
expect(actual).to be true     # passes if actual == true
expect(actual).to be_falsy    # passes if actual is falsy (nil or false)
expect(actual).to be_nil      # passes if actual is nil
expect { ... }.to raise_error(ErrorClass)

expect(actual).to be_xxx         # passes if actual.xxx?
expect(obj).to be_completed # if obj.completed?

expect(actual).to have_xxx(:arg) # passes if actual.has_xxx?(:arg)
expect(obj).to have_member(:one) # passes if obj.has_member?(:one)

expect(actual).to include(expected)
Equivalence
expect(actual).to eq(expected)  # passes if actual == expected
expect(actual).to eql(expected) # passes if actual.eql?(expected)
expect(actual).not_to eql(not_expected) # passes if not(actual.eql?(expected))

Note: The new expect syntax no longer supports the == matcher.

Identity
expect(actual).to be(expected)    # passes if actual.equal?(expected)
expect(actual).to equal(expected) # passes if actual.equal?(expected)
Comparisons
expect(actual).to be >  expected
expect(actual).to be >= expected
expect(actual).to be <= expected
expect(actual).to be <  expected
expect(actual).to be_within(delta).of(expected)
Regular expressions
expect(actual).to match(/expression/)
Types/Classes
expect(actual).to be_an_instance_of(expected) # passes if actual.class == expected
expect(actual).to be_a(expected)              # passes if actual.is_a?(expected)
expect(actual).to be_an(expected)             # an alias for be_a
expect(actual).to be_a_kind_of(expected)      # another alias
Thruthiness
expect(actual).to be_truthy   # passes if actual is truthy (not nil or false)
expect(actual).to be true     # passes if actual == true
expect(actual).to be_falsy    # passes if actual is falsy (nil or false)
expect(actual).to be false    # passes if actual == false
expect(actual).to be_nil      # passes if actual is nil
expect(actual).to_not be_nil  # passes if actual is not nil
Expecting errors (notice the blocks, no arguments)
expect { ... }.to raise_error
expect { ... }.to raise_error(ErrorClass)
expect { ... }.to raise_error("message")
expect { ... }.to raise_error(ErrorClass, "message")
Expecting throws (notice the blocks, no arguments)
expect { ... }.to throw_symbol
expect { ... }.to throw_symbol(:symbol)
expect { ... }.to throw_symbol(:symbol, 'value')
Yielding
expect { |b| 5.tap(&b) }.to yield_control # passes regardless of yielded args

expect { |b| yield_if_true(true, &b) }.to yield_with_no_args # passes only if no args are yielded

expect { |b| 5.tap(&b) }.to yield_with_args(5)
expect { |b| 5.tap(&b) }.to yield_with_args(Fixnum)
expect { |b| "a string".tap(&b) }.to yield_with_args(/str/)

expect { |b| [1, 2, 3].each(&b) }.to yield_successive_args(1, 2, 3)
expect { |b| { :a => 1, :b => 2 }.each(&b) }.to yield_successive_args([:a, 1], [:b, 2])
Predicate matchers
expect(actual).to be_xxx         # passes if actual.xxx?
expect(actual).to have_xxx(:arg) # passes if actual.has_xxx?(:arg)
Ranges (Ruby >= 1.9 only)
expect(1..10).to cover(3)
Collection membership
expect(actual).to include(expected)
expect(actual).to start_with(expected)
expect(actual).to end_with(expected)

expect(actual).to contain_exactly(individual, items)
# ...which is the same as:
expect(actual).to match_array(expected_array)
Should syntax (old deprecated style)
actual.should eq expected
actual.should be > 3
[1, 2, 3].should_not include 4
Compound Matcher Expressions
expect(alphabet).to start_with("a").and end_with("z")
expect(stoplight.color).to eq("red").or eq("green").or eq("yellow")

ActiveRecord Official Docs

Basics (TODO)

http://guides.rubyonrails.org/active_record_basics.html

Migrations (TODO)

http://guides.rubyonrails.org/active_record_migrations.html

Validations (TODO)

http://guides.rubyonrails.org/active_record_validations.html

Callbacks (TODO)

http://guides.rubyonrails.org/active_record_callbacks.html

Associations (TODO)

http://guides.rubyonrails.org/association_basics.html

Querying (TODO)

http://guides.rubyonrails.org/active_record_querying.html

FactoryGirl Docs

http://www.rubydoc.info/gems/factory_girl/file/GETTING_STARTED.md

Design Patterns In Ruby

It’s commonly agreed that the most useful thing about patterns is the way in which they form a vocabulary for articulating design decisions during the normal course of development conversations among programmers.

So much for the principles originally cited by the GoF in 1995. To this formidable set, I would like to add one more principle that I think is critical to building, and actually finishing, real systems. This design principle comes out of the Extreme Programming world and is elegantly summed up by the phrase You Ain’t Gonna Need It (YAGNI for short). The YAGNI principle says simply that you should not implement features, or design in flexibility, that you don’t need right now. Why? Because chances are, you ain’t gonna need it later, either.

I think this is everthing like RoR, EventMachine, or a lot of other frameworks.

The Template Method pattern is simply a fancy way of saying that if you want to vary an algorithm, one way to do so is to code the invariant part in a base class and to encapsulate the variable parts in methods that are defined by a number of subclasses. The base class can simply leave the methods completely undefined—in that case, the subclasses must supply the methods. Alternatively, the base class can provide a default implementation for the methods that the subclasses can override if they want.

Also known as the policy pattern

  • defines a family of algorithms
  • encapsulate each algorithms
  • makes the algorithms interchangeable within that family

With the Template Method pattern, we make our decision when we pick our concrete subclass. In the Strategy pattern, we make our decision by selecting a strategy class at runtime.

Simple strategy pattern can be implemented with a passed in code block.

general.execute do |specific|
  ...
  specific.execute
  ...
end

The easiest way to go wrong with the Strategy pattern is to get the interface between the context and the strategy object wrong. Bear in mind that you are trying to tease an entire, consistent, and more or less self-contained job out of the context object and delegate it to the strategy. You need to pay particular attention to the details of the interface between the context and the strategy as well as to the coupling between them. Remember, the Strategy pattern will do you little good if you couple the con- text and your first strategy so tightly together that you cannot wedge a second or a third strategy into the design.

The Strategy pattern is a delegation-based approach to solving the same problem as the Template Method pattern. Instead of teasing out the variable parts of your algo- rithm and pushing them down into subclasses, you simply implement each version of your algorithm as a separate object. You can then vary the algorithm by supplying dif- ferent strategy objects to the context

We have a couple of choices regarding how we get the appropriate data from the context object over to the strategy object. We can pass all of the data as parameters as we call methods on the strategy object, or we can simply pass a reference to the whole context object to the strategy. (dont forget code blocks)

The motive behind the Strategy pattern is to supply the context with an object that knows how to perform some variation on an algorithm.

The observer pattern is a software design pattern in which an object, called the subject, maintains a list of its dependents, called observers, and notifies them automatically of any state changes, usually by calling one of their methods.

Observable in ruby stdlib

Code block as an observer

fred = Employee.new('Fred', 'Crane Operator', 30000)
fred.add_observer do |changed_employee|
  puts(“Cut a new check for #{changed_employee.name}!”) 
  puts(“His salary is now #{changed_employee.salary}!”)
end

The key decisions that you need to make when implementing the Observer pattern all center on the interface between the subject and the observer. At the simple end of the spectrum, you might do what we did in the example above: Just have a single method in the observer whose only argument is the subject. The GoF term for this strategy is the pull method, because the observers have to pull whatever details about the change that they need out of the subject. The other possibility—logically enough termed the push method—has the subject send the observers a lot of details about the change.

Example in AR (this might be a little too magical for me)

class EmployeeObserver < ActiveRecord::Observer
  def after_create(employee)
    # New employee record created
  end
  def after_update(employee)
    # Employee record updated
  end
  def after_destroy(employee)
    # Employee record deleted
  end 
end

In a nice example of the Convention Over Configuration pattern, ActiveRecord does not require you to register your observer: It just figures out that EmployeeObserver is there to observe Employees, based on the class name.

The Observer pattern allows you to build components that know about the activities of other components without having to tightly couple everything together in an unmanageable mess of code-flavored spaghetti. By creating a clean interface between the source of the news (the observable object) and the consumer of that news (the observers), the Observer pattern moves the news without tangling things up.

The interface between observer and observable can be a complex as you like, but if you are building a simple observer, code blocks work well.

Composite should be used when clients ignore the difference between compositions of objects and individual objects. If programmers find that they are using multiple objects in the same way, and often have nearly identical code to handle each of them, then composite is a good choice; it is less complex in this situation to treat primitives and composites as homogeneous.

Three parts

  1. component - Common interface or base class for all of your objects.
  2. leaf - Implements all Component methods
  3. composit - The composite is a component, but it is also a higher-level object that is built from subcomponents.

...there is one unavoidable difference between a composite and a leaf: The composite has to manage its children, which probably means that it needs to have a method to get at the children and possibly methods to add and remove child objects. The leaf classes, of course, really do not have any children to manage; that is the nature of leafyness.

Problem: Application needs to manipulate a hierarchical collection of "primitive" and "composite" objects. Processing of a primitive object is handled one way, and processing of a composite object is handled differently. Having to query the "type" of each object before attempting to process it is not desirable.

Provide a way to access the elements of an aggregate object sequentially without exposing its underlying representation

In other words, an Iterator provides the outside world with a sort of movable pointer into the objects stored inside an otherwise opaque aggregate object.

To mix in Enumerable, you need only make sure that your internal iterator method is named each and that the individual elements that you are going to iterate over have a rea- sonable implementation of the <=> comparison operator.

class Account
  def <=>(other)
    balance <=> other.balance
  end
  ...
end

class Portfolio
  include Enumerable
  def each(&block)
    @accounts.each(&block)
  end
  ...
end

Gang of four says:

Commands are an object-oriented replacement for callbacks.

What more is there to say?

Modify existing object dynamically to avoid an adapter (not sure why its here)

bto = BritishTextObject.new('hello', 50.8, :blue)
class << bto
  def color
    colour
  end
end

same as but doesn't look as cool

bto = BritishTextObject.new('hello', 50.8, :blue)
def bto.color
  colour
end

An adapter is an adapter only if you are stuck with objects that have the wrong interface and you are trying to keep the pain of dealing with these ill-fitting interfaces from spreading throughout your system.

#####Proxy Pattern

A proxy, in its most general form, is a class functioning as an interface to something else. The proxy could interface to anything: a network connection, a large object in memory, a file, or some other resource that is expensive or impossible to duplicate.

The Proxy pattern is essentially built around a little white lie. When the client asks us for an object - perhaps the bank account object mentioned earlier - we do indeed give the client back an object. However, the object that we give back is not quite the object that the client expected. What we hand to the client is an object that looks and acts like the object the client expected, but is actually an imposter.

  • Protection Proxy - proxy that controls access to an object
  • Remote Proxy - local access to remote service/object
  • Virtual Proxy - It pretends to be the real object, but it does not even have a reference to the real object until the client code calls a method

Use method_missing for easy proxies:

class AccountProxy

  def initialize(real_account)
    @subject = real_account
  end

  def method_missing(name, *args)
    puts("Delegating #{name} message to subject.")
    @subject.send(name, *args)
  end

end

a.k.a Wrapper

A design pattern that allows behavior to be added to an individual object, either statically or dynamically, without affecting the behavior of other objects from the same class (only objects?)

writer = CheckSummingWriter.new(TimeStampingWriter.new( NumberingWriter.new(SimpleWriter.new('final.txt'))))

Make writing wrappers easier for all the simple pass through methods by using Forwardable.

require 'forwardable'

class WriterDecorator

  extend Forwardable

  def_delegators :@real_writer, :write_line, :rewind, :pos, :close

  def initialize(real_writer)
    @real_writer = real_writer
  end

end

The forwardable module is more of a precision weapon than the method_missing technique. With forwardable, you have control over which methods you delegate. Although you could certainly put logic in method_missing to pick and choose which methods to delegate, the method_missing technique really shines when you want to delegate large numbers of calls.

Wrapping methods

w = SimpleWriter.new('out')

class << w #<-- use this instead of def w.write_line
  alias old_write_line write_line #<-- because of this
  def write_line(line)
    old_write_line("#{Time.new}: #{line}")
  end
end

Just use the module

require 'singleton'

class SimpleLogger
  include Singleton
  # Lots of code deleted...
end

SimpleLogger.instance

Properly applied, singletons are not global variables. Rather, they are meant to model things that occur exactly once. Yes, because it occurs only once, you can use a singleton as a unique communications conduit between bits of your program. But don't do that.

Isn't every class and module a singleton? Its an actual object, and there's only one.

Factory method pattern is a creational pattern which uses factory methods to deal with the problem of creating objects without specifying the exact class of object that will be created.

Does "creating objects without specifying the exact class of object" mean that every factory is capable of creating multiple types of objects, or that the factory name doesn't specify the class type of the object returned?

The key thing that we discovered in this chapter is how both of these patterns morphed in Ruby’s dynamic environment—specifically, how they became much sim- pler. While the GoF concentrated on inheritance-based implementations of their factories, we can get the same results with much less code by taking advantage of the fact that in Ruby, classes are just objects. In Ruby we can look up classes by name, pass them around, and store them away for future use.

The builder pattern is an object creation software design pattern. Unlike the abstract factory pattern and the factory method pattern whose intention is to enable polymorphism, the intention of the builder pattern is to find a solution to the telescoping constructor anti-pattern. The telescoping constructor anti-pattern occurs when the increase of object constructor parameter combination leads to an exponential list of constructors. Instead of using numerous constructors, the builder pattern uses another object, a builder, that receives each initialization parameter step by step and then returns the resulting constructed object at once.

Builder pattern is like a wind-up constructor. I'm not sure how the following code solves the "telescoping constructor anti-pattern" mentioned above. I'm not sure how the following code solves anything at all. Why not just allow the Computer class itself to build its object step by step?

class Computer

  def initialize(drive_size, cd, memory)
    @drive_size, @cd, @memory = drive_size, cd, memory
  end

end

class ComputerBuilder

  attr_accessor :drive_size, :cd, :memory

  def build
    Computer.new(self.drive_size, self.cd, self.memory)
  end

end

builder = ComputerBuilder.new
builder.drive_size = 700
builder.cd = false
builder.memory = 1000
computer = builder.build

Good rationale for Builder Pattern from wikipedia

The builder pattern has another benefit. It can be used for objects that contain flat data (html code, SQL query, X.509 certificate...), that is to say, data that can't be easily edited. This type of data cannot be edited step by step and must be edited at once. The best way to construct such an object is to use a builder class.

The idea behind the Builder pattern is that if your object is hard to build, if you have to write a lot of code to configure each object, then you should factor all of that creation code into a separate class, the builder.

Builders, because they are in control of configuring your object, can also prevent you from constructing an invalid object.

The Interpreter pattern is built around a very simple idea: Some programming problems are best solved by creating a specialized language and expressing the solution in that language.

Another clue that your problem might be right for the Interpreter pattern is that you find yourself creating lots of discrete chunks of code, chunks that seem easy enough to write in themselves, but which you find yourself combining in an ever expanding array of combinations. Perhaps a simple interpreter could do all of the combining work for you.

Interpreters typically work in two phases. First, the parser reads in the program text and produces a data structure, called an abstract syntax tree (AST).

I'm skipping most of this because I think ruby DSL's are a better option for the rare occasion that the Interpreter Pattern would be applicable. And this stuff is so challanging that I'll forget it by tomorrow anyway.

Domain Specific Languages Chapter

Mostly skipping for now, maybe go back to it later. I've delt with DSL often.

Meta-programming Chapter

Skipping. Will do more with meta-programming with more indepth sources.

Convention Over Configuration Chapter

I'm a rails developer. This stuff looks too familiar.

Metaprogramming Ruby

  • An object is composed of a bunch of instance variables and a link to a class.

  • The methods of an object live in the object’s class. (From the point of view of the class, they’re called instance methods.)

  • The class itself is just an object of class Class. The name of the class is just a constant.

  • Class is a subclass of Module. A module is basically a package of methods. In addition to that, a class can also be instantiated (with new) or arranged in a hierarchy (through its superclass).

  • Constants are arranged in a tree similar to a file system, where the names of modules and classes play the part of directories and regular constants play the part of files.

  • Each class has an ancestors chain, beginning with the class itself and going up to BasicObject.

  • When you call a method, Ruby goes right into the class of the receiver and then up the ancestors chain, until it either finds the method or reaches the end of the chain.

  • When you include a module in a class, the module is inserted in the ancestors chain right above the class itself. When you prepend the module, it is inserted in the ancestors chain right below the class.

  • When you call a method, the receiver takes the role of self.

  • When you’re defining a module (or a class), the module takes the role of self.

  • Instance variables are always assumed to be instance variables of self.

  • Any method called without an explicit receiver is assumed to be a method of self.

  • Refinements are like pieces of code patched right over a class, and they override normal method lookup. On the other hand, a Refinement works in a limited area of the program: the lines of code between the call to using and the end of the file, or the end of the module definition.

You can use public_send in place of send if you want to be careful.

There is one important reason to use define_method over the more familiar def keyword: define_method allows you to decide the name of the defined method at runtime.

define_method is private so you cant define method on another class unless you use the send() trick. Same with method_missing.

Dropping this book for now. Getting fatigued by the plot driven slow motion explainations, probably because I'm already familiar with the topic, so new-to-me material is spread extra thin because of the plotting... might go back to it later.

Pickaxe Part I Ruby.new

skimming...

Default value of Hash

histogram = Hash.new(0) # The default value is zero

Don't forget there's a sub() along with that gsub()

newline = line.sub(/Perl/, 'Ruby') # replace first 'Perl' with 'Ruby'
newerline = newline.gsub(/Python/, 'Ruby') # replace every 'Python' with 'Ruby'

You have many ways to read input into your program. Probably the most traditional is to use the method gets, which returns the next line from your program's standard input stream:

line = gets
print line

gets returns nil when done, so you can use this

while line = gets
  print line
end

ARGV contains each of the arguments passed to the running program

p ARGV #=> [arg1, arg2, arg3]
p ARGV[0] #=> arg1
p ARGV.length #=> 3

Don't forget slicing an array

a = [ 1, 3, 5, 7, 9 ] 
a[1,3] #=>[3,5,7] 
a[3, 1] # => [7] 
a[-3,2] #=>[5,7]

Same as string, but doesn't accept regex

s = '12345'
s[0,2] #=> '12'
s[/../] #=> '12' #<-- won't work with an array (don't know how it could)

Splat seems to work, makes sense, Array.[] is just a method

[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10][ *[0,3] ]

Select non-adjacent values in array with value_at

[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10].values_at(1,3,5) #=> [1, 3, 5]

Ruby version of perl's splice - If the index to [ ]= is two numbers (a start and a length) or a range, then those elements in the original array are replaced by whatever is on the right side of the assignment. If the length is zero, the right side is inserted into the array before the start position; no elements are removed.

a = [0, 1, 2, 3, 4, 5]
a[2,4] = :cat
a # => [0, 1, :cat] 

Word counter one liner I just made up

"one two two three three three".scan(/[\w']+/).inject(Hash.new(0)){|a,v| a[v] +=1; a }

Block local variables (does anybody care?, maybe?, be more confident?)

square = "some shape"
sum = 0
[1, 2, 3, 4].each do |value; square| #<-- can't touch outer square now
  square = value * value
  sum += square
end

Add a buncha numbers together

[1,3,5,7].inject(:+) # => 16

Enumerators—External Iterators

It’s also worth spending another paragraph looking at why Ruby’s internal iterators aren’t always the best solution. One area where they fall down badly is where you need to treat an iterator as an object in its own right (for example, passing the iterator into a method that needs to access each of the values returned by that iterator). It’s also difficult to iterate over two collections in parallel using Ruby’s internal iterator scheme. (When would anybody ever want to pass an external iterator to a method?)

You can create an Enumerator object by calling the to_enum method (or its synonym, enum_for) on a collection such as an array or a hash:

a = [ 1, 3, "cat" ]
h = { dog: "canine", fox: "vulpine" }

# Create Enumerators
enum_a = a.to_enum
enum_h = h.to_enum
enum_a.peek # => 1
enum_a.entries # => [ 1, 3, "cat" ]
enum_a.next # => 1
enum_h.next # => [:dog, "canine"] enum_a.next # => 3
enum_h.next # => [:fox, "vulpine"]

There's also peek, enteries

Most of the internal iterator methods—the ones that normally yield successive values to a block—will also return an Enumerator object if called without a block:

e = [1,2,3].each_with_index
e.next #=> [1, 0]

Ruby has a method called loop that does nothing but repeatedly invoke its block. Typically, your code in the block will break out of the loop when some condition occurs. But loop is also smart when you use an Enumerator—when an enumerator object runs out of values inside a loop, the loop will terminate cleanly. The following example shows this in action—the loop ends when the three-element enumerator runs out of values. You can also handle this in your own iterator methods by rescuing the StopIteration exception, but because we haven’t talked about exceptions yet, we won’t go into details here.

short_enum = [1, 2, 3].to_enum 
long_enum = ('a'..'z').to_enum
loop do
  puts "#{short_enum.next} - #{long_enum.next}"
end

Note from http://www.ruby-doc.org/core-2.2.0/Enumerator.html Chaining enumerators lets you chain them together to create new possiblities.

Enumerator mixes in Enumerable, and has these methods

each
each_with_index
with_index # shorter version
next
peek
rewind
each_with_object
with_object

Example

%w[foo bar baz].map.with_index { |w, i| "#{i}:#{w}" }

Bocks for Transactions

# read the counter using read lock
File.open("counter", "r") {|f|
  f.flock(File::LOCK_SH)
  p f.read
}

Blocks Can Be Objects

Here’s an example where we create a Proc object in one instance method and store it in an instance variable. We then invoke the proc from a second instance method.

class ProcExample
  def pass_in_block(&action)
    @stored_proc = action
  end
  def use_proc(parameter)
    @stored_proc.call(parameter)
  end 
end

eg = ProcExample.new
eg.pass_in_block { |param| puts "The parameter is #{param}" } 
eg.use_proc(99)

Blocks can be closures

def power_proc_generator
  value = 1
  lambda { value += value }
end

power_proc = power_proc_generator

puts power_proc.call #=> 2
puts power_proc.call #=> 4
puts power_proc.call #=> 8

An Alternative Notation for lambda

lambda { |params| ... }

-> params { ... }
proc1 = -> arg { puts "In proc1 with #{arg}" }
proc2 = -> arg1, arg2 { puts "In proc2 with #{arg1} and #{arg2}" }
proc3 = ->(arg1, arg2) { puts "In proc3 with #{arg1} and #{arg2}" }

def my_while(cond, &body)
  while cond.call
    body.call
  end 
end

a = 0

my_while ->{ a<3} do
  puts a
  a += 1
end

Oh shit, procs can take blocks

proc1 = lambda do |a, *b, &block|
  puts "a = #{a.inspect}"
  puts "b = #{b.inspect}" 
  block.call
end

proc1.call(1, 2, 3, 4) { puts "in block1" }

# new syntax
proc2 = -> a, *b, &block do 
  puts "a = #{a.inspect}"
  puts "b = #{b.inspect}" 
  block.call
end

Sharing Functionality, Inheritance, Modules, and Mixins

The Comparable mixin adds the comparison operators (<, <=, ==, >=, and >), as well as the method between?, to a class. You must define <=> 'method', which will also work allow sort.

class DominantPair

  include Comparable
  attr_accessor :data

  def initialize(letter, number)
    @data = [letter, number]
  end

  def <=>(other)
    @data[1] <=> other.data[1]
  end

  def inspect
    "[#{data[0]},#{data[1]}]"
  end

end

a = DominantPair.new('a', 2)
b = DominantPair.new('b', 3)
c = DominantPair.new('c', 1)
[a,b,c].sort #=> [[c,1], [a,2], [b,3]]
a > c #=> true
b > a #=> true
c < b #=> true

The Enumerable mixin provides collection classes with several traversal and searching methods, and with the ability to sort. The class must provide a method each, which yields successive members of the collection. If Enumerable#max, #min, or #sort is used, the objects in the collection must also implement a meaningful <=> operator, as these methods rely on an ordering between members of the collection.

class DominantPairCollection

  include Enumerable

  def initialize(pairs)
    @pairs = pairs
  end

  def each
    @pairs.each do |pair|
      yield pair
    end
  end

end

a = DominantPair.new('a', 2)
b = DominantPair.new('b', 3)
c = DominantPair.new('c', 1)

collection = DominantPairCollection.new([a,b,c])

collection.each do |pair|
  p pair
end

Do the same thing but with Forwardable

require 'forwardable'
class DominantPairCollection

  include Enumerable

  extend Forwardable
  def_delegators :@pairs, :each

  def initialize(pairs)
    @pairs = pairs
  end

end

a = DominantPair.new('a', 2)
b = DominantPair.new('b', 3)
c = DominantPair.new('c', 1)

collection = DominantPairCollection.new([a,b,c])

collection.each do |pair|
  p pair
end

A mixin’s instance variables can clash with those of the host class or with those of other mixins.

For the most part, mixin modules don’t use instance variables directly—they use accessors to retrieve data from the client object. But if you need to create a mixin that has to have its own state, ensure that the instance variables have unique names to distinguish them from any other mixins in the system (perhaps by using the module’s name as part of the variable name).

In general, a mixin that requires its own state is not a mixin—it should be written as a class.

Liskov Substitution Principle: What this means is that you should be able to substitute an object of a child class wher- ever you use an object of the parent class—the child should honor the parent’s contract.

In the real world, there really aren’t that many true is a relationships. Instead, it’s far more common to have has a or uses a relationships between things. The real world is built using composition, not strict hierarchies.

Finally, we’ll offer a warning for Perl users. Strings that contain just digits are not automatically converted into numbers when used in expressions. Use the Integer() method to convert the strings to integers. Why not to_i ?

3.times { print "X " } 1.upto(5).to_a #=> [1, 2, 3, 4, 5] 99.downto(95).to_a #=> [99, 98, 97, 96, 95] 50.step(80, 5).to_a # => [50, 55, 60, 65, 70, 75, 80]

10.downto(7).with_index.to_a #=> [[10, 0], [9, 1], [8, 2], [7, 3]]

Ranges and enumerators

e = (1..10).to_enum e.next #=> 1 e.next #=> 2 e.next #=> 3

Ranges as Conditions (I don't understand this at all, maybe more to come)

As well as representing sequences, ranges can also be used as conditional expressions. Here, they act as a kind of toggle switch—they turn on when the condition in the first part of the range becomes true, and they turn off when the condition in the second part becomes true. For example, the following code fragment prints sets of lines from standard input, where the first line in each set contains the word start and the last line contains the word end:

while line = gets
  puts line if line =~ /start/ .. line =~ /end/
end

Ranges as intervals

(1..10) === 5 # => true
(1..10) === 15 # => false

case car_age
when 0..0
  puts "Mmm.. new car smell"
when 1..2
  puts "Nice and new"
when 3..9
  puts "Reliable but slightly dinged"
when 10..29
  puts "Clunker"
else
  puts "Vintage gem"
end

Don't forget there's also a !~ along with =~

File.foreach("testfile").with_index do |line, index|
  puts "#{index}: #{line}" if line !~ /on/
end

Unlike sub and gsub, sub! and gsub! return the string only if the pattern was matched. If no match for the pattern is found in the string, they return nil instead. This means it can make sense (depending on your need) to use the ! forms in conditions.

The match operators are defined for both String and Regexp objects.

/abc/.match "abc" #=> MatchData
"abc".match(/abc/) #=> MatchData

Various methods of MatchData

m = 'abcdef'.match(/(cd)/)
m.begin(0) #=> 2
m.captures #=> ["cd"]
m.regexp #=> /(cd)/
m.values_at(0,2,4) #=> ['cd', nil, nil]
m.pre_match #=> "ab"
m.post_match #=> "ef"
m.captures[0] #=> ["cd"]
m[0] #=> ["cd"]

MatchData#[] and MatchData.captures are not the same:

m = 'abcdef'.match(/(cd)/)
m.captures[0] #=> "cd"
m[0] #=> "cd" (part of string that was matched)
m.captures[1] #=> nil
m[1] #=> "cd" #=> first memory
m.captures[2] #=> nil
m[2] #=> nil

m[0] #=> "cd" (part of string that was matched)
m[1] #=> "cd" #=> first memory

Earlier we noted that the sequences \1, \2, and so on, are available in the pattern, standing for the nth group matched so far. The same sequences can be used in the second argument of sub and gsub.

"fred:smith".sub(/(\w+):(\w+)/, '\2, \1') #=> smith, fred
"nercpyitno".gsub(/(.)(.)/, '\2\1') #=> encryption

Good section on regexen, but I'll probably forget it all. I'll remember its here in case I need to reference it.

You won’t get an immediate error if you start a method name with an uppercase letter, but when Ruby sees you calling the method, it might guess that it is a constant, not a method invocation, and as a result it may parse the call incorrectly.

By convention, methods names starting with an uppercase letter are used for type conversion. The Integer method, for example, converts its parameter to an integer.

Default arguments values can reference previous arguments.

def surround(word, pad_width=word.length/2)
  "[" * pad_width + word + "]" * pad_width
end

I know from experience that a argument can be a instance variable

def my_meth(@my_var)
  ...
end

You can use split to indicate a method that doesn't use any arguments but that are perhaps used by a method in a superclass

class Child < Parent
  def do_something(*not_used)
      # our processing
      super 
  end
end

You can put the splat argument anywhere in a method’s parameter list, allowing you to write this:

def split_apart(first, *splat, last)

If you cared only about the first and last parameters, you could define this method using this:

def split_apart(first, *, last)

Private methods may not be called with a receiver, so they must be methods available in the current object.

As of Ruby 1.9, splat arguments can appear anywhere in the parameter list, and you can intermix splat and regular arguments.

Didn't think of it before but a block param is a possible way to interject some variable code into a chain

if operator =~ /^t/
  calc = lambda {|n| n*number }
else
  calc = lambda {|n| n+number }
end
(1..10).collect(&calc).join(", ")

(This is just wrong) If the last argument to a method is preceded by an ampersand, Ruby assumes that it is a Proc object. It removes it from the parameter list, converts the Proc object into a block, and associates it with the method.

Proof:

procinator = Object.new

def procinator.to_proc
  puts "called to_proc"
  Proc.new {|x| x}
end


p (1..3).map(&procinator) 

outputs:
called to_proc
[1, 2, 3]

Ruby keyword arguments in 2.0!

def search(field, genre: nil, duration: 120) 
  p [field, genre, duration ]
end

search(:title)
search(:title, duration: 432)
search(:title, duration: 432, genre: "jazz")

You can collect these extra hash arguments as a hash parameter—just prefix one element of your argument list with two asterisks (a double splat).

def search(field, genre: nil, duration: 120, **rest) 
  p [field, genre, duration, rest ]
end

Expressions

a * b + c # => 5

Same as

(a.*(b)).+(c) # => 5

Define operators like <<

class Appender

  attr_accessor :data

  def initialize
    @data = ''
  end

  def <<(x)
    @data += x
  end

end


appender = Appender.new

appender << 'a'
appender << 'b'
appender << 'c'

p appender.data #=> 'abc'

Note: return self so we can chain

def <<(x)
  @data += x
  self
end

appender << 'a' << 'b' << 'c' << 'd'

How to define [] and []= methods

obj = Object.new

def obj.[](*args)
  p args
end

obj[1,2,3] #=> [1,2,3]


def obj.[]=(*args)
  p args # last arg is value of assignment
end

obj[1,2,3] = 4 #=> [1, 2, 3, 4]

Ruby is authoritarian about its assignment methods. The value of an assignment is always the value of the parameter, the return value of the method is discarded.

class MyClass
  def val=(v)
    @v = v
    44
  end
end

o = MyClass.new
p o.val = 1 #=> 1 (not 44)

Assignment isn't so obvious

a=1,2,3,4   # a=[1,2,3,4] # WTH?  I've been wasting so much time typing brackets
b=[1,2,3,4] # b=[1,2,3,4]

a,b=1,2,3,4    # a=1, b=2 
c,=1,2,3,4     # c=1

# look again
c = 1,2,3,4
p c #=> [1,2,3,4]
c, = 1,2,3,4 #=> value of assignment is [1,2,3,4]
p c #=> 1  but c is only 1

Splat works on rvalues

a, b, c, d, e = *(1..2), 3, *[4, 5] # a=1, b=2, c=3, d=4, e=5

Seems to do something with ranges too

a = *(1..6)
a #=> [1, 2, 3, 4, 5, 6]

Splat can be applied to lvalues too (exactly one - works like a sponge)

a,*b=1,2,3 # a=1, b=[2,3] 
a,*b=1 # a=1, b=[]

*a,b=1,2,3,4         # a=[1,2,3], b=4
c,*d,e=1,2,3,4       # c=1, d=[2,3], e=4
f,*g,h,i,j=1,2,3,4   # f=1, g=[], h=2, i=3, j=4

I always forget this thing

var ||= "default value"

defined? returns useful values too

defined? 1           #=> "expression"
defined? dummy       #=> nil
defined? printf      #=> method
defined? String      #=> 'constant'
defined? $_          #=> 'global-variable'
defined? Math::PI    #=> "constant"
defined? a = 1       #=>  "assignment"
defined? 42.abs      #=> "method"
defined? nil         #=> "nil"

In addition to the boolean operators, Ruby objects support comparison using the methods ==, ===, <=>, =~, eql?, and equal? All but <=> are defined in class Object but are often overridden by descendants to provide appropriate semantics.

Both == and =~ have negated forms, != and !. Ruby first looks for methods called != or !, calling them if found. If not, it will then invoke either == or =~, negating the result.

ruby has a if/then/else control structure? (the then is optional if formatted like below)

if x == 1 then 
  do_1
elsif x == 2 then 
  do_2
else
  do_else
end

then required for this formatting

if artist == "Gillespie" then handle = "Dizzy" 
elsif artist == "Parker" then handle = "Bird" 
else handle = "unknown"
end

The unless statement does support else, but most people seem to agree that it’s clearer to switch to an if statement in these cases.

case also supports 'then formatting'

kind = case year
       when 1850..1889 then "Blues"
       when 1890..1909 then "Ragtime"
       when 1910..1929 then "New Orleans Jazz" 
       when 1930..1939 then "Swing"
       else                 "Jazz"
       end

Don't forget the statment modifier form of while loop

x = 0
x += 1 while x < 100
puts x #=> 100

The ruby do-while construct

print "Hello\n" while false
begin
  print "Goodbye\n"
end while false
  • break: terminates the immediately enclosing loop; control resumes at the statement following the block.

  • redo: repeats the current iteration of the loop from the start but withoutreevaluating the condition or fetching the next element (in an iterator).

  • next: skips to the end of the loop, effectively starting the next iteration:

  • retry: was removed in 1.9

break can return a value

x = 0
y = loop do
  break(x) if x > 50
  x += 1
end

p y #=> 51

Exceptions

raise # <--- reraises current exception
raise "My stupid exception" #<---  raises RuntimeError with message

Raise a particular exception, using caller to automatically remove the current routine from the stack backtrace (useful for modules)

raise InterfaceException, "Keyboard failure", caller #<-- particular exception with stacktrace

Remove more from backtrace

caller[1..-2]

Creating custom exception

class RetryException < RuntimeError
  attr :ok_to_retry
  def initialize(ok_to_retry)
    @ok_to_retry = ok_to_retry
  end
end

Will I ever use catch and throw???

When Ruby encounters a throw, it zips back up the call stack looking for a catch block with a matching symbol. If the throw is called with the optional second parameter, that value is returned as the value of the catch.

x = catch(:mycatch) do
  throw(:mycatch, :hello)
end

p x #=> hello

Diversion with scan, match and captures. Conclusions: Don't use scan for single captures, use match. Use named captures, so much better.

str = 'this is a string right here'

p str.scan(/\w+/) #=> ["this", "is", "a", "string", "right", "here"]
p str.scan(/(\w+)/) #=> [["this"], ["is"], ["a"], ["string"], ["right"], ["here"]] 

str = 'find me in all this stuff'
p str.scan(/me/) #=> ["me"]

# scan version
str = 'first: Larry last:Wall'
p str.scan(/first:\s*(.*)\s*last:\s*(.*)\s*/) #=> [["Larry ", "Wall"]]
p str.scan(/first:\s*(.*)\s*last:\s*(.*)\s*/)[0][0] #=> Larry
p str.scan(/first:\s*(.*)\s*last:\s*(.*)\s*/)[0][1] #=> Wall

# match version
match = str.match(/first:\s*(.*)\s*last:\s*(.*)\s*/)
p match[0]
p match[1] #=> "Larry" # index at 1 because 0 is whole matched string
p match[2] #=> "Wall"
p match.captures #=> ["Larry ", "Wall"], same as equivalent to mtch.to_a[1..-1].
p match.captures[0] #=> "Larry" # lowers index to 0
p match.captures[1] #=> "Wall"

# matched with named captures (So much better!)
match = str.match(/first:\s*(?<first>.*)\s*last:\s*(?<last>.*)\s*/)
p match[:first]
p match[:last]

Use %r{} regex form for HTML

if page =~ %r{<title>(.*?)</title>}m
  puts "Title is #{$1.inspect}"
end

Previous code from the book, this is my improved example

if match = page.match(%r{<title>(?<title>.*?)</title>}m)
  puts "Title is #{match[:title].inspect}"
end

Block comments in ruby

=begin
...
=end

Fibers, Threads, and Processes

Fibers are often used to generate values from infinite sequences on demand. Here’s a fiber that returns successive integers divisible by 2 and not divisible by 3:

twos = Fiber.new do 
  num = 2
  loop do
    Fiber.yield(num) unless num % 3 == 0 
    num += 2
  end 
end
10.times { print twos.resume, " " }

Because fibers are just objects, you can pass them around, store them in variables, and so on.

Fibers can be resumed only in the thread that created them. (unless you require the fiber lib)

A related but more general mechanism is the continuation. A continuation is a way of recording the state of your running program (where it is, the current binding, and so on) and then resuming from that state at some point in the future. You can use continuations to implement coroutines (and other new control structures). Continuations have also been used to store the state of a running web application between requests—a continuation is created when the application sends a response to the browser; then, when the next request arrives from that browser, the continuation is invoked, and the application continues from where it left off. You enable continuations in Ruby by requiring the continuation library.

Running multiple processes

The method Object#system executes the given command in a subprocess; it returns true if the command was found and executed properly. It raises an exception if the command cannot be found. It returns false if the command ran but returned an error. In case of failure, you’ll find the subprocess’s exit code in the global variable $?.

One problem with system is that the command’s output will simply go to the same destination as your program’s output, which may not be what you want. To capture the standard output of a subprocess, you can use the backquote characters, as with date in the previous example. Remember that you may need to use String#chomp to remove the line-ending characters from the result.

Pickaxe Part III Ruby Crystalized

Ruby expressions and statements are terminated at the end of a line unless the parser can determine that the statement is incomplete, such as if the last token on a line is an operator or comma.

Must be why this works

mls = "line one\n" + 
      "line two\n"

And this is a syntax error

mls =   "line one\n"
      + "line two\n"

If Ruby comes across a line anywhere in the source containing just END, with no leading or trailing whitespace, it treats that line as the end of the program—any subsequent lines will not be treated as program code.

However, these lines can be read into the running pro- gram using the global IO object DATA

#!/usr/bin/env ruby

p DATA.readlines.map(&:chomp)

__END__
One
Two
Three

prints

["One", "Two", "Three"]

Every Ruby source file can declare blocks of code to be run as the file is being loaded (the BEGIN blocks) and after the program has finished executing (the END blocks):

BEGIN {

}

END {

}

This I didn't know, you can drop the Q when %Q for a double quotes string

var = 'here'
puts %Q{var = #{var}} #=> "var = here"
puts %{var = #{var}} #=> "var = here"

Array of symbols in Ruby 2 is similar to array of strings

%w{one two three} #=> ["one", "two", "three"]
%i{one two three} #=> [:one, :two, :three]
%I{one two three} #=> [:one, :two, :three]

Unlike their lowercase counterparts, %I, %Q, and %W will preform interpolation:

%I{ one digit#{1+1} three } # => [:one, :digit2, :three]

Regex in sub/gsub using named captures (note that substitution must be in single quotes or every backslash must be escaped)

'this that'.sub(/^(?<first>.+)\s(?<second>.+)$/, '\k<second> \k<first>')
'this that'.sub(/^(?<first>.+)\s(?<second>.+)$/, "\\k<second> \\k<first>")

# without named captures
'this that'.sub(/^(.+)\s(.+)$/, '\2 \1')

Class variables belong to the innermost enclosing class or module. Class variables used at the top level are defined in Object and behave like global variables. In Ruby 1.9, class variables are supposed to be private to the defining class, although as the following example shows, there seems to be some leakage.

class BugVar
  @@var = 99
  def var
    @@var
  end
end

@@var = 100 #=> warning: class variable access from toplevel
p BugVar.new.var #=> 100

Class variables are inherited by children but propagate upward if first defined in a child, this is messy.

Pickaxe recommends avoiding class variables.

I'm going to try to avoid class variables now, maybe just use class instance variables and instance variables.

class BetterVar

  @var = 99 #<-- class instance var

  class << self         #<-- secret sauce
    attr_accessor :var  #<-- secret sauce
  end

  def var
    self.class.var      #<-- secret sauce
  end

  def var=(x)
    self.class.var = x
  end

end

p BetterVar.var #=> 99
bv = BetterVar.new
bv.var=200
p BetterVar.new.var #=> 200
p BetterVar.var #=> 200

run if the current file is the script being run, can put tests here

if __FILE__ == $0 
  # tests...
end

Element reference vs actual method call

var[] = "one"                   var.[ ]=("one")
var[1] = "two"                  var.[ ]=(1, "two")
var["a", /^cat/ ] = "three"     var.[ ]=("a", /^cat/, "three")

If you are writing an [ ]= method that accepts a variable number of indices, it might be con- venient to define it using this:

def []=(*indices, value) 
  # ...
end

Ranges in Boolean Expressions (I don't think anybody uses these, and they would baffle anybody reading my code)

if expr1 .. expr2
while expr1 .. expr2

A range used in a boolean expression acts as a flip-flop. It has two states, set and unset, and is initially unset.

  1. For the three-dot form of a range, if the flip-flop is unset and expr1 is true, the flip-flop becomes set and the the flip-flop returns true.

  2. If the flip-flop is set, it will return true. However, if expr2 is not true, the flip-flop becomes unset.

  3. If the flip-flop is unset, it returns false.

The first step differs for the two-dot form of a range. If the flip-flop is unset and expr1 is true, then Ruby only sets the flip-flop if expr2 is not also true.

I've never ever seen this used!

case Expressions

Ruby has two forms of case statement. The first allows a series of conditions to be evaluated, executing code corresponding to the first condition that is true:

case
when something
...
when something_else
...
when yet_something_else
...
else
...
end

Second form takes a target

case target
when <test against target>
...
when <test against target>
...
else
...
end

then is optional but good for scrunching down the statement

case target when then ... when then ... else ... end

while and until statements can have a "do"?

until something do ... end

while something do ... end

loop, which iterates its associated block, is not a language construct—it is a method in module Kernel.

while and until modifiers: (not sure if these are a good idea, doesn't look loopy enough)

expression while boolean-expression
expression until boolean-expression

break and next may optionally take one or more arguments. If used within a block, the given argument(s) are returned as the value of the yield. If used within a while, until, or for loop, the value given to break is returned as the value of the statement. If break is never called or if it is called with no value, the loop returns nil. (stop abusing the word yield!, its losing all meaning)

var = [1,2,3].each do |x|
  break :hello
end
var #=> hello

var = [1,2,3].each do |x|
  break
end
var #=> nil

confirmed:

def my_method 
  yield
end

var = my_method do 
  break :there
  :here
end

p var #=> :there

Outside a class or module definition, a definition with an unadorned method name is added as a private method to class Object. It may be called in any context without an explicit receiver.

Confirmed:

def my_method
end

p Object.private_instance_methods.grep /my/ #=> [:my_method]

A definition using a method name of the form expr.methodname creates a method associated with the object that is the value of the expression; the method will be callable only by sup- plying the object referenced by the expression as a receiver. This style of definition creates per-object or singleton methods. You’ll find it most often inside class or module definitions, where the expr is either self or the name of the class/module. This effectively creates a class or module method (as opposed to an instance method).

Even works on the default 'main' object

def self.my_method
end

p self.methods.grep /my/ #=> [:my_method]

Method definitions may not contain class or module definitions. They may contain nested instance or singleton method definitions. The internal method is defined when the enclosing method is executed. The internal method does not act as a closure in the context of the nested method—it is self-contained.

nested methods aren't scoped to the outer method - use a proc for that

def my_method
  def my_other_method
    :here
  end
end

my_other_method rescue puts $! #=> undefined local variable or method `my_other_method' for main:Object
my_method
my_other_method rescue puts $! #=> :here

This is an interesting use of nested methods:

def clock
  def clock
    puts :tock
  end
  puts :tick
end

clock #=> tick
clock #=> tock

A method definition may have zero or more regular arguments, zero or more keyword arguments, a optional splat argument, an optional double splat argument, and an optional block argument. Arguments are separated by commas, and the argument list may be enclosed in parentheses.

This is a little confusing, I wouldn't push my luck with this stuff:

def mixed(a, b=50, c=b+10, d)
  [ a, b, c, d ]
  end

p mixed(1, 2)       #=> [1, 50, 60, 2]
p mixed(1, 2, 3)    #=> [1, 2, 12, 3]
p mixed(1, 2, 3, 4) #=> [1, 2, 3, 4]

Splats can be in the middle. This is probably a bad idea

def splat(first, *middle, last) 
  [ first, middle, last ]
end
splat(1, 2)       # => [1, [], 2] 
splat(1, 2, 3)    # => [1, [2], 3] 
splat(1, 2, 3, 4) # => [1, [2, 3], 4]

This might be useful

If an array argument follows arguments with default values, parameters will first be used to override the defaults. The remainder will then be used to populate the array.

def mixed(a, b=99, *c) 
  [ a, b, c]
end

mixed(1)          #=> [1, 99, []]
mixed(1, 2)       #=> [1, 2, []]
mixed(1, 2, 3)    #=> [1, 2, [3]]
mixed(1, 2, 3, 4) # => [1, 2, [3, 4]]

Any parameter may be a prefixed with an asterisk. If a starred parameter supports the to_a method, that method is called, and the resulting array is expanded inline to provide parameters to the method call. If a starred argument does not support to_a, it is simply passed through unaltered.

class MyClass
  def to_a;
    [1,2,3]
  end
end

def my_method(*x)
  x.class
end

my_method( MyClass.new ) #=> Array

The above paragraph is not quite right. If the argument doesn't support to_a, it is not passed through unaltered, but rather in a single element array.

Keyword arguments

Ruby 2 methods may declare keyword arguments using the syntax name: default_value for each. These arguments must follow any regular arguments in the list.

If you call a method that has keyword arguments and do not provide corresponding values in the method call’s parameter list, the default values will be used. If you pass keyword parameters that are not defined as arguments, an error will be raised unless you also define a double splat argument, **arg. The double splat argument will be set up as a hash containing any uncollected keyword parameters passed to the method.

def my_method(x,y, one: :one, two: :two, three: :three)
  p [x,y,one,two,three]
end

my_method rescue puts $!           #=> wrong number of arguments (0 for 2)
my_method(1) rescue puts $!        #=> wrong number of arguments (1 for 2)
my_method(1,2)                     #=> [1, 2, :one, :two, :three]
my_method(1,2, one: :a)            #=> [1, 2, :a, :two, :three]
my_method(1,2, one: :ONE, two: :b) #=> [1, 2, :a, :b, :three]

Any parameter may be prefixed with two asterisks (a double splat). Such parameters are treated as hashes, and their key-value pairs are added as additional parameters to the method call.

Ruby 2 methods may declare keyword arguments using the syntax name: default_value for each. These arguments must follow any regular arguments in the list.

keyword arguments with defaults (not required because they have defaults, duh)

def mymeth1(a:1, b:2, c:3)
end

mymeth1(a: 1)

required keyword arguments

def mymeth2(a:, b:, c:)
end

mymeth2(a: 1) rescue puts $! #=> missing keywords: b, c
mymeth2(a: 1, b: 2, c: 3)

old school default hash for comparison

def oldmeth(options = {})
  bar = options.fetch(:bar, 'default')
  puts bar
end

oldmeth #=> default

old school hash passing, but not so old we needed =>

oldmeth(bar: 'not_default') #=> not_default

blocks can now how keyword arguments

define_method(:foo) do |bar: 'default'|
  puts bar
end

foo #=> default
foo(bar: 'baz') # => 'baz'


define_method(:foo) do |bar:| #<-- required
  puts bar
end

foo #=> missing keyword: bar

If you call a method that has keyword arguments and do not provide corresponding values in the method call’s parameter list, the default values will be used. If you pass keyword parameters that are not defined as arguments, an error will be raised unless you also define a double splat argument, **arg. The double splat argument will be set up as a hash containing any uncollected keyword parameters passed to the method.

def mymeth(a:, b:, c:, **splat)
  [a,b,c,splat]
end

mymeth(x:100,y:101,z:102,a:1,b:2,c:3) #=> [1, 2, 3, {:x=>100, :y=>101, :z=>102}]

undefine methods with undef (not quite sure why)

undef mymethod
undef :mymethod

The below paragraph indicates its even more useless.

An undefined method still exists; it is simply marked as being undefined. If you undefine a method in a child class and then call that method on an instance of that child class, Ruby will immediately raise a NoMethodError—it will not look for the method in the child’s parents.

Not quite sure why the following happens, but it could be useful if the passed in method returns a proc

def my_meth(&block)
end

def hello
  puts :hello
end

my_meth(&hello) #=> hello

This is a more normal procedure

def my_meth(&block)
  block.call
end

hello = ->{ puts :hello }
my_meth(&hello) #=> hello

block_given? seems to work also for parameter blocks

def my_meth(&block)
  block.call if block_given?
end

my_meth() #=> no error

When a receiver is explicitly specified in a method invocation, it may be separated from the method name using either a period (.) or two colons (::). The only difference between these two forms occurs if the method name starts with an uppercase letter. In this case, Ruby will assume that receiver::Thing is actually an attempt to access a constant called Thing in the receiver unless the method invocation has a parameter list between parentheses. Using :: to indicate a method call is mildly deprecated.

Foo.Bar()  # method call
Foo.Bar    # method call
Foo::Bar() # method call
Foo::Bar   # constant access

Confirmed;

class MyClass

  def my_meth
    :hello
  end

  def My_Meth
    :goodbye
  end

end

p MyClass.new.my_meth  #=> :hello
p MyClass.new::my_meth #=> :hello
p MyClass.new.My_Meth  #=> :goodbye
p MyClass.new::My_Meth rescue puts $! #=> #<MyClass:0x00000101915be8> is not a class/module

In method aliasing the first param is the new name

alias new_name name

When a method is aliased, the new name refers to a copy of the original method's body. If the original method is subsequently redefined, the aliased name will still invoke the original implementation.

Singleton form of class definition, this extends the eigenclass:

class << obj
  body
end

extends() method definition: adds to obj the instance methods from each module given as a paremeter

A Ruby class definition creates or extends an object of class Class by executing the code in body.

Playing with extending eigenclass

extending singleton of 'main' object

class << self
  def my_method
    puts "in my_method"
  end
end

my_method #=> "in my_method"

extending some object

obj = Object.new

class << obj
  def my_method
    puts 'in obj my_method'
  end
end

obj.my_method # "in obj my_method"

extending a class, this proves that class methods are instance methods of a classes eigenclass

class MyClass
  class << self
    def my_method
      puts "in MyClass my_method"
    end
  end
end

MyClass.my_method #=> "in MyClass my_method"

Remember that a class definition is executable code. Many of the directives used in class definitions (such as attr and include) are actually simply private instance methods of class Module. (note: doesn't have to be private, see below)

Writing class macro

class MyBase

  def self.my_macro
    puts :my_macro
  end

end

class MyClass < MyBase
  my_macro
end

MyClass.new #=> "my_macro"

Overriding class allocation for finer control or to take control over object initialization. This follows the principal of most surprise, so try to not do this.

# normal class definition
class MyClass
  def self.new
    obj = allocate
    obj.send(:initialize, :hello)
  end
end

class MyClass
  def initialize(x)
    puts x
  end
end

# class can-opener
class << MyClass
  def new
    obj = allocate
    obj.send(:initialize, :goodbye)
  end
end

MyClass.new #=> "goodbye"

A module is basically a class that cannot be instantiated.

A module may define an initialize method, which will be called upon the creation of an object of a class that mixes in the module if either the class does not define its own initialize method or the class's initialize method invokes super.

module MyModule
  def initialize(greeting: 'hello')
    puts greeting
  end
end

class MyClass
  include MyModule
end

MyClass.new #=> 'hello'
MyClass.new(greeting:'goodbye') #=> 'goodbye'

A module may also be included at the top level, in which case the module's constants, class variables, and instance methods become available at the top level.

Instance methods defined in modules can be mixed-in to a class using include. But what if you want to call the instance methods in a module directly?

module MyModule
  def say_hello()
    puts :hello
  end
end
include MyModule # The only way to access MyModule.say_hello
say_hello #=> hello


module MyModule
  def say_hello()
    puts :hello
  end
  module_function :say_hello #<-- improvement over include MyModule
end

MyModule.say_hello #=> hello

module_function: Creates module functions for the named methods. These functions may be called with the module as a receiver, and also become available as instance methods to classes that mix in the module. Module functions are copies of the original, and so may be changed independently. The instance-method versions are made private. If used with no arguments, subsequently defined methods become module functions.

module MyModule

  module_function #<-- applies to everything that follows if no arguments

  def say_hello
    puts :hello
  end
  def say_goodbye
    puts :goodbye
  end
end


MyModule.say_hello   #=> hello
MyModule.say_goodbye #=> goodbye

The instance method and module method are two different methods: the method definition is copied by module_function, not aliased.

Rationale for module functions: so you can write a module that have methods that can both be called Module.method and can be included into classes. Why its called module_function and not module_method I have no idea.

Access control

  • public: accessible to anyone
  • protected: Can be invoked only by objects of the defining class and its subs
  • private: Can be called only in functional form (with implicit receiver of self)

The return value of a yield is the value of the last expression evaluated in the block or the value passed to a next statement executed in the block.

Bock Arguments

  • You can specify default values
  • You can specify splat arguments
  • The last argument can be prefixed with & for a block
  • Block local variables are declared after a semi

This gets kinda crazy, not sure if I'd ever use this

def gimme_a_proc(&block)
  block.call(&->{ puts :here })
end

gimme_a_proc() do |&proc|
  proc.call
end #=> here

Proc objects

Four ways to create a proc object

  • By passing a block to a method whose last parameter is prefixed with an ampersand. That parameter will receive the block as a Proc object.

      def meth1(p1, p2, &block) #<-- will run to_proc on block
        puts block.inspect
      end
    
  • By calling Proc.new {}

      block = Proc.new { :hello }
    
  • By calling Object#lambda

      block = lambda { :hello }
    
  • lambda syntax

      lam = ->{ :hello }
      lam = ->(x,y){ puts x,y }
    

Note that there cannot be a space between -> and the opening parenthesis.

Here’s the big thing to remember: raw procs are basically designed to work as the bodies of control structures such as loops. Lambdas are intended to act like methods. So, lambdas are stricter when checking the parameters passed to them, and a return in a lambda exits much as it would from a method.

Calling a proc

proc = Proc.new {}
proc.call
proc.()  #<--- cool?
proc[]
yield

Within both raw procs and lambdas, executing next causes the block to exit back to the caller of the block. The return value is the value (or values) passed to next, or nil if no values are passed.

Dont use - deprecated

proc = proc {}

Exeptions

When an exception is raised, Ruby places a reference to the Exception object in the global variable $!.

Exceptions may be handled in the following ways:

  • Within the scope of a begin/end block:

      begin
        ..
      rescue
        ..
      else
        ..
      ensure
        ..
      end
    
  • Within the body of a method:

      begin
        # something which might raise an exception
      rescue SomeExceptionClass => some_variable
        # code that deals with some exception
      rescue SomeOtherException => some_other_variable
        # code that deals with some other exception
      else
        # code that runs only if *no* exception was raised
      ensure
        # ensure that this code always runs, no matter what
      end
    
  • After the execution of a single statement:

      x/y rescue puts $!
    

A rescue clause with no parameter is treated as if it had a parameter of StandardError

If you want to rescue every exception, use this: (note: re-raise this dammit)

rescue Exception => e

The rescue modifier takes no exception parameter and rescues StandardError and its children.

values = [ "1", "2.3", /pattern/ ]
result = values.map {|v| Integer(v) rescue Float(v) rescue String(v) }
result # => [1, 2.3, "(?-mix:pattern)"]

The method Object#catch executes its associated block:

catch ( object ) do 
  code...
end

The method Object#throw interrupts the normal processing of statements:

throw( object ‹ , obj › )

My pointless example (not sure if its actually showing throw working)

val = catch(:outer) do
  catch(:inner) do
    5.times do |x|
      throw(:outer, x) if x==3
    end
  end
end

p val #=> 3

If the throw is passed a second parameter, that value is returned as the value of the catch.

Ruby honors the ensure clauses of any block expressions it traverses while looking for a corresponding catch.

If no catch block matches the throw, Ruby raises an ArgumentError exception at the location of the throw.

Standard Protocols and Coercions

from stackoverflow

  • call to_s to get a string that describes the object.
  • call to_str to verify that an object really acts like a string.
  • implement to_s when you can build a string that describes your object.
  • implement to_str when your object can fully behave like a string.

As we can see, to_s is happy to turn any object into a string. On the other hand, to_str raises an error when its parameter does not look like a string.

Array#join calls to_str on its param. If you try a.join(3) it will try to_str on 3 and raise and exception TypeError: no implicit conversion of Fixnum into String. But 3.to_s does have a stringy conversion.

I guess this is called a protocal because its not really a set of strict rules.

The symbol.to_proc trick

names = %w{ant bee cat}
result = names.map(&:upcase)

Closer look:

class MyData

  def initialize(data)
    @data = data
  end

  def yo
    "yo #{@data}"
  end

end
  
a = [MyData.new(1), MyData.new(2), MyData.new(3)]
p a.map(&:yo) #=> ["yo 1", "yo 2", "yo 3"]

yoproc = :yo.to_proc
p yoproc.call(MyData.new(1))  #=> "yo 1"

Even closer look, defining our own Symbol#to_proc (this stuff is deep)

class MyData

  def initialize(data)
    @data = data
  end

  def yo
    "yo #{@data}"
  end

end
  
class Symbol
  def to_proc
    print "HERE! "
    proc { |obj, *args| obj.send(self, *args) } #<--- self is :yo
  end
end

a = [MyData.new(1), MyData.new(2), MyData.new(3)]
p a.map(&:yo) #=> HERE! ["yo 1", "yo 2", "yo 3"]

yoproc = :yo.to_proc
p yoproc.call(MyData.new(1))  #=> HERE! "yo 1"

Note: inject doesn't seem to need the &

[1,2,3].inject(&:+) #=> 6
[1,2,3].inject(:+)  #=> 6

Metaprogramming

Accessing class instance variables

class Test
  @var = 99 
  class << self
    attr_accessor :var
  end 
end

Object is the superclass of our custom classes. That's where the free class methods come from. Its also where "class macros" come from. Class macros and class methods are the same if they are defined one class up.

class Object
  def hello
    :hello
  end
  def goodbye
    :goodbye
  end
end

class MyClass
  print hello
end

p MyClass.goodbye #=> hello:goodbye

Access eigenclass of object (not sure why)

animal = 'dog' singleton = class << animal self end p singleton #=> #<Class:#String:0x0000010118bcb0>

When you include a module you are effectively adding it as a new superclass.

Ruby 2 introduced the prepend method. Logically, this behaves just like include, but the methods in the prepended module take precedence over those in the host class. Ruby pulls off this magic by inserting a dummy class in place of the original host class2 and then inserting the prepended module between the two.

If a method inside a prepended module has the same name as one in the original class, it will be invoked instead of the original. The prepended method can then call the original using super.

The include method effectively adds a module as a superclass of self. It is used inside a class definition to make the instance methods in the module available to instances of the class. However, it is sometimes useful to add the instance methods to a particular object. You do this using extend

Extending a class is the same as adding instance methods to its eigenclass, which end up being class methods because instance methods of a classes eigenclass are that classes class methods.

module MyModule
  def hello
    :hello
  end
end

class MyClass
  include MyModule #<----- include
end

p MyClass.new.hello #=> :hello

Object.send(:remove_const, :MyModule)
Object.send(:remove_const, :MyClass)

module MyModule
  def hello
    :hello
  end
end

class MyClass
  extend MyModule #<----- extend
end

p MyClass.hello #=> :hello

extending an object is the same as extending a class - adding instance methods to its eigenclass.

module MyModule
  def hello
    :hello
  end
end

obj = Object.new
obj.extend MyModule

p obj.hello #=> :hello

simple macro mechanism:

class MyClass

  def self.my_macro
    def say_hello
      puts :hello
    end
  end

  my_macro

end

MyClass.new.say_hello #=> hello

more dynamic macro

class MySuperClass
  def self.say_macro(word)
    define_method(:say) do
      puts word
    end
  end
end

class MyClass < MySuperClass
  say_macro(:goodbye)
end

MyClass.new.say #=> goodbye

macro from a module:

module MyModule
  def say_macro(word) #<-- now a instance method
    define_method(:say) do
      puts word
    end
  end
end

class MyClass
  extend MyModule #<-- extend, not include
  say_macro(:goodbye)
end

MyClass.new.say #=> goodbye

If you want to add both class and instance methods into a class at once you can use the included hook:

module MyModule

  def say_this(this) #<-- instance method
    puts this
  end

  module ClassMethods
    def say_macro(word)
      define_method(:say) do
        puts word
      end
    end
  end

  def self.included(host_class) #<--- hook
    host_class.extend(ClassMethods)
  end

end

class MyClass
  include MyModule #<-- include
  say_macro(:goodbye)
end

MyClass.new.say_this(:hello) #=> hello
MyClass.new.say #=> goodbye

#####Subclassing Expressions

The return value of a Struct is a class. which you ordinarily would use to make objects.

Person = Struct.new(:name, :address, :likes) #=> Person
kevin = Person.new('kevin', 'ma') #=> #<struct Person name="kevin", address="ma", likes=nil>

in the class defintion MySuperClass can be any class object

class MyClass < MySuperClass
end

class MyClass < Struct.new(:name, :address, :likes)
  def to_s
    ...
  end
end
Creating single classes
myclass = Class.new do
  def hello
    :hello
  end
end

myclass.new.hello #=> hello

passing argument to Class.new

myclass = Class.new()
p myclass.superclass #=> Object

myclass = Class.new(String)
p myclass.superclass #=> String

class MyClass < String
end
p MyClass.superclass #=> String

You may have noticed that the classes created by Class.new have no name. However, if you assign the class object for a class with no name to a constant, Ruby automatically names the class afterthe constant:

MyClass = Class.new()
p MyClass #=> MyClass

We can use these dynamically constructed classes to extend Ruby in interesting ways. For example, here's a simple reimplementation of the Ruby Struct class:

def MyStruct(*keys)
Class.new do
  attr_accessor *keys
    def initialize(hash)
      hash.each do |key, value|
        instance_variable_set("@#{key}", value)
      end
    end
  end
end

Person = MyStruct :name, :address, :likes
dave = Person.new(name: "dave", address: "TX", likes: "Stilton")

The methods Object#instance_eval, Module#class_eval, and Module#module_eval let you set self to be some arbitrary object, evaluate the code in a block with, and then reset self:

'dog'.instance_eval do
  puts self.upcase #=> DOG
  puts upcase      #=> DOG
end

class MyClass
  def self.hello
    puts :hello
  end
end

MyClass.class_eval do
  hello
end

Both forms also take a string, but this is considered a little dangerous.

class_eval and instance_eval both set self for the duration of the block. However, they differ in the way they set up the environment for method definition. class_eval sets things up as if you were in the body of a class definition, so method definitions will define instance methods. In contrast, calling instance_eval on a class acts as if you were working inside the singleton class of self. Therefore, any methods you define will become class methods.

MyClass = Class.new

MyClass.class_eval do
  def hello
    puts :hello
  end
end

MyClass.new.hello #=> hello

MyClass.instance_eval do
  def goodbye
    puts :goodbye
  end
end

MyClass.goodbye #=> goodbye

Ruby has variants of these methods. Object#instance_exec, Module#class_exec, and Module#module_exec behave identically to their _eval counterparts but take only a block (that is, they do not take a string). Any arguments given to the methods are passed in as block parameters. This is an important feature. Previously it was impossible to pass an instance variable into a block given to one of the _eval methods - because self is changed by the call, these variables go out of scope.

'cat'.instance_exec('dog') do |animal|
  puts upcase         #=> CAT
  puts animal.upcase  #=> DOG
end

Yikes

Ruby 1.9 changed the way Ruby looks up constants when executing a block using instance_eval and class_eval. Ruby 1.9.2 then reverted the change. In Ruby 1.8 and Ruby 1.9.2, constants are looked up in the lexical scope in which they were referenced. In Ruby 1.9.0, they are looked up in the scope in which instance_eval is called. This (artificial) example shows the behavior at the time I last built this book - it may well have changed again by the time you run it....

instance_eval and domain-specfic languages

class Conversation

  def start(&b)
    instance_eval(&b) #<-- secret sauce
  end

  def hello
    puts :hello
  end

  def goodbye
    puts :goodbye
  end

end

Conversation.new.start do
  hello   #=> hello
  goodbye #=> goodbye
end 

There’s a drawback, though. Inside the block, scope isn’t what you think it is, so this code wouldn’t work:

@size = 4 
turtle.walk do
  4.times do 
    turtle.forward(@size) 
    turtle.left
  end 
end

Instance variables are looked up in self, and self in the block isn’t the same as self in the code that sets the instance variable @size. Because of this, most people are moving away from this style of instance_evaled block.

Hook methods

Method related

method_added 
method_missing 
method_removed
method_undefined
singleton_method_added 
singleton_method_removed
singleton_method_undefined

Class and module related

append_features 
const_missing 
extend_object 
extended 
included 
inherited 
initialize_clone 
initialize_copy 
initialize_dup

inherited hook

class MyClassBase
  def self.inherited(child)
    p child
  end
end

class MyClassA < MyClassBase
end #=> MyClassA

MyClassC = Class.new(MyClassBase) #=> #<Class:0x0000010118b260>

method_missing Hook

The built-in method_missing basically raises an exception (either a NoMethodError or a NameError depending on the circumstances). The key here is that method_missing is simply a Ruby method. We can override it in our own classes to handle calls to otherwise undefined methods in an application-specific way.

method_missing has a simple signature, but many people get it wrong:

def method_missing(name, *args, &block)

Before we get too deep into the details, I’ll offer a tip about etiquette. There are two main ways that people use method_missing. The first intercepts every use of an undefined method and handles it. The second is more subtle; it intercepts all calls but handles only some of them. In the latter case, it is important to forward on the call to a superclass if you decide not to handle it in your method_missing implementation:

class MyClass < OtherClass
  def method_missing(name, *args, &block)
    if <some condition> # handle call
    else
      super # otherwise pass it on
    end 
  end
end

Object is the superclass of normal class defintion

class MyClassA #<--- default 'Object'
end

class MyClassB < Object
end

p MyClassA.superclass
p MyClassB.superclass

Looking Inside Classes

false argument in the following classes will prevent recurse into parent classes

Demo = Class.new
Demo.private_instance_methods(false)
Demo.protected_instance_methods(false)
Demo.public_instance_methods(false)
Demo.singleton_methods(false)
Demo.class_variables
Demo.constants(false)

demo = Demo.new
demo.instance_variables
demo.public_method

Calling methods dynamically

Object#send:

'cat'.send(:length) #=> 3
'cat'.send(:upcase) #=> CAT

Method Objects

cl = 'cat'.method(:length)
p cl #=> #<Method: String#length>
p cl.call #=>3

cl = 'cat'.method(:[])
p cl #=> #<Method: String#[]> 
p cl[1] #=> a

Another example

def plus1(x)
  x += 1
end

meth_obj = method(:plus1)

p meth_obj.to_proc #=> #<Proc:0x0000010190fbf8 (lambda)> lambda!

p [1,2,3].map(&meth_obj) #=> [2,3,4]

Method objects are bound to one particular object. You can create unbound methods (of class UnboundMethod) and then subsequently bind them to one or more objects. The binding creates a new Method object. As with aliases, unbound methods are references to the definition of the method at the time they are created. (Note: this seems very restrictive and useless)

class MyClass
  def hello; p hello end
end

meth = MyClass.instance_method(:hello)
some_random_obj = Object.new
meth.bind(some_random_obj)
some_random_obj.hello #=> bind argument must be an instance of MyClass

Hooking method calls

Old method with alias_method

class Object

  alias_method :old_puts, :puts

  def puts(*args)
    old_puts "About to puts something..."
    old_puts(*args)
  end

end

puts :hello

New method with prepend

module MyStuff
  def puts(*args)
    print "... "
    super
  end
end

class Object
  prepend MyStuff
end

puts :hello #=> "... hello"

Another (ridiculous) way to hook a method - using unbound methods of course

class MyClass
  def goodbye
    puts :goodbye
  end
end

class MyClass

  old_method = instance_method :goodbye

  # redefine goodbye to say hello first
  define_method :goodbye do
    print 'hello '
    old_method.bind(self).call
  end

end

MyClass.new.goodbye #=> hello goodbye

Print current file:

print File.read(__FILE__)

Finished - now I'm going to forget all the above.

The Well Grounded Rubyist

Part 1 - Ruby Foundations

Moving on

Part 2 - Built-in classes and modules

Ruby has a lot of built-in classes. Most of them can be instantiated using new:

str = String.new
arr = Array.new

Some can’t; for example, you can’t create a new instance of the class Integer. But for the most part, you can create new instances of the built-in classes.

In addition, a lucky, select few built-in classes enjoy the privilege of having literal constructors. That means you can use special notation, instead of a call to new, to create a new object of that class.

String      'new string'
Symbol       :symbol
Array        [1,2,3]
Hash         {:one => 1}
Range        0..9 or 0...10
Regexp       /.*/
Proc         ->(x,y){ x * y }

By defining + you get += for free

class MyInt

  def initialize(int)
    @int = int
  end

  def +(other)
    @int + other
  end

end

int = MyInt.new(1)

p int + 1 #=> 2

int += 1
p int     #=> 2

Section is about defining sugar, but don't forget Forwardable can make this easier:

require 'forwardable'
class MyInt

  extend Forwardable
  def_delegators :@int, :to_s, :+

  def initialize(str)
    @int = str
  end

end

int = MyInt.new(1)

p int
p int.to_s #=> "1"
p int + 1 #=> 2

Don't forget, too, the conditional assignment operator ||=, as well as its rarely spotted cousin &&= , both of which provide the same kind of shortcut as the pseudooperator methods but are based on operators, namely || and && , which you can't override.

Rare &&= : assigns value only if target is 'true'

var = 1
var &&= 2 # same as next line
# var = var && 2 
p var #=> 2

var = nil
var &&= 2 # same as next line
# var = var && 2
p var #=> nil

Defining unary operator

class MyClass
  def +@ # <--- weird but right
    :+
  end

  def -@ # <--- weird but right
    :-
  end
end

a = MyClass.new
p +a #=> :+

b = MyClass.new
p -b #=> :-

There's an overridable ! method, huh

class MyClass

  def !
    'calling !'
  end

end

o = MyClass.new
p !o #=> 'calling !'
p (not o) #=> 'calling !' 

I wonder how the following happens

When it comes to generating string representations of their instances, some built-in classes do things a little differently from the defaults. For example, if you call puts on an array, you get a cyclical representation based on calling to_s on each of the elements in the array and outputting one per line. That is a special behavior; it doesn't correspond to what you get when you call to_s on an array, namely a string representation of the array in square brackets, as an array literal.

You’ve already seen the star operator used in method parameter lists, where it denotes a parameter that sponges up the optional arguments into an array. In the more general case, the star turns any array, or any object that responds to to_a, into the equivalent of a bare list.

The term bare list means several identifiers or literal objects separated by commas. Bare lists are valid syntax only in certain contexts. For example, you can put a bare list inside the literal array constructor brackets:

[1,2,3,4,5]

It’s a subtle distinction, but the notation lying between the brackets isn’t, itself, an array; it’s a list, and the array is constructed from the list, thanks to the brackets.

All these work the same, assinging x=1, y=2, z=3

x,y,z = 1,2,3
puts x,y,z

x=y=z = nil

x,y,z = [1,2,3]
puts x,y,z

x=y=z = nil

x,y,z = *[1,2,3]
puts x,y,z

The star has a kind of bracket-removing or un-arraying effect. What starts as an array becomes a list.

Use to_s for display purposes, use to_str if you want your object to be a string.

obj + "there." #=> "Hello there." (uses to_str)

obj << " that" #=> "this that" (uses to_str)

to_str is required for the following to work (to_s) didn't do it

class MyClass

  def initialize(data)
    @data = data
  end
  
  def to_str
    @data.to_s 
  end

end

p "my string " + MyClass.new([1,2,3]) #=> "my string [1, 2, 3]"

Objects can masquerade as arrays if they have a to_ary method. If such a method is present, it’s called on the object in cases where an array, and only an array, will do— for example, in an array-concatenation operation.

[1,2,3].concat(my_obj)

If you define ==, your objects will automatically have the != method.

But for classes that do need full comparison functionality, Ruby provides a convenient way to get it. If you want objects of class MyClass to have the full suite of comparison methods, all you have to do is the following:

  1. Mix a module called Comparable (which comes with Ruby) into MyClass.

  2. Define a comparison method with the name <=> as an instance method in MyClass.

    class MyInt

    attr_accessor :int

    def initialize(int) self.int = int end

    end

    one = MyInt.new 1 two = MyInt.new 1

    p one == two #=> false

    reopen class

    class MyInt

    include Comparable

    def <=>(other) self.int <=> other.int end

    end

    one = MyInt.new 1 two = MyInt.new 1

    p one == two #=> true

method reflection method grouping (to help remembering)

MyClass.methods #<-- class methods

MyClass.instance_methods #<-- instance methods defined in class
MyClass.public_instance_methods
MyClass.private_instance_methods
MyClass.protected_instance_methods


obj.methods #<-- instance methods
obj.private_methods    
obj.public_methods
obj.protected_methods
obj.singleton_methods

MyClass.new.methods == MyClass.instance_methods
MyClass.instance_methods == MyClass.public_instance_methods

Extract a string with [] (also known as slice)

"my string"[3, 100] #=> "string" (second arg is length, not position)
"my string"[3..-1] #=> "string" use a range if you have two positions

Test for substring without a regex

p "my string"['string'] #=> 'string' # finds and returns a substring
p "my string"['no string'] #=> nil

target = 'str'
if 'my string'[target]
  puts "found target #{target}"
end

The regex option

target = /str/
if 'my string'[target]
  puts "found target #{target}" #+> found target (?-mix:str)
end

Above better than this? (maybe because I always type =~ wrong)

target = /str/
if 'my string' =~ target
  puts "found target #{target}" #+> found target (?-mix:str)
end

One more

target = /str/
if 'my string'.match(target)
  puts "found target #{target}" #+> found target (?-mix:str)
end

Found another one

target = 'str'
if 'my string'.include? target
  puts "found target #{target}"
end

slice has a bang!

str = 'my string'
p str.slice!('my ') #=> 'my '
p str #=> "string"

To set part of a string to a new value, you use the []= method. Or just use sub!/gsub!

str = 'my string'
str['my'] = 'not my'
p str #=> 'not my string'

str = 'my string'
str.sub!('my', 'not my')
p str #=> 'not my string'

Count the letters in a string

'this is my string'.count('i') #=> 3
'this is my string'.count('a-z') #=> character range

Consider allowing symbols or strings as method arguments When you’re writing a method that will take an argument that could conceivably be a string or a symbol, it’s often nice to allow both. It’s not necessary in cases where you’re dealing with user-generated, arbitrary strings, or where text read in from a file is involved; those won’t be in symbol form anyway. But if you have a method that expects, say, a method name, or perhaps a value from a finite table of tags or labels, it’s polite to allow strings or symbols. That means avoiding doing anything to the object that requires it to be one or the other and that will cause an error if it’s the wrong one. You can normalize the argument with a call to to_sym (or to_s, if you want to normalize to strings) so that whatever gets passed in fits into the operations you need to perform.

Weird parens when using hash.each_with_index

a.each_with_index {|(k,v), i| p k,v,i }

Spliltting also works with procs but not lambdas because theres no || in lambdas

p = Proc.new {|(x,y)| puts "first:#{x} second:#{y}" }
p.call([1,2]) #=> first:1 second:2

l = lambda { |(a,b)| puts "first:#{a} second:#{b}" }
l.call([1,2])

Doesn't work with arrow lambda syntax because there's no | |

l = ->(x){ puts "first:#{?} second:#{?}" }
l.call([1,2])

Hash to array intermission

h = Hash[*[1,2,3,4]]  

# note, this is how you construct a hash with keys and values, not Hash.new(a:1, b:2)
h = Hash[a:1, b:2]

# new() is for defaults
h2 = Hash.new(99)
h2[:one] #=> 99

Array constructor can take a block (note the 3)

x = 1
a = Array.new(3) do
  x *= 10
end

p a #=> [10, 100, 1000]

hash intermission (note, Hash() is a method), Hash[] and Hash() are the same

p Hash(a:1, b:2, c:3) #=> {:a=>1, :b=>2, :c=>3}
p Hash[a:1, b:2, c:3] #=> {:a=>1, :b=>2, :c=>3}

[]= and Array() method

class Arrayish

  def initialize(a)
    @a = a
  end

  def []=(index, value)
    @a[index] = value
  end

  def to_a
    @a
  end

end

a = Arrayish.new(%i{a b c})
a[3] = :d # because we implemented []=
p a.class #=> Arrayish
real_array = Array(a) # because we implemented to_a
p real_array.class #=> Array

To make slices work you need to build []= like this

def []=(*index, value)
  @a[*index] = value
end

a = Arrayish.new(%i{a b c})
a[0,2]  = :x, :y
p a.to_a #=> [:x, :y, :c]

Array#push can take multiple values, Array#<< can't

a = %i{ one two three}
a << [:four, :five]
p a #=> [:one, :two, :three, [:four], [:five]]

a = %i{ one two three}
a.push *[:four, :five] #=> note the splat
p a #=> [:one, :two, :three, :four, :five]

Array#pop can take a number argument

a = %i{ one two three}
p a.pop(2) #=> [:two, :three]
p a        #=> :one

Array#flatten can take level argument - there's also a flatten!

array = [1,2,[3,4,[5],[6,[7,8]]]]
array.flatten(2) #=> [1, 2, 3, 4, 5, 6, [7, 8]]      

You can uniq arrays - also uniq!

[1,2,3,1,4,3,5,1].uniq #=> [1, 2, 3, 4, 5] 

Array#compact removes nils from arrays (not false)

[1, nil, 2, nil, 3].compact #=> [1,2,3]

Three ways to create a Hash

{one:1, two:2, three:3}
Hash.new(:default)
Hash[:one, 1, :two, 2, :three, 3] #<-- takes a 'list'

Hash#store takes two arguments

hash = {}
hash.store(:one, 1) #<-- list
hash.store(*[:two, 2]) #<-- splat array
hash.store( *'three 3'.split ) #<-- really grasping
p hash #=> {:one=>1, :two=>2, "three"=>"3"}

My attempt to unique values of hash elegantly

h = {a:1, b:2, c:3, d:1, e:2}
h = h.invert.invert
p h #=> {:d=>1, :e=>2, :c=>3}

My attempt to sort hash by values

h = {a:1, b:2, c:3, d:1, e:2}
h = h.sort { |a,b| a[1] <=> b[1] } #--> there is no Hash#sort!
p Hash[*h.flatten] #=> {:a=>1, :d=>1, :b=>2, :e=>2, :c=>3} 

breaking news... there's a sort_by and a to_h, use _k for unused block param

hash = {a:1, b:2, c:3, d:1, e:2}
hash = hash.sort_by { |_k,v| v }.to_h
p hash #=> {:a=>1, :d=>1, :b=>2, :e=>2, :c=>3} 

Waring Array#to_h is weird

[:one, 1, :two, 2, :three, 3].to_h #=> wrong element type Symbol at 0 (expected array) 

[[:one, 1], [:two, 2], [:three, 3]].to_h #=> {:one=>1, :two=>2, :three=>3}

To change array into hash in one line, use each_slice, enumerator.to_a

[:one, 1, :two, 2, :three,3].each_slice(2).to_h

Hash#fetch is versitile

h = {one: 1, two: 2, three: 3}

p h.fetch(:three) #=> 3
p h.fetch(:four, 4) #=> 4
h.fetch(:four) rescue p $! #=> key not found: :four (KeyError) 
h.fetch(:four) {|k| p "sorry no #{k}" } #=> sorry no four

Hash#values_at is similar to Array#values_at

h = {one: 1, two: 2, three: 3}

p h.values_at(:one, :three) #=> [1,3]

Hash constructor takes a default value

h = Hash.new(0)
h[:nope] #=> 0

Hash constructor can autovivify

h = Hash.new {|hash,key| hash[key] = 0 }

h.store(:one,1)
p h #=> {:one => 1}

h[:two] #=> 0
p h #=> {:one => 1, :two => 2} 

Ranges

Ranges are enclusive or exclusive

(1..10).include?  10 #=> true
(1...10).include? 10 #=> false

Think of ... as pushing the last value outside the range

Ranges and include? and cover?, cover? is simplier and faster, won't use with strings because its confusing. Note that cover? is exclusive to ranges

p ('a'..'c').cover? 'abc' #=> true
p ('a'..'c').include? 'abc' #=> false

p ('a'..'c').cover? 'bcd' #=> true
p ('a'..'c').include? 'bcd' #=> false

p ('a'..'c').cover? 'cde' #=> false
p ('a'..'c').include? 'cde' #=> false

This test is useless

[1,2,3,nil,4,5,6].find {|n| n.nil? }

This could work

[1,2,3,nil,4,5,6].find_all {|n| n.nil? }.count > 0

This doesn't because [nil].any? is false

[1,2,3,nil,4,5,6].find_all {|n| n.nil? }.any?

Enumerable#grip operates on === and works like:

enumerable.select {|element| expression === element }

Enumerable#group_by is cool

colors = %w{ red orange yellow blue indigo violet }

colors.group_by{|c| c[0]} #=> {"r"=>["red"], "o"=>["orange"], "y"=>["yellow"], "b"=>["blue"], "i"=>["indigo"], "v"=>["violet"]}

colors.group_by{|c| c.size} #=> {3=>["red"], 6=>["orange", "yellow", "indigo", "violet"], 4=>["blue"]}

Enumerable#partition is also cool

a = [1,2,3,4,5,6]
p a.partition {|x| (1..3).cover?(x)} #=> [[1, 2, 3], [4, 5, 6]] 

Enumerable#take and Enumerable#drop are complementary (neither alters original array) drop is like an undestructive pop and take is like an undestructive shift

a = [1,2,3,4,5,6]
a.drop(2) #=> [3, 4, 5, 6]
a         #=> [1, 2, 3, 4, 5, 6]
a.pop(4)  #=> [3, 4, 5, 6]
a         #=> [1, 2]

a = [1,2,3,4,5,6]
a.take(2)  #=> [1,2]
a          #=> [1, 2, 3, 4, 5, 6]
a.shift(2) #=> [1,2]
a          #=> [3, 4, 5, 6]

Constrain Enumerable#take and Enumerable#drop with take_while and drop_while

[1,2,3,4,5,6].take_while {|x| x < 4 } #=> [1,2,3]

loop catches StopIteration exception, so yo can use it like this

e = [1,2,3,4,5,6].each_slice(2)

loop do
  p e.next
end

which looks more pro?

e.inject([]){|acc, x| acc << x; acc} 
e.reduce([]){|acc, x| acc << x; acc} 

fancy map in place

names = %w{ David Yukihiro Chad Amy }
names.map!(&:upcase)

Even though Ruby strings aren’t enumerable in the technical sense (String does not include Enumerable), the language thus provides you with the necessary tools to address them as character, byte, codepoint, and/or line collections when you need to.

Sorting enumerables: If you have a class, and you want to be able to arrange multiple instances of it in order, you need to do the following:

  1. Define a comparison method for the class (<=>).
  2. Place the multiple instances in a container, probably an array.
  3. Sort the container.
class MyClass

  attr_accessor :var

  def initialize(x)
    self.var = x
  end

  def <=>(other)
    self.var <=> other.var
  end

end

collection = [5,1,3,4,2].map {|x| MyClass.new(x)}
collection.sort.map {|c| c.var} #=> [1,2,3,4,5]

Comparable, enumerables and <=>:

  • If you define <=> for a class, then instances of that class can be put inside an array or other enumerable for sorting.
  • If you don’t define <=>, you can still sort objects if you put them inside an array and provide a code block telling the array how it should rank any two of the objects.
  • If you define <=> and also include Comparable in your class, then you get sortability inside an array and you can perform all the comparison operations between any two of your objects (>, <, and so on)

Poor mans sorting (class doesn't have <=> method)

class MyClass

  attr_accessor :var

  def initialize(x)
    self.var = x
  end

end

collection = [5,1,3,4,2].map {|x| MyClass.new(x)}
collection.sort{|a,b| a.var <=> b.var }.map {|c| c.var} #=> [1,2,3,4,5]

Enumerable#sort_by makes it even cleaner

class MyClass

  attr_accessor :var

  def initialize(x)
    self.var = x
  end

end

collection = [5,1,3,4,2].map {|x| MyClass.new(x)}
collection.sort_by{|x| x.var }.map {|c| c.var} #=> [1,2,3,4,5]

# even shorter!
collection.sort_by(&:var).map {|c| c.var} #=> [1,2,3,4,5]

Enumerators

At heart, an enumerator is a simple enumerable object. It has an each method, and it employs the Enumerable module to define all the usual methods—select, inject, map, and friends—directly on top of its each.

An enumerator isn’t a container object. It has no “natural” basis for an each operation, the way an array does (start at element 0; yield it; go to element 1; yield it; and so on). The each iteration logic of every enumerator has to be explicitly specified. After you’ve told it how to do each, the enumerator takes over from there and figures out how to do map, find, take, drop, and all the rest.

Simple enumerator with new

e = Enumerator.new do |y|
  y.yield :a #<-- each call and stop
  y.yield :b #<-- each call and stop
  y.yield :c #<-- each call and stop
end


loop do
  p e.next
end

Same

Object#to_enum
Object#to_enum(:each)

From a blog: How to make your class return an enumerator instead of including Enumerable

class UsersWithGravatar
  def each

    return enum_for(:each) unless block_given? # Sparkling magic!

    User.find_each do |user|
      hash  = Digest::MD5.hexdigest(user.email)
      image = "http://www.gravatar.com/avatar/#{hash}"
      yield user unless Net::HTTP.get_response(URI.parse(image)).body == missing_avatar
    end
  end
end

proof that to_enum hooks up to each

o = Object.new

def o.each
  print 'running each '
  yield 1
  yield 2
  yield 3
end

e = o.to_enum p e.next #=> "running each 1"

My example of implementing an enumerator for MyClass without cheating with an existing each()

class MyClass

  attr_accessor :data

  def initialize(data)
    self.data = data
  end

  def each(&block)
    for x in self.data #<-- no state needed here, uses fibers behind the scene
      block.call(x) 
    end
  end

  def to_enum
    Enumerator.new do |y|
      each do |x| #<-- our implemented each
        y << x #<-- blocks after each next
      end
    end
  end

end

o = MyClass.new([:a,:b,:c])
e = o.to_enum

p e.map.with_index {|x, i| [x,i]}.to_h #=> {:a=>0, :b=>1, :c=>2}

Enumerators really aren't tied to each. It just happens that the to_enum method of most classes will default to using each, enumerators aren't dependent on any particular method to work.

a = [1,2,3,4,5]
e = Enumerator.new do |y|
  total = 0
  until a.empty?
    total += a.pop
    y << total
  end
end

loop do 
  print e.next #=> 5912141
end

An enumerator just needs something to feed to <<

e = Enumerator.new do |y|
  loop do
    y << Time.now
  end
end

e.first(3) #=> [2015-02-19 18:51:13 -0500, 2015-02-19 18:51:13 -0500, 2015-02-19 18:51:13 -0500]

An enumerator is an object, and can therefore maintain state. It remembers where it is in the enumeration. An iterator is a method. When you call it, the call is atomic; the entire call happens, and then it’s over.

My example of how to implement to_enum that takes a method parameter

class MyClass

  attr_accessor :data

  def initialize(data)
    self.data = data
  end

  def to_enum(method)
    Enumerator.new do |y|
      self.send(method) do |x| #<-- special sauce
        y << x
      end
    end
  end

  def myeach(&b)
    for i in data
      b.call(i)
    end
  end

end

o = MyClass.new(%i{one two three four five})
e = o.to_enum(:myeach)

loop do
  p e.next
end

Regex

I'm leaning to accept that the below are only good for their bool value (as a more rubyish replacement for =~) and are otherwise not very useful without captures - nobody cares about prematch or postmatch, etc.

regex.match(string)
string.match(regex)

str.match(/(my)/) #=> #<MatchData "my" 1:"my">
str.match(/not my/) #=> nil
str.match(/(my)/).begin(0) #=> 8
str.match(/(my)/).to_a #=> ['my', 'my'] [whole string, captures...]
str.match(/(my)/).pre_match #=> 'this is '
str.match(/(my)/).post_match #=> ' string'

With captures

str.match(/my/)[0] #=> my
str.match(/my/)[1] #=> nil, whoops forgot to capture
str.match(/(my)/)[1] #=> my

Two ways to get at captures. I like the second if there's room because it doesn't have the indexing confusion

str.match(/(my)/)[1] #=> my
str.match(/(my)/).captures(0) #=> my

m[1] == m.captures[0]
m[2] == m.captures[1]

Anything inside a (?:) grouping will be matched based on the grouping, but not saved to a capture.

Escaping regex

search_for = 'a.c'
re = /#{Regexp.escape(str)}/ #=>  /a\.c/
re.match("a.c") #=>  #<MatchData "a.c">
re.match("abc") #=> nil, because '.' is not /./

'Zero or one' is slightly unintuitive to me when used like this

''.match(/(.?)/)[1] #=> ""

But I've often used it like this:

'name: kevin'.match(/name:? (.*)/)[1] #=> 'kevin'

Don't forget ignores match (bad exaple)

'one two three'.match(/(\w*) (?:\w*) (\w*)/).captures

Regular expressions, it should be noted, can't do everything. In particular, it's a commonplace and correct observation that you can't parse arbitrary XML with regular expressions, for reasons having to do with nesting of elements and the ways in which character data are represented.

String#scan is great

 "testing 1 2 3 testing 4 5 6".scan(/\d/) 
 #=> ["1", "2", "3", "4", "5", "6"]

'one two three'.scan(/\w+/) 
#=> ["one", "two", "three"]

If you use parenthetical groupings in the regexp you give to scan, the operation returns an array of arrays. Each inner array contains the results of one scan through the string:

# without
"first:Kevin last:Swope first:Bob last:Smith".scan(/first:\w+ last:\w+/) 
#=> ["first:Kevin last:Swope", "first:Bob last:Smith"]

# with
"first:Kevin last:Swope first:Bob last:Smith".scan(/first:(\w+) last:(\w+)/) 
#=> [["Kevin", "Swope"], ["Bob", "Smith"]]


"first:Kevin last:Swope first:Bob last:Smith".scan(/first:(\w+) last:(\w+)/).each do |(f,l)|
  p "#{f} => #{l}"
end

"Kevin => Swope"
"Bob => Smith"

# don't need an each, scan takes a block and knows what to do

"first:Kevin last:Swope first:Bob last:Smith".scan(/first:(\w+) last:(\w+)/) do |(f,l)|
  p "#{f} => #{l}"
end

Note that if you provide a block, scan doesn't store the results up in an array and return them; it sends each result to the block and then discards it. That way, you can scan through long strings, doing something with the results along the way, and avoid using up a lot of memory-saving substrings you've already seen and used.

Using the notion of a pointer into the string, StringScanner lets you traverse across the string as well as examine what's already been matched and what remains. StringScanner is a useful complement to the built-in string scanning facilities.

Enumerable#grep does a filtering operation from an enumerable object based on the case equality operator (===), returning all the elements in the enumerable that return a true value when threequaled against grep's argument. Thus if the argument to grep is a regexp, the selection is based on pattern matches, as per the behavior of Regexp#===:

Part 3 - Ruby Dynamics

Class methods are just instance methods defined on a classes eigenclass:

MyClass = Class.new

# define class method "outside" class
def MyClass.hello
  p :hello
end

# define class method "inside" class
class MyClass
  def self.goodbye
    p :goodbye
  end
end

MyClass.hello   #=> :hello
MyClass.goodbye #=> :goodbye

But to get inside the definition body of a singleton class, you use a special notation:

class << object
 # method and constant definitions here
end

Example:

class MyClass
  class << self

    attr_accessor :var #<-- access to class instance var

    def mymethod #<-- class method
    end

  end
end

The << object notation means the anonymous, singleton class of object. When you're inside the singleton class definition body, you can define methods - and these methods will be singleton methods of the object whose singleton class you're in.

Comparing class and instance variables inside eigenclass

MyClass = Class.new

class << MyClass
  def hello
    p :hello
  end
end

MyClass.hello #=> :hello


obj = Object.new

class << obj
  def hello
    p :hello
  end
end

obj.hello #=> :hello

class keyword either accepts a constant or a << object:

class MyClass
  #...
end

class << myobj
  #...
end

Most common use of class << is to open a classes eigenclass to create class methods, for those too lazy to type def self.method

class MyClass
  class << self
    def mymethod
  end
end

Mixing a module into an objects eigenclass puts the modules method in front of the classes instance method, which seems to be the same effect as prepend()

class MyClass
  def hello
    p :hello
  end
end

o = MyClass.new
o.hello #=> :hello

module MyModule
  def hello
    p :goodbye
  end
end

class << o
  include MyModule
end

o.hello #=> :goodbye

Almost every object in Ruby can have methods added to it. The exceptions are instances of certain Numeric subclasses, including integer classes and floats, and symbols.

Class methods are eigenclass methods but there's a lookup chain, unlike an objects eigenclass methods.

class C
  def self.hello
    p :hello
  end
end

class D < C
end

D.hello #=> :hello

Singleton classes of class objects are sometimes called metaclasses

Kernel#extend - Adds to obj the instance methods from each module given as a parameter.

module MyModule
  def hello
    p :hello
  end
end

class MyClass
end

MyClass.extend(MyModule) # adds to eigenclass of class
MyClass.hello #=> :hello

o = Object.new
o.extend(MyModule) # adds to eigenclass of object
o.hello #=> :hello

Looks like extend is include that targets the eigenclass

Using extend to modify a single object in a convenient way

module HashWithLogging
  def []=(key, value)
    puts "Assigning #{value} to #{key}"
    super
  end
end

hash = {}
hash.extend(HashWithLogging)
hash[:one] = 1 #=> Assigning 1 to :one

This does the same but looks terrible

module HashWithLogging
  def []=(key, value)
    puts "Assigning #{value} to #{key}"
    super
  end
end

hash = {}

class << hash
  include HashWithLogging
end

hash[:one] = 1 #=> Assigning 1 to :one

In other code, these two are the same

hash.extend(HashWithLogging)

class << hash
  include HashWithLogging
end

Side note, I'm thinking that alias_method exists only to create decorators for methods. Its a way to redefine a method (not override in a subclass) without losing it. The only reason to do that is to impersonate the existing method and wrap it.

BasicObject

obj.class.ancestors.last == BasicObject

Moving on

Callable and Runnable Objects

class Printer

  def self.to_proc
    Proc.new {|x| print x}
  end

end

%i{a b c d e}.each(&Printer) #=> abcde

# same as

myproc = Printer.to_proc
%i{a b c d e}.each(&myproc) #=> abcde

# same as - because proc == proc.to_proc

myproc = Printer.to_proc
%i{a b c d e}.each(&myproc.to_proc.to_proc.to_proc) #=> abcde

Symbol.to_proc trick useful for when a method doesn't take any arguments

class Symbol
  def to_proc
    Proc.new {|obj| obj.send(self) } #<-- note self is the method as a symbol, like :capitalize
  end
end

class Fixnum
  def one
    puts self
  end
end

[1,2,3].map(&:one) #=> 123

Procs can accept a splat in the block argument list

pr = Proc.new {|*x| p x }
pr.call()
pr.call(1)
pr.call(2)

instance_eval

instance_eval is mostly useful for breaking in to what would normally be another object's private data - particularly instance variables.

class MyClass

  attr_accessor :var

  def initialize(data)
    self.var = data
  end

end

o = MyClass.new(:hello)

o.instance_eval do
  p self.var #=> hello
end

instance_eval has a close cousin called instance_exec. The difference between the two is that instance_exec can take arguments. Any arguments you pass it will be passed, in turn, to the code block. This enables you to do things like this:

string = "A sample string"
string.instance_exec("s") {|delim| self.split(delim) } #=> all this to use 'self'?

The most useful eval: class_eval (a.k.a. module_eval)

Class eval puts you inside the class definition

class MyClass
  attr_accessor :var
  def initialize(data)
    self.var = data
  end
end

MyClass.class_eval do
  def upcase
    var.upcase
  end
end

o = MyClass.new(:hello)

p o.upcase #=> :HELLO

But you can do some things with class_eval that you can't do with the regular class keyword:

  • Evaluate a string in a class-definition context
  • Open the class definition of any anonymous class (not just singleton classes)
  • Use existing local variables inside a class definition body

Using local variable inside class definition, but you have to also use define method (flatten scopes) class_eval "flattens scopes" when creating classes

C = Class.new

var = :hello

C.class_eval do

  define_method(:return_var) do
    var
  end

end

p C.new.return_var #=> :hello

When you open a class with the class keyword, you start a new local-variable scope. But the block you use with class_eval can see the variables created in the scope surrounding it.

def inside a instance_eval creates a method on the classes eigenclass, aka, class method, just like it would defining a method on an ordinary object with instance_eval

C = Class.new
C.instance_eval do
  def hello
    :hello_from_instance
  end
end

p C.hello #=> :hello_from_instance (instance of Class, aka eigenclass)

class_eval on a class just reopens the class, but its a way to dynamically define method with define_method(:method)

C = Class.new
C.class_eval do
  def hello
    :hello_from_class
  end
end

p C.new.hello #=> :hello_from_class

NOTE: don't rely on bindings and closures for instnce_eval, it can do weird things

Parallel execution with threads

Moving on - I'm more of an event-driven guy

Callbacks, hooks, and runtime introspection

method_missing as a delegate/forwardable

class Cookbook
  attr_accessor :title, :author
  def initialize
    @recipes = []
  end
  def method_missing(m,*args,&block)
    @recipes.send(m,*args,&block)
  end
end

c = Cookbook.new
c << :rice
c.concat [:beer, :flour, :pepper]

p c.to_a #=> [:rice, :beer, :flour, :pepper]

forwardable alternative, limitation is that all delegated methods need to be explicit

require 'forwardable'

class Cookbook

  extend Forwardable #<-----
  def_delegators :@recipes, :<<, :concat, :to_a

  attr_accessor :title, :author

  def initialize
    @recipes = []
  end

end

c = Cookbook.new
c << :rice
c.concat [:beer, :flour, :pepper]

p c.to_a #=> [:rice, :beer, :flour, :pepper]

Module#included

module MyModule
  def self.included(c) #<-- class method
    p "#{c} included"
  end
end

MyClass = Class.new

MyClass.class_eval do
  include MyModule
end

Module#extend works the same way

module MyModule
  def self.extended(c)
    print "#{c} extended "
  end
  def hello
    p :hello
  end
end

MyClass = Class.new

MyClass.class_eval do
  extend MyModule
end

MyClass.hello #=> MyClass extended :hello

When would it be useful for a module to intercept its own inclusion like this? One commonly discussed case revolves around the difference between instance and class methods. When you mix a module into a class, you're ensuring that all the instance methods defined in the module become available to instances of the class. But the class object isn't affected. The following question often arises: What if you want to add class methods to the class by mixing in the module along with adding the instance methods?

module MyMixedModule

  def self.included(klass)
    klass.extend(ClassMethods)
  end

  def instance_method
    p "hello from instance_method"
  end

  module ClassMethods
    def class_method
      p "hello from class_method"
    end
  end

end

class MyClass
  include MyMixedModule
end

MyClass.new.instance_method #=> "hello from instance_method"
MyClass.class_method        #=> "hello from class_method"

In effect, extending an object with a module is the same as including that module in the object's singleton class. Whichever way you describe it, the upshot is that the module is added to the object's method-lookup path, entering the chain right after the object's singleton class.

module MyModule
  def hello
    p :hello
  end
end

o1 = Object.new
o1.extend(MyModule) # extend includes a module into the eigenclass
o1.hello #=> :hello

o2 = Object.new
class << o2 # enter eigenclass
  include MyModule
end

o2.hello #=> :hello

Class#inherited

class A
  def self.inherited(klass)
    p "#{B} inherited #{A}"
  end
end

class B < A #=> B inherited A
end

Module#const_missing and const_set

class MyClass
  def self.const_missing(const)
    const_set(const, :hello)
    "#{self} #{const}"
  end
end

p MyClass::A #=> "MyClass A"
p MyClass::A #=> :hello

method_added and singleton_method_added methods exist but you probably wont ever use them... Moving on

Interpreting object capability queries ( introspection? )

With method_missing, you can arrange for an object to provide a response when sent a message for which it has no corresponding method. But respond_to? won't know about such messages and will tell you that the object doesn't respond to the message even though you get a useful response when you send the message to the object. Some Rubyists like to override respond_to? so that it incorporates the same logic as method_missing for a given class. That way, the results of respond_to? correspond more closely to the specifics of what messages an object can and can't make sense of.

Others prefer to leave respond_to? as it stands, on the grounds that it's a way to check whether an object already has the ability to respond to a particular message without the intervention of method_missing. Given that interpretation, respond_to? corresponds closely to the results of methods. In both cases, the scope of operations is the entirety of all public methods of a given object.

If you want to know which of the methods defined in the Enumerable module are overridden in Range? You can find out by performing an and (&) operation on the two lists of instance methods: those defined in Enumerable and those defined in Range:

Range.instance_methods(false) & Enumerable.instance_methods(false)

Getting object singleton methods

MyClass = Class.new

class << MyClass
  def hello
    p :hello_from_singleton_method
  end
end

p MyClass.singleton_methods #=> [:hello]

Getting methods that are only inherited

File.singleton_methods - File.singleton_methods(false)

Getting variables

local_variables
global_variables

page 461

Effective Javascript

Item 1: Know Which Javascript You Are Using

...many JavaScript engines support a const keyword for defining variables, yet the ECMAScript standard does not provide any definition for the syntax or behavior of const. Moreover, the behavior of const differs from implementation to implementation.

ES5 introduced another versioning consideration with its strict mode.

"use strict";
  • Decide which versions of JavaScript your application supports.
  • Be sure that any JavaScript features you use are supported by all environments where your application runs.
  • Always test strict code in environments that perform the strictmode checks.
  • Beware of concatenating scripts that differ in their expectations about strict mode.

Item 2: Understand Javascripts Floating Point Numbers

... all numbers in JavaScript are double-precision floating-point numbers, that is, the 64-bit encoding of numbers specified by the IEEE 754 standard - commonly known as "doubles."

typeof 1 //=> "number"
typeof 2 //=> "number"

Ugh

.2 + .1 //=> 0.30000000000000004 
  • JavaScript numbers are double-precision floating-point numbers.
  • Integers in JavaScript are just a subset of doubles rather than a separate datatype.
  • Bitwise operators treat numbers as if they were 32-bit signed integers.
  • Be aware of limitations of precisions in floating-point arithmetic.

Item 3: Beware of Implicit Coercions

Ugh

3 + true //=> 4

Since NaN is the only JavaScript value that is treated as unequal to itself, you can always test if a value is NaN by checking it for equality to itself:

var a = NaN;
a !== a; //=> true

Objects can also be coerced to primitives. This is most commonly used for converting to strings:

"the Math object: " + Math; //=> "the Math object: [object Math]"
"the JSON object: " + JSON; //=> "the JSON object: [object JSON]"

Objects are converted to strings by implicitly calling their toString method. You can test this out by calling it yourself:

Math.toString(); // "[object Math]"
JSON.toString(); // "[object JSON]"

Similarly, objects can be converted to numbers via their valueOf method. You can control the type conversion of objects by defining these methods:

"J" + { toString: function() { return "S"; } }; //=> "JS"
2 * { valueOf: function() { return 3; } }; //=> 6

valueOf overrides toString, so only use valueOf for numberish objects

var obj = {
  toString: function() {
    return "[object MyObject]";
  },
  valueOf: function() {
    return 17;
  }
};

"object: " + obj; //=> "object: 17"

Most JavaScript values are truthy, that is, implicitly coerced to true. There are exactly seven falsy values: false, 0, -0, "", NaN, null, and undefined. All other values are truthy.

The more precise way to check for undefined is to use typeof:

typeof y === "undefined"

# or

if (x === undefined) { ... }
  • Type errors can be silently hidden by implicit coercions.
  • The + operator is overloaded to do addition or string concatenation depending on its argument types.
  • Objects are coerced to numbers via valueOf and to strings via toString.
  • Objects with valueOf methods should implement a toString method that provides a string representation of the number produced by valueOf.
  • Use typeof or comparison to undefined rather than truthiness to test for undefined values.

Item 4: Prefer Primitives to Object Wrappers

In addition to objects, JavaScript has five types of primitive values: booleans, numbers, strings, null, and undefined. (Confusingly, the typeof operator reports the type of null as "object", but the ECMAScript standard describes it as a distinct type.)

At the same time, the standard library provides constructors for wrapping booleans, numbers, and strings as objects. You can create a String object that wraps a string value:

var s = new String("hello");

That you can't compare the contents of two distinct String objects using built-in operators.

console.log("hello" === 'hello'); //=> true
console.log(new String("hello") === new String('hello')); //=> false

Since these wrappers don't behave quite right, they don't serve much of a purpose. The main justification for their existence is their utility methods. JavaScript makes these convenient to use with another implicit coercion: You can extract properties and call methods of a primitive value, and it acts as though you had wrapped the value with its corresponding object type. For example, the String prototype object has a toUpperCase method, which converts a string to uppercase. You can use this method on a primitive string value:

"hello".toUpperCase(); // "HELLO"

Ugh... A strange consequence of this implicit wrapping is that you can set properties on primitive values with essentially no effect. Since the implicit wrapping produces a new String object each time it occurs, the update to the first wrapper object has no lasting effect.

"hello".someProperty = 17;
"hello".someProperty; // undefined

The wrapping doesn't seem to change the type, is that right?

var str = 'hello';
console.log(typeof str) //=> string
str.toUpperCase()
console.log(typeof str) //=> string
  • Object wrappers for primitive types do not have the same behavior as their primitive values when compared for equality.
  • Getting and setting properties on primitives implicitly creates object wrappers.

Item 5: Avoid using == with Mixed Types

Ugh

"1.0e0" == { valueOf: function() { return true; } };

When the two arguments are of the same type, there's no difference in behavior between == and ===.

Looks like good advice.

So if you know that the arguments are of the same type, they are interchangeable. But using strict equality is a good way to make it clear to readers that there is no conversion involved in the comparison. Otherwise, you require readers to recall the exact coercion rules to decipher your code's behavior.

var date = new Date("1999/12/31");
date == "1999/12/31"; //=> false

But the mistake is symptomatic of a more general misunderstanding of coercions. The == operator does not infer and unify arbitrary data formats. It requires both you and your readers to understand its subtle coercion rules.

  • The == operator applies a confusing set of implicit coercions when its arguments are of different types.
  • Use === to make it clear to your readers that your comparison does not involve any implicit coercions.
  • Use your own explicit coercions when comparing values of different types to make your program's behavior clearer.

Item 6: Learn the Limits of Semicolon Insertion

Um, I'll just use semis so I don't have to know these rules.

Item 7: Think of Strings As Sequences of 16-Bit Code Units

  • JavaScript strings consist of 16-bit code units, not Unicode code points.
  • Unicode code points 216 and above are represented in JavaScript by two code units, known as a surrogate pair.
  • Surrogate pairs throw off string element counts, affecting length, charAt, charCodeAt, and regular expression patterns such as ".".
  • Use third-party libraries for writing code point-aware string manipulation.
  • Whenever you are using a library that works with strings, consult the documentation to see how it handles the full range of code points.

Item 8: Minimize Use of the Global Object

duh

Item 9: Always Declare Local Variables

Watch out for accidental globals

function swap(a, i, j) {
 temp = a[i]; // global
 a[i] = a[j];
 a[j] = temp;
}

Using "use strict" wouln't let the above pass.

Item 10: Avoid with

No problem - didn't even know about it until you mentioned it

Item 11: Get Comfortable with Closures

OK done

Item 12: Understand Variable Hoisting

JavaScript does not support block scoping: Variable definitions are not scoped to their nearest enclosing statement or block, but rather to their containing function.

... the variable is in scope for the entire function, but it is only assigned at the point where the var statement appears.

Hoisting can also lead to confusion about variable redeclaration. It is legal to declare the same variable multiple times within the same function.

Strict doesn't help this runaway loop

"use strict";

for ( var i = 0; i < 10; i++ ) {
  var i = 0;
  console.log( i );
}

The one exception to JavaScript's lack of block scoping is, appropriately enough, exceptions. That is, try..catch binds a caught exception to a variable that is scoped just to the catch block:

  • Variable declarations within a block are implicitly hoisted to the top of their enclosing function.
  • Redeclarations of a variable are treated as a single variable.
  • Consider manually hoisting local variable declarations to avoid confusion.

Item 13: Use Immediately Invoked Function Expressions to Create Local Scopes

  • Closures capture their outer variables by reference, not by value.
  • Use immediately invoked function expressions (IIFEs) to create local scopes.

Item 14: Beware of Unportable Scoping of Named Function Expressions

Use named function for recursion

var cd = function count_down( start ) { //<--- count_down

  if ( start == 0 ) {
    return;
  }
  console.log( start );
  count_down( start - 1 ); //<--- count_down

}

cd(10);

// scoped to function only
console.log(typeof count_down) //=> 'undefined'
console.log(typeof count_down === 'undefined') //=> true

NOTE: you can use the 'cd' var in place of count_down, so we didn't get much for naming the function

Javascript is a mess;

var a;
console.log(a) //=> undefined
console.log(typeof a) //=> undefined (might be 'undefined' depending on how console.log works
console.log(a === undefined) //=> true
console.log(typeof a === undefined) //=> false
console.log(typeof a === 'undefined') //=> true

The real usefulness of named function expressions, though, is for debugging. Most modern JavaScript environments produce stack traces for Error objects, and the name of a function expression is typically used for its entry in a stack trace.

  • Use named function expressions to improve stack traces in Error objects and debuggers.
  • Consider avoiding named function expressions or removing them before shipping.
  • If you are shipping in properly implemented ES5 environments, you've got nothing to worry about.

Item 15: Beware of Unportable Scoping of Block-Local Function Declarations

  • Always keep function declarations at the outermost level of a program or a containing function to avoid unportable behavior.
  • Use var declarations with conditional assignment instead of conditional function declarations.

Item 19: Get Comfortable Using Higher-Order Functions

  • Higher-order functions are functions that take other functions as arguments or return functions as their result.
  • Familiarize yourself with higher-order functions in existing libraries.
  • Learn to detect common coding patterns that can be replaced by higher-order functions.

Item 20: Use call to Call Methods with a Custom Receiver

Note: call() is a method property of functions, not ordinary objects

  • Use the call method to call a function with a custom receiver
  • Use the call method for calling methods that may not exist on a given object.
  • Use the call method for defining higher-order functions that allow clients to provide a receiver for the callback.

Item 21: Use apply to Call Functions with Different Numbers of Arguments

javascripts version of splat

function average(){
  var count = 0;
  for(var i = 0; i<arguments.length; i++){
    count = count + arguments[i];
  }
  return count/arguments.length;
}

var a = [1,2,3,4,5];
console.log(average.apply(null, a)) //=> 3
  • Use the apply method to call variadic functions with a computed array of arguments.
  • Use the first argument of apply to provide a receiver for variadic methods.

Item 22: Use arguments to Create Variadic Functions

  • Use the implicit arguments object to implement variable-arity functions.
  • Consider providing additional fixed-arity versions of the variadic functions you provide so that your consumers don't need to use the apply method.

Item 23: Never Modify the arguments Object

Make real arguments array from existing one;

var args = [].slice.call(arguments);
  • Never modify the arguments object.
  • Copy the arguments object to a real array using [].slice.call(arguments) before modifying it.

Item 24: Use a Variable to Save a Reference to arguments

  • Be aware of the function nesting level when referring to arguments.
  • Bind an explicitly scoped reference to arguments in order to refer to it from nested functions.

Item 25: Use bind to Extract Methods with a Fixed Receiver

var a = [ 1, 2, 3, 4, 5 ];
var s = a.slice;
console.log( s() ) //=> []

var s = a.slice.bind( a );
console.log( s() ) //=> [1,2,3,4,5]

Another example, replacing a callback function definition with a function

var buffer = { data: [], add: function( s ) { this.data.push( s ) }, }

buffer.add( 1 ); buffer.add( 2 ); buffer.add( 3 ); console.log( buffer.data );

var a = ['a', 'b', 'c']

// using a wrapper function for calling buffer.add a.forEach(function(x){ buffer.add(x) })

var a = ['x', 'y', 'z']

// directly calling buffer add, but we need a bind a.forEach(buffer.add.bind(buffer))

console.log( buffer.data ); //=> [ 1, 2, 3, 'a', 'b', 'c', 'x', 'y', 'z' ]

  • Beware that extracting a method does not bind the method's receiver to its object.
  • When passing an object's method to a higher-order function, use an anonymous function to call the method on the appropriate receiver.
  • Use bind as a shorthand for creating a function bound to the appropriate receiver.

Item 26: Use bind to Curry Functions

Eventhough the first arg to bind is 'this', you don't have to use it.

Example of conditioning method to pass to map()

var highlighter = function(pointer, value){
  console.log(pointer + value);
};

// pass wrapper to map
[1,2,3,4,5].map(function(x){
  highlighter('*** ', x);
});

// use bind to convert highlighter into one argument method
[1,2,3,4,5].map(highlighter.bind(null, '--> '));
  • Use bind to curry a function, that is, to create a delegating function with a fixed subset of the required arguments.
  • Pass null or undefined as the receiver argument to curry a function that ignores its receiver.

Item 28: Avoid Relying on the toString Method of Function

  • JavaScript engines are not required to produce accurate reflections of function source code via toString.
  • Never rely on precise details of function source, since different engines may produce different results from toString.
  • The results of toString do not expose the values of local variables stored in a closure.
  • In general, avoid using toString on functions

Item 29: Avoid Nonstandard Stack Inspection Properties

  • Avoid the nonstandard arguments.caller and arguments.callee, because they are not reliably portable.
  • Avoid the nonstandard caller property of functions, because it does not reliably contain complete information about the stack.

Item 30: Understand the Difference between prototype, getPrototypeOf, and proto

Prototypes involve three separate but related accessors, all of which are named with some variation on the word prototype. This unfortunate overlap naturally leads to quite a bit of confusion.

Classes in JavaScript are essentially the combination of a constructor function (User) and a prototype object used to share methods between instances of the class (User.prototype).

  • C.prototype determines the prototype of objects created by new C().
  • Object.getPrototypeOf(obj) is the standard ES5 function for retrieving the prototype of an object.
  • obj.proto is a nonstandard mechanism for retrieving the prototype of an object.
  • A class is a design pattern consisting of a constructor function and an associated prototype.

Item 31: Prefer Object.getPrototypeOf to proto

ES5 introduced Object.getPrototypeOf as the standard API for retrieving an object's prototype, but only after a number of JavaScript engines had long provided the special proto property for the same purpose.

Unreliable in some envs

var empty = Object.create(null); // object with no prototype
"__proto__" in empty; // false (in some environments)
"__proto__" in empty; // true (in some environments)
  • Prefer the standards-compliant Object.getPrototypeOf to the nonstandard proto property.
  • Implement Object.getPrototypeOf in non-ES5 environments that support proto.

Item 32: Never Modify proto

  • Never modify an object's proto property.
  • Use Object.create to provide a custom prototype for new objects.

Item 33: Make Your Constructors new-Agnostic

function User( name, passwordHash ) {
  if ( !( this instanceof User ) ) {
    return new User( name, passwordHash );
  }
  this.name = name;
  this.passwordHash = passwordHash;
}

Version using Object.create()

function User( name, passwordHash ) {
  var self = this instanceof User ? this : Object.create( User.prototype );
  self.name = name;
  self.passwordHash = passwordHash;
  return self;
}

Protecting a constructor against misuse may not always be worth the trouble, especially when you are only using a constructor locally.

  • Make a constructor agnostic to its caller's syntax by reinvoking itself with new or with Object.create.
  • Document clearly when a function expects to be called with new

Item 34: Store Methods on Prototypes

  • Storing methods on instance objects creates multiple copies of the functions, one per instance object.
  • Prefer storing methods on prototypes over storing them on instance objects.

Item 35: Use Closures to Store Private Data

no good example provided by good

My attempt at an example:

function MyConstructor(){
  var data = data;
  return {
    get_data: function(){ return data},
    set_data: function(value){ data = value }
  }
}

var obj = new MyConstructor; obj.set_data(123); console.log(obj.get_data());

_Downside to the above is that the methods need to be defined within the constructor and not on the prototype.

  • Closure variables are private, accessible only to local references.
  • Use local variables as private data to enforce information hiding within methods.

Item 36: Store Instance State Only on Instance Objects

function Tree(x) {
  this.value = x;
}

Tree.prototype = {
  children: [], //<---- should be instance state!!!!
  addChild: function( x ) {
    this.children.push( x );
  }
};
  • Mutable data can be problematic when shared, and prototypes are shared between all their instances.
  • Store mutable per-instance state on instance objects.

Item 37: Recognize the Implicit Binding of this

Every function has an implicit binding of this, whose value is determined when the function is called.

remember that everytime we enter a function, this is assigned to something else, and if we are using 'use strict' its probably being set to undefined.

Bug version, wrong this

function Data( data, splitter ) {
  this._data = data;
  this._splitter = splitter;
}

Data.prototype.split = function() {
  return this._data.map( function( x ) {
    console.log(this) //<-- either global this or undefined because of 'use strict'
    return x.split( this._splitter ) //<-- cannot read property of this
  } )
}

var data = new Data( ['abc','def'], new RegExp( '' ) )
console.log( data.split() );

that this technique

function Data( data, splitter ) {
  this._data = data;
  this._splitter = splitter;
}

Data.prototype.split = function() {
  var that = this //<-- that trick
  return this._data.map( function( x ) {
    return x.split( that._splitter ) //<-- that
  } )
}

var data = new Data( ['abc','def'], new RegExp( '' ) )
console.log( data.split() );

bind technique ( more elegant but difficult to read? )

function Data( data, splitter ) {
  this._data = data;
  this._splitter = splitter;
}

Data.prototype.split = function() {
  return this._data.map( function( x ) {
    return x.split( this._splitter ) //<-- that
  }.bind(this) ) //<-- bind()
}

var data = new Data( ['abc','def'], new RegExp( '' ) )
console.log( data.split() );
  • The scope of this is always determined by its nearest enclosing function.
  • Use a local variable, usually called self, me, or that, to make a this-binding available to inner functions.

Item 38: Call Superclass Constructors from Subclass Constructors

revisit this one, but I don't favor class inheritance in javascript

Item 40: Avoid Inheriting from Standard Classes

The ECMAScript standard library is small, but it comes with a hand- ful of important classes such as Array, Function, and Date. It can be tempting to extend these with subclasses, but unfortunately their definitions have enough special behavior that well-behaved sub- classes are impossible to write.

My limited example of delegation

function MyArray(data){
  this._data = data
}

MyArray.prototype.forEach = function(f){
  this._data.forEach(f);
}

var a = new MyArray([1,2,3]);

a.forEach(function(x){
  console.log(x);
})
  • Inheriting from standard classes tends to break due to special internal properties such as [[Class]].
  • Prefer delegating to properties instead of inheriting from standard classes.

Item 41: Treat Prototypes As an Implementation Detail

  • Objects are interfaces; prototypes are implementations.
  • Avoid inspecting the prototype structure of objects you don’t control.
  • Avoid inspecting properties that implement the internals of objects you don’t control.

Item 42: Avoid Reckless Monkey-Patching

  • Avoid reckless monkey-patching.
  • Document any monkey-patching performed by a library.
  • Consider making monkey-patching optional by performing the modifications in an exported function.
  • Use monkey-patching to provide polyfills for missing standard APIs.

Item 43: Build Lightweight Dictionaries from Direct Instances of Object

At its heart, a JavaScript object is a table mapping string property names to values.

prototype pollution, where properties on a prototype object can cause unexpected properties to appear when enumerating dictionary entries.

  • Use object literals to construct lightweight dictionaries.
  • Lightweight dictionaries should be direct descendants of Object.prototype to protect against prototype pollution in for...in loops.

Item 44: Use null Prototypes to Prevent Prototype Pollution

ES5 offers the first standard way to create an object with no prototype.

var x = Object.create(null);
Object.getPrototypeOf(o) === null; // true

No amount of prototype pollution can affect the behavior of such an object.

  • In ES5, use Object.create(null) to create prototype-free empty objects that are less susceptible to pollution.
  • In older environments, consider using { proto: null }.
  • But beware that proto is neither standard nor entirely portable and may be removed in future JavaScript environments.
  • Never use the name "proto" as a dictionary key since some environments treat this property specially.

Item 45: Use hasOwnProperty to Protect Against Prototype Pollution

This doesn't seem to be a problem. for..in seems to be skipping everything except properties I define on object. Maybe just in chrome and node, which is v8 right? Respecting enumerable properties with for..in loops but still showing with 'in' operator?

var dict = {data:null};
console.log("toString" in dict); // true
console.log("data" in dict); // true

for(var p in dict){ console.log(p) }

//=> 'data' nothing else
  • Use hasOwnProperty to protect against prototype pollution.
  • Use lexical scope and call to protect against overriding of the hasOwnProperty method. *Consider implementing dictionary operations in a class that encapsulates the boilerplate hasOwnProperty tests.
  • Use a dictionary class to protect against the use of "proto" as a key.

Item 46: Prefer Arrays to Dictionaries for Ordered Collections

The ECMAScript standard does not specify any particular order of property storage and is even largely mum on the subject of enumeration.

  • Avoid relying on the order in which for...in loops enumerate object properties.
  • If you aggregate data in a dictionary, make sure the aggregate operations are order-insensitive.
  • Use arrays instead of dictionary objects for ordered collections.

Item 47: Never Add Enumerable Properties to Object.prototype

  • Avoid adding properties to Object.prototype.
  • Consider writing a function instead of an Object.prototype method.
  • If you do add properties to Object.prototype, use ES5's Object.defineProperty to define them as nonenumerable properties.

Item 48: Avoid Modifying an Object during Enumeration

ECMA standard states:

If new properties are added to the object being enumerated during enumeration, the newly added properties are not guaranteed to be visited in the active enumeration.

  • Make sure not to modify an object while enumerating its properties with a for...in loop.
  • Use a while loop or classic for loop instead of a for...in loop when iterating over an object whose contents might change during the loop.
  • For predictable enumeration over a changing data structure, consider using a sequential data structure such as an array instead of a dictionary object.

Item 49: Prefer for Loops to for...in Loops for Array Iteration

  • Always use a for loop rather than a for...in loop for iterating over the indexed properties of an array.
  • Consider storing the length property of an array in a local variable before a loop to avoid recomputing the property lookup.

Item 50: Prefer Iteration Methods to Loops

  • Use iteration methods such as Array.prototype.forEach and Array.prototype.map in place of for loops to make code more readable and avoid duplicating loop control logic.
  • Use custom iteration functions to abstract common loop patterns that are not provided by the standard library.
  • Traditional loops can still be appropriate in cases where early exit is necessary; alternatively, the some and every methods can be used for early exit.

Item 51: Reuse Generic Array Methods on Array-Like Objects

The standard methods of Array.prototype were designed to be reusable as methods of other objects - even objects that do not inherit from Array. As it turns out, a number of such array-like objects crop up in various places in JavaScript.

function my_array() {
  [].forEach.call( arguments, function( x ) {
    console.log( x );
  } )
}

So what exactly makes an object "array-like"? The basic contract of an array object amounts to two simple rules.

  • It has an integer length property in the range 0...232-1.
  • The length property is greater than the largest index of the object. An index is an integer in the range 0...232-2 whose string representation is the key of a property of the object.

Good enough to be array like:

var arrayLike = { 0: "a", 1: "b", 2: "c", length: 3 };

Strings act like immutable arrays, too, since they can be indexed and their length can be accessed as a length property.

var str = 'this is a string';

[].forEach.call(str, function(x){
  console.log(x)
})
  • Reuse generic Array methods on array-like objects by extracting method objects and using their call method.
  • Any object can be used with generic Array methods if it has indexed properties and an appropriate length property.

Item 52: Prefer Array Literals to the Array Constructor

  • The Array constructor behaves differently if its first argument is a number.
  • Use array literals instead of the Array constructor.

Item 54: Treat undefined As "No Value"

The undefined value is special: Whenever JavaScript has no specific value to provide it just produces undefined. Unassigned variables start out with the value undefined

var x;
x; // undefined

var obj = {};
obj.x; // undefined

function f() {
  return;
}
f(); // undefined

function g() {}
g(); // undefined

function f(x) {
  return x;
}
f(); // undefined
  • Avoid using undefined to represent anything other than the absence of a specific value.
  • Use descriptive string values or objects with named boolean properties, rather than undefined or null, to represent application- specific flags.
  • Test for undefined instead of checking arguments.length to provide parameter default values.
  • Never use truthiness tests for parameter default values that should allow 0, NaN, or the empty string as valid arguments.

Item 55: Accept Options Objects for Keyword Arguments

  • Use options objects to make APIs more readable and memorable.
  • The arguments provided by an options object should all be treated as optional.
  • Use an extend utility function to abstract out the logic of extracting values from options objects.

Item 56: Avoid Unnecessary State

  • Prefer stateless APIs where possible. (not sure about this one)
  • When providing stateful APIs, document the relevant state that each operation depends on.

Item 58: Distinguish between Array and Array-Like

  • Never overload structural types with other overlapping types.
  • When overloading a structural type with other types, test for the other types first.
  • Accept true arrays instead of array-like objects when overloading with other object types.
  • Document whether your API accepts true arrays or array-like values.
  • Use ES5's Array.isArray to test for true arrays.

Item 60: Support Method Chaining

  • Use method chaining to combine stateless operations.
  • Support method chaining by designing stateless methods that produce new objects.
  • Support method chaining in stateful methods by returning this.

The Principles of Object-Oriented Javascript

Primitive Types

  • Boolean
  • Number
  • String
  • Null
  • Undefined

In JavaScript, as in many other languages, a variable holding a primitive directly contains the primitive value (rather than a pointer to an object). When you assign a primitive value to a variable, the value is copied into that variable. This means that if you set one variable equal to another, each variable gets its own copy of the data.

var color1 = 'red';
var color2 = color1; //<-- 'red' copied into color2, not a reference
var color1 = 'blue'
console.log(color1, color2) //=> blue red

Identifying Primitive Types

The best way to identify primitive types is with the typeof operator, which works on any variable and returns a string indicating the type of data.

console.log(typeof "Nicholas"); // "string"
console.log(typeof 10); // "number"
console.log(typeof 5.1); // "number"
console.log(typeof true); // "boolean"
console.log(typeof undefined); // "undefined"

Bug:

console.log(typeof null); //=> "object"

The best way to determine if a value is null is to compare it against null directly, like this:

console.log(value === null); // true or false

Primitive Methods

Despite the fact that they're primitive types, strings, numbers, and Booleans actually have methods. (The null and undefined types have no methods.)

Despite the fact that they have methods, primitive values themselves are not objects. JavaScript makes them look like objects to provide a consistent experience in the language, as you'll see later in this chapter.

Reference Types

Unlike primitive types, assigning a reference type to a variable assigns its pointer (maybe that's why they are called reference types?)

var o1 = new Object(); var o2 = o1

console.log(o1 == o2) //=> true console.log(o1 === o2) //=> true

JavaScript is a garbage-collected language, so you don’t really need to worry about memory allocations when you use reference types. However, it’s best to dereference objects that you no longer need so that the garbage collector can free up that memory. The best way to do this is to set the object variable to null.

var object1 = new Object();
// do something
object1 = null;

Instantiate built-in reference types

var items = new Array();
var now = new Date();
var error = new Error("Something bad happened.");
var func = new Function("console.log('Hi');");
var object = new Object();
var re = new RegExp("\\d+");

Example of object literal syntax

var book = {
    "name": "The Principles of Object-Oriented JavaScript",
    "year": 2014
};

Exaple of same with new (constructor form?)

var book = new Object();
book.name = "The Principles of Object-Oriented JavaScript";
book.year = 2014;

Using an object literal doesn’t actually call new Object(). Instead, the JavaScript engine follows the same steps it does when using new Object() without actually calling the constructor. This is true for all reference literals.

Array literal and constructor forms

var colors = [ "red", "blue", "green" ];

var colors = new Array("red", "blue", "green")

You almost always define functions using their literal form.

Not this

var reflect = new Function("value", "return value;");

Regex literal and constructor forms

var numbers = /\d+/g;
var numbers = new RegExp("\\d+", "g");

The literal form of regular expressions in JavaScript is a bit easier to deal with than the constructor form because you don’t need to worry about escaping characters within strings.

Regular expression literals are preferred over the construc- tor form in JavaScript except when the regular expression is being con- structed dynamically from one or more strings.

That said, with the exception of Function, there really isn’t any right or wrong way to instantiate built-in types. Many developers prefer literals, while some prefer constructors. Choose whichever method you find more comfortable to use.

identifying reference Types

var func = function(){}
console.log(typeof func) //=> function

Other reference types are trickier to identify because, for all reference types other than functions, typeof returns "object".

var a = [];
console.log(typeof a) //=> object

The instanceof operator takes an object and a constructor as param- eters. When the value is an instance of the type that the constructor speci- fies, instanceof returns true; otherwise, it returns false

var a = [];
console.log(a instanceof Array) //=> true
console.log(a instanceof Object) //=> true

var b = {};
console.log(b instanceof Object) //=> true

The instanceof operator can identify inherited types. That means every object is actually an instance of Object because every reference type inherits from Object.

Primative Wrapper Types

Perhaps one of the most confusing parts of JavaScript is the concept of primitive wrapper types. There are three primitive wrapper types (String, Number, and Boolean). These special reference types exist to make working with primitive values as easy as working with objects. (It would be very confusing if you had to use a different syntax or switch to a procedural style just to get a substring of text.)

A temporary wrapper object is automatically created for the primative time so it looks like it has methods, but it goes away quickly.

var name = "Nicholas";
var firstChar = name.charAt(0);
console.log(firstChar); //=> "N"

instanceof doesn't work of primatives because the primative wrapper isn't automatically created because its only created when the value is read (or written to?).

var name = "Nicholas";
var count = 10;
var found = false;
console.log(name instanceof String);    // false
console.log(count instanceof Number);   // false
console.log(found instanceof Boolean);  // false

Just creates an object

var name = new String("Nicholas");
var count = new Number(10);
var found = new Boolean(false);
console.log(typeof name); //=> object
console.log(typeof count); //=> object
console.log(typeof found); //=> object

You can create the wrapper by hand but its garbage

var found = new Boolean(false);
if (found) {
  console.log("Found"); //=> Found
}

Manually instantiating primitive wrappers can also be confusing in other ways, so unless you find a special case where it makes sense to do so, you should avoid it. Most of the time, using primitive wrapper objects instead of primitives only leads to errors.

You can use typeof to identify primitive types with the exception of null, which must be compared directly against the special value null.

Function Declarations vs. Expressions

Declaraton

function add(num1, num2) {
 return num1 + num2;
}

Function Expression

var add = function(num1, num2) {
 return num1 + num2;
};

Although these two forms are quite similar, they differ in a very important way. Function declarations are hoisted to the top of the context (either the function in which the declaration occurs or the global scope) when the code is executed. That means you can actually define a function after it is used in code without generating an error.

The arguments object is automatically available inside any function. This means named parameters in a function exist mostly for convenience and don't actually limit the number of arguments that a function can accept.

The arguments object is not an instance of Array and therefore doesn't have the same methods as an array; Array.isArray(arguments) always returns false.

In practice, checking the named parameter against undefined is more common than relying on arguments.length.

var func = function( a, b, c ) {
  if ( undefined == a ) {
    console.log( "a is undefined" )
  }
}

methods that manipulate 'this'

call()

First of three methods that manipulates 'this'

var obj = {
  name: 'kevin'
}

var func = function(greeting) {
  console.log( greeting +" "+ this.name )
}

func.call( obj , "hello") //=> hello kevin
apply()

Second of three methods that manipulates 'this'

var obj = {
  name: 'kevin'
}

var func = function( greeting ) {
  console.log( greeting + " " + this.name )
}

func.apply( obj, [ "hello" ] ) //=> hello kevin
bind()

Third of three methods that manipulates 'this' added in ECMAScript5

bind is like a sticky call(), nerds call it currying

var obj = {
  name: 'kevin'
}

var func = function( greeting ) {
  console.log( greeting + " " + this.name )
}

bind first argument. like call(), its first argument is 'this'

var boundFunc = func.bind(obj)
boundFunc('goodbye') //=> goodbye kevin

bind two arguments

var boundFunc = func.bind(obj, 'adios')
boundFunc() //=> adios kevin

Understanding Objects

Unreliable check for a property on objects

if (person1.age) {
 // do something with age
}

The problem with this pattern is how JavaScript's type coercion affects the outcome. The if condition evaluates to true if the value is truthy (an object, a nonempty string, a nonzero number, or true) and evaluates to false if the value is falsy (null, undefined, 0, false, NaN, or an empty string).

A more reliable way to test for the existence of a property is with the in operator. Like as key?() in ruby.

var person1 = {
  name: "Nicholas",
  sayName: function() {
    console.log( this.name );
  }
};

console.log( "sayName" in person1 ); //=> true

In most cases, the in operator is the best way to determine whether the property exists in an object. It has the added benefit of not evaluating the value of the property, which can be important if such an evaluation is likely to cause a performance issue or an error.

In some cases, however, you might want to check for the existence of a property only if it is an own property. The in operator checks for both own properties and prototype properties, so you'll need to take a different approach.

hasOwnProperty(), which is present on all objects and returns true only if the given property exists and is an own property.

console.log("toString" in person1); // true
console.log(person1.hasOwnProperty("toString")); // false

By default, all properties that you add to an object are enumerable, which means that you can iterate over them using a for-in loop.

var property;
for (property in object) {
  console.log("Name: " + property);
  console.log("Value: " + object[property]);
}

If you just need a list of an object's properties to use later in your program, ECMAScript 5 introduced the Object.keys() method to retrieve an array of enumerable property names

var obj = new Object();
obj.one = 1;
obj.two = 2;
obj.three = 3;

console.log( Object.keys( obj ) ); //=> ['one', 'two', 'three']

There is a difference between the enumerable properties returned in a for-in loop and the ones returned by Object.keys(). The for-in loop also enumerates prototype properties, while Object.keys() returns only own (instance) properties.

Types of Properties

There are two different types of properties: data properties and accessor properties.

kludge accessor property syntax (node the 'get' and 'set')

var myobj = {

  _data: null,

  get data() {
    return this._data;
  },

  set data( data ) {
    this._data = data;
  }

};

myobj.data = [ 'a', 'b', 'c' ];
console.log( myobj.data ) //=> [ 'a', 'b', 'c' ]

You don't need to define both a getter and a setter; you can choose one or both. If you define only a getter, then the property becomes read-only, and attempts to write to it will fail silently in nonstrict mode and throw an error in strict mode. If you define only a setter, then the property becomes write-only, and attempts to read the value will fail silently in both strict and nonstrict modes.

Note, access to myobj._null is still possible even with getters and setters.

defineProperty() with "enumerable"

var myobj = {
  hello: 'there'
};

// change enumerable
console.log( 'hello' in myobj ) //=> true
console.log( Object.keys( myobj ) ) //=> ['hello']
console.log( myobj.propertyIsEnumerable( 'hello' ) ) //=> true

Object.defineProperty( myobj, 'hello', {
  enumerable: false
} )

console.log( 'hello' in myobj ) //=> true
console.log( Object.keys( myobj ) ) //=> []
console.log( myobj.propertyIsEnumerable( 'hello' ) ) //=> false

defineProperty() with "configurable" seems to be a solution in search of a problem?

You can defineProperty along with its data

Object.defineProperty( myobj, 'hello', {
  value: 'there',
  writable: true
} )

console.log(myobj.hello) //=> there

myobj.hello = 'goodbye';
console.log(myobj.hello) //=> goodbye


Object.defineProperty( myobj, 'hello', {
  value: 'there',
  writable: false
} )

myobj.hello = 'adios'; //=> TypeError: Cannot assign to read only property 'hello' of #<Object>

When JavaScript is running in strict mode, attempting to delete a nonconfigurable property results in an error. In nonstrict mode, the operation silently fails.

The advantage of using accessor property attributes instead of object literal notation to define accessor properties is that you can also define those properties on existing objects.

var obj = {
  _data: null
};

Object.defineProperty( obj, 'data', {

  get: function() {
    return this._data;
  },

  set: function( val ) {
    this._data = val;
  },

} );

obj.data = '123';
console.log( obj.data );

Setters and getters defined with defineProperty seem to confuse enumerable:

Without Object.defineProperty:

var obj = {
  _myData: null,

  get: function() {
    return this._myData;
  },

  set: function( val ) {
    this._myData = val;
  },

};

obj.myData = '123';

console.log('myData' in obj) //=> true
console.log(obj.propertyIsEnumerable('myData')) //=> TRUE

With Object.defineProperty:

var obj2 = {_myData:null};

Object.defineProperty( obj2, 'myData', {

  get: function() {
    return this._myData;
  },

  set: function( val ) {
    this._myData = val;
  },

} );

obj2.myData = '123';
console.log( obj2.myData );

console.log('myData' in obj2) //=> true <--- shouldn't this be false?
console.log(obj2.propertyIsEnumerable('myData')) //=> FALSE

Object.defineProperties

var obj = {};

Object.defineProperties( obj, {

  _name: {
    value: 'kevin',
    enumerable: true,
  },

  name: {
    get: function() {
      return this._name
    },
    enumerable: true
  }

} );

Constructors and Prototypes

A constructor is simply a function that is used with new to create an object.

The new operator automatically creates an object of the given type and returns it.

function MyObject() {}
var obj = new MyObject;

Every object instance is automatically created with a constructor property that contains a reference to the constructor function that created it.

console.log( obj instanceof MyObject ); //=> true
console.log( obj.constructor ); //=> [Function: MyObject]
console.log( obj.constructor === MyObject); //=> true

Ugh...

The constructor property can be overwritten and therefore may not be completely accurate.

You can also explicitly call return inside of a constructor. If the returned value is an object, it will be returned instead of the newly created object instance. If the returned value is a primitive, the newly created object is used and the returned value is ignored.

This is good...

An error occurs if you call the a constructor in strict mode without using new. This is because strict mode doesn’t assign this to the global object. Instead, this remains undefined, and an error occurs whenever you attempt to create a property on undefined.

Constructors allow you to configure object instances with the same properties, but constructors alone don’t eliminate code redundancy... prototypes

You can determine whether a property is on the prototype by using a function such as:

function hasPrototypeProperty(object, name) {
  return name in object && !object.hasOwnProperty(name);
}

Reflection example using Object.getPrototypeOf(() and instanceof

function MyConstructor() {}

MyConstructor.prototype = {
  hello: function() {
    console.log( 'hello' )
  }
}

var obj = new MyConstructor;
obj.hello() //=> hello

console.log(Object.getPrototypeOf(obj)) //=> { hello: [Function] }
console.log(obj instanceof MyConstructor) //=> true

A function called with new is:

  1. A constructor.
  2. Creates an object, assigns it to this, and returns it implicitly
  3. Sets the prototype of the created object to the 'constructor property' of the constructor function.

Some JavaScript engines also support a property called proto on all objects. This property allows you to both read from and write to the [[Prototype]] property. Firefox, Safari, Chrome, and Node.js all support this property, and proto is on the path for standardization in ECMAScript 6.

console.log(obj.__proto__) //=> { hello: [Function] }

_proto_ can be assigned to (why does it have to look like crap?)

var MyPrototype = {
  hello: function() {
    console.log( 'Goodbye' )
  }
};

var obj = new Object;
obj.__proto__ = MyPrototype;

obj.hello();

As a prototype property if its the prototype of a particular object

console.log(MyConstructor.prototype.isPrototypeOf(obj)) //=> true

Define prototype before or after?

// before
var MyPrototype = {
  hello: function() {
    console.log( 'Goodbye' )
  }
};

function MyConstructor(){}
MyConstructor.prototype = MyPrototype;

var obj = new MyConstructor();
obj.hello(); //=> goodbye

// after
function MyConstructor(){}

MyConstructor.prototype = {
  hello: function() {
    console.log( 'Goodbye' )
  }
};

var obj = new MyConstructor();
obj.hello(); //=> goodbye

Use a prototype sort of like a class variable:

function MyConstructor() {}

MyConstructor.prototype.myArray = []

var obj1 = new MyConstructor();
var obj2 = new MyConstructor();
var obj3 = new MyConstructor();

obj1.myArray.push( 1 );
obj2.myArray.push( 2 );
obj3.myArray.push( 3 );

console.log( obj1.myArray ); //=> [1,2,3]

Common pattern of defining a prototype by assigning an object literal

function MyConstructor() {}

MyConstructor.prototype = {
  sayHello: function() {
    console.log( 'hello' )
  },
  sayGoodbye: function() {
    console.log( 'goodbye' )
  },
};

var obj = new MyConstructor;
obj.sayHello(); //=> hello
obj.sayGoodbye(); //=> goodbye

The constructor property is actually defined on the prototype because it is shared among object instances.

Assigning object literal to prototype wipes out the constructor property, you might want to put it back.

function MyConstructor() {}

MyConstructor.prototype = {
  constructor: 'MyConstructor', //<---- put it back
  sayHello: function() {
    console.log( 'hello' )
  },
};

var obj = new MyConstructor;

console.log(obj.constructor)

Every function has a prototype property that defines any properties shared by objects created with a particular constructor.

Shared methods and primitive value properties are typically defined on prototypes, while all other properties are defined within the constructor.

The constructor property is actually defined on the prototype because it is shared among object instances. (wiped out by assignment of object literal)

JavaScript's built-in approach for inheritance is called prototype chaining, or prototypal inheritance.

Any object defined via an object literal has its [[Prototype]] set to Object.prototype, meaning that it inherits properties from Object.prototype

The valueOf() method gets called whenever an operator is used on an object. You can always define your own valueOf() method if your objects are intended to be used with operators.

The toString() method is called as a fallback whenever valueOf() returns a reference value instead of a primitive value.

valueOf() seems to override toString even when concating with + (makes no sense)

Douglas Crockford recommends using hasOwnProperty() in for-in loops all the time as a defense against monkey patching:

for (var property in empty) {
 if (empty.hasOwnProperty(property)) {
  console.log(property);
 }
}

Object literals have Object.prototype set as their [[Prototype]] implicitly, but you can also explicitly specify [[Prototype]] with the Object.create() method.

var obj = Object.create(Object.prototype)

// same as

obj = {}

Second param to Object.create takes arguments as Object.defineProperties

var obj = Object.create( Object.prototype, {
  data: {
    value: '123',
    enumerable: true
  }
} )

console.log( obj.data ) //=> 123

Object Inheritance: (made possible with Object.create)

var objA = {
  hello: function() {
    console.log( 'hello' )
  }
}

objA.hello(); //=> hello

var objB = Object.create(objA);

objB.hello(); //=> hello

console.log(objA.isPrototypeOf(objB)) //=> true

Constructor Inheritance: (assign an objects prototype to a constructor)

function ObjA(){} //<-- constructor

ObjA.prototype.hello = function(){
  console.log( 'hello' )
}

function ObjB(){} //<-- constructor

ObjB.prototype = new ObjA; //<--- inheritance

var objB = new ObjB;
objB.hello(); //=> hello

Always make sure that you overwrite the prototype before adding properties to it, or you will lose the added methods when the overwrite happens.

Data hiding with immediate functions (module pattern?)

var obj = ( function() { // immediate function

  var data = null;

  return { // literal object

    get: function() {
      return data;
    },
    set: function( val ) {
      data = val;
    }

  }

} )();


obj.set('123');
console.log(obj.get()); //=> 123

Creating scope safe constructors. i.e. user forgets to use new

function Person( name ) {
  if ( this instanceof Person ) {
    this.name = name;
  } else { // whoops!
    return new Person( name );
  }
}

done!

Speaking Javascript

Semicolons terminate statements, but not blocks. There is one case where you will see a semicolon after a block: a function expression is an expression that ends with a block. If such an expression comes last in a statement, it is followed by a semicolon

var myfunc = function(){};

Roughly, the first character of an identifier can be any Unicode letter, a dollar sign ($), or an underscore (_). Subsequent characters can additionally be any Unicode digit.

var あ = 1;
var び = 2;
console.log(あ + び); //=> 3

All values in JavaScript have properties.

  • The primitive values are booleans, numbers, strings, null, and undefined.
  • All other values are objects.

Each object has a unique identity and is only (strictly) equal to itself.

var obj1 = {};
var obj2 = {};

console.log(obj1 === obj1) //=> true
console.log(obj1 === obj2) //=> false

All primitive values encoding the same value are considered the same

var a = 1;
var b = 1;

console.log(a == b) //=> true
console.log(a === b) //=> true

Primitives have the following characteristics:

  • Compared by value
  • Always immutable

All nonprimitive values are objects

  • Plain objects, which can be created by object literals
  • Arrays, which can be created by array literals
  • Regular expressions, which can be created by regular expression literals
  • Compared by reference
  • Mutable by default

Javascript has two 'nonvalue' objects, undefined and null

  • undefined means 'no value'.
    • Uninitialized vars are undefined.
    • Missing parameters are undefined.
    • If you read an nonexistent property
  • null means "no object"
    • used as a 'nonvalue' whenever an object is expected

There are two operators for categorizing values: typeof is mainly used for primitive values, while instanceof is used for objects.

console.log( typeof 1 ) //=> number
console.log( typeof true ) //=> boolean
console.log( typeof 'string' ) //=> string
console.log( typeof {} ) //=> object
console.log( typeof [] ) //=> object
console.log( typeof undefined ) //=> undefined
console.log( typeof null) //=> object <--- bug!

typeof null returning 'object' is a bug that can’t be fixed, because it would break existing code. It does not mean that null is an object.

console.log( [] instanceof Array ); //=> true
console.log( [] instanceof Object ); //=> true
console.log( {} instanceof Object ); //=> true
console.log( undefined instanceof Object ); //=> false
console.log( null instanceof Object ); //=> false

Falsy

  • undefined
  • null
  • false
  • 0
  • NaN
  • ''

Truthy

  • everything else

Boolean(), called as a function, converts its parameter to a boolean. You can use it to test how a value is interpreted

console.log( Boolean(-0) ); //=> false

All numbers in JS are floats

console.log( 1 === 1.00 ); //=> true

String += same as in ruby

var str = 'a';
str += 'b';
str += 'c';

console.log(str) //=> 'abc'

Loops

for (var i=0; i < arr.length; i++) {
 console.log(arr[i]);
}

var i = 0;
while (i < arr.length) {
 console.log(arr[i]);
 i++;
}

do {
 // ...
} while (condition);

Optional function parameters:

function myFunc(x,y){

  x = x || 0;
  y = y || 0;

}

throw and catch

function myFunc( x, y ) {
  if ( arguments.length < 2 ) {
    throw new Error( "not enough args" );
  }
}

try {
  myFunc();
} catch ( exception ) {
  console.log( exception )
}

The scope of a variable is always the complete function (as opposed to the current block).

Each variable declaration is hoisted: the declaration is moved to the beginning of the function, but assignments that it makes stay put.

Counter using closure:

function counter( start ) {
  return function() {
    return start++;
  }
}

var c = counter( 1 );
console.log( c() ); //=> 1
console.log( c() ); //=> 2
console.log( c() ); //=> 3

Something I dind't notice before, 'in' works differently in differnt contexts:

var obj = {};
obj.one = 1;
obj.two = 2;

console.log(obj); //=> { one: 1, two: 2 }

console.log('one' in obj); //=> returns true in this context

for (var p in obj){ //=> for context
  console.log(p);
}

nothing happens in this context

var p;
while (p in obj){
  console.log(p);
}

undefined or in, your choice

console.log( 'one' in obj ); //=> true
console.log( obj.one !== undefined ); //=> true

Stealing methods

var objA = {
  name: 'objA',
  hello: function(){ console.log("hello I'm " + this.name) }
};

var objB ={name: 'objB'};

objB.hello = objA.hello;
objB.hello(); //=> "hello I'm objB"

Need to bind if not calling on a similar object

var objA = {
  name: 'objA',
  hello: function() {
    console.log( "hello I'm " + this.name )
  }
};

var extracted_func = objA.hello;
extracted_func(); //=> TypeError, cannot read property 'name'

// bind it with 'this'
extracted_func = extracted_func.bind(objA)
extracted_func(); //=> hello I'm objA

"for in" works with arrays, not sure if its the best choice:

var a = [1,2,3,4,5];

for(var p in a){
  console.log(p)
}

Array length can trucate arrays

var a = [1,2,3,4,5];
console.log(a); //=> [1,2,3,4,5]
a.length = 3
console.log(a); //=> [1,2,3]

Iterating over arrays

forEach()

var a = [ 1, 2, 3, 4, 5 ];

a.forEach( function( e, i ) {
  console.log( e, i )
} );

map() exact same thing

var a = [ 1, 2, 3, 4, 5 ];

a.map( function( e, i ) {
  console.log( e, i )
} );

regex

console.log( /b/.test( 'abc' ) ) // true
console.log( /z/.test( 'abc' ) ) // false

var rematch = /(a.*c).*(g.*i)/.exec('abcdefghi')
console.log(rematch[0]) //=> 'abcdefghi'
console.log(rematch[1]) //=> 'abc'
console.log(rematch[2]) //=> 'ghi'

str = '<tag>Content</tag>';

str = str.replace(/<(.*?)>/, '[$1]');
console.log(str); //> "[tag]Content</tag>" <--- only first

str = str.replace(/<(.*?)>/g, '[$1]'); // <--- global
console.log(str); //> "[tag]Content[/tag]" <--- only first

Strict Mode: Recommended, with Caveats

  • Enabling strict mode for existing code may break it
  • Package with care: When you concatenate and/or minify files, you have to be careful that strict mode isn't switched off where it should be switched on or vice versa. Both can break code.

undefined and null are the only values for which any kind of property access results in an exception

var i;
i.myval; //=> TypeError: Cannot read property 'myval' of undefined

var i = undefined;
i.myval //=> TypeError: Cannot read property 'myval' of undefined

var i = null;
i.myval; //=> TypeError: Cannot read property 'myval' of null

var i = {};
i.myval; //=> no error

always undefined (useful in ES3)

if(x === void 0){
  //...
}

Use wrapper objects

var str = new String('abc') // NOT LIKE THIS
var str = String(123)       // like this

var w = Number(formData.width); // like this

The following functions are the preferred way of converting a value to a boolean, number, string, or object (Don't use these as constructors):

Boolean()
Number()
String()

The + operator for arrays is useless, it converts them to strings first:

[1, 2] + [3] //=> '1,23'

There Are No Valid Use Cases for ==

if(x != null) // dont
if(x)         // do

// not sure if x is number or string
if (x == 123) // dont
if (Number(x) === 123) // do

void

void 0 // as a synonym for undefined

// discarding the result of an expression
javascript:void window.open("http://example.com/")

I added the void operator to JS before Netscape 2 shipped to make it easy to discard any non-undefined value in a javascript: URL

Categorizing Values via typeof and instanceof

  • The typeof operator distinguishes primitives from objects and determines the types of primitives.
  • The instanceof operator determines whether an object is an instance of a given constructor

NaN is ridiculous

NaN === NaN // false
typeof NaN // 'number'

String.prototype.slice(start, end?)

'abc'.slice(2) //=> 'c'
'abc'.slice(1, 2) //=> 'b' (2 is end, which is exclusive)

String.prototype.substring(start, end?) Should be avoided in favor of slice(),

String.prototype.trim()

'\r\nabc \t'.trim() //=> 'abc'

String.prototype.concat(str1?, str2?, ...)

String.prototype.toLowerCase()

String.prototype.toUpperCase()

String.prototype.indexOf(searchString, position?)

String.prototype.lastIndexOf(searchString, position?)

String.prototype.search(regexp) //<--- regex

String.prototype.match(regexp) //<--- regex String.prototype.replace(regexp|string, replacement) //<--- regex

Use Error constructor for throwing

if (somethingBadHappened) {
 throw new Error('Something bad happened');
}

Don't forget finally

try{

}catch(e){

console.log(e.stack)

}finally{ // always executed

}

Handling specific errors (from stackexchange)

function SpecificError () { }
SpecificError.prototype = new Error();

// ...

try {
  throw new SpecificError;
} catch (e) {
  if (e instanceof SpecificError) { //<--- instanceof
   // specific error
  } else {
    throw e; // let others bubble up
  }
}

Use the delete operator sparingly. Most modern JavaScript engines optimize the performance of instances created by constructors if their "shape" doesn't change (roughly: no properties are removed or added). Deleting a property prevents that optimization.

Two techniques for overcoming 'this' shadowing

// that = this
var obj = {
  myvar: 'hello',
  func: function() {
    var that = this; // that = this
    var inner = function() {
      log( that.myvar ); //<--- that
    }
    inner()
  }
}

obj.func(); //=> hello

// bind(this)
var obj = {
  myvar: 'hello',
  func: function() {
    var inner = function() {
      log( this.myvar );
    }.bind(this) //<--- secret sauce
    inner()
  }
}

obj.func(); //=> hello

proto is not part of the ECMAScript 5 standard. Therefore, you must not use it if you want your code to conform to that standard and run reliably across current JavaScript engines.

However, more and more engines are adding support for proto and it will be part of ECMAScript 6.

Only getting a property considers the complete prototype chain of an object. Setting and deleting ignores inheritance and affects only own properties. (should this be obvious?)

Setting a property creates an own property, even if there is an inherited property with that key.

Avoid invoking hasOwnProperty on an object because it may have been overridden.

Object.prototype.hasOwnProperty.call(obj, 'foo') // safe
{}.hasOwnProperty.call(obj, 'foo') // shorter true

in works for array indices

var a = [1,2,3];
log( 2 in a ) //=> true
log( 3 in a ) //=> false

delete works for arrays (but not so great)

var a = [ 1, 2, 3 ];
log( a ); //=> [1,2,3]
delete a[ 0 ]
log( a ) //=> [ , 2, 3 ] //<-- what?

use splice to delete array elements

// array.splice(start, count)
var a = [ 1, 2, 3 ];
a.splice(0,1);
log(a); //=> [2,3]

clearing a shared reference to an array danger

var a = [1,2,3];
var b = a;

a = []
log(b); //=> [1,2,3] didn't work

var a = [1,2,3];
var b = a;

a.length = 0
log(b); //=> []

use filter to remove holes from array

var a = [ 0, , , 1, , , 2, , , 3 ]
a = a.filter(function(x){ return true });
log(a); //=> [ 0, 1, 2, 3 ]

JavaScript Patterns

JavaScript uses functions to manage scope.

Every JavaScript environment has a global object accessible when you use this outside of any function.

Every global variable you create becomes a property of the global object.

In browsers, for convenience, there is an additional property of the global object called window that (usually) points to the global object itself.

// in browser console
> myglobal = 'hello';
> "hello"
> window.myglobal
> "hello"
> this.myglobal
> "hello"

...the most important pattern for having fewer globals is to always use var to declare variables. (seems to be the only choice when using 'use strict')

problem with hoisting

var myname = "global"; // global variable
function func() {
  log( myname ); // "undefined" //<--- hoisted up from below, not the global
  var myname = "local";
  log( myname ); // "local"
}
func();

"single var pattern" (not sure if I like this)

var i = 0,
  max = 10,
  array = [];

...substitute i++ with either one of these expressions: JSLint prompts you to do it; the reason being that ++ and -- promote "excessive trickiness."

i = i + 1
i += 1

for-in loops should be used to iterate over nonarray objects. Looping with for-in is also called enumeration.

It's important to use the method hasOwnProperty() when iterating over object properties to filter out properties that come down the prototype chain.

hasOwnProperty filter on enumeration

for ( var i in obj ) {
  if ( obj.hasOwnProperty( i ) ) { // filter
    console.log( i, ":", obj[ i ] );
  }
}

Another pattern for using hasOwnProperty() is to call that method off of the Object.prototype, in case that hasOwnProperty has been redefined (paranoid?, not sure if this is still important with ES5 (maybe for legacy libraries))

for (var i in man) {
  if (Object.prototype.hasOwnProperty.call(man, i)) { // filter
    console.log(i, ":", man[i]);
  }
}

Using parseInt() you can get a numeric value from a string. The function accepts a second radix parameter, which is often omitted but shouldn't be. The problems occur when the string to parse starts with 0: for example, a part of a date entered into a form field. Strings that start with 0 are treated as octal numbers (base 8) in ECMAScript 3;

You have no reason to use the new Object() constructor when you can use an object literal

I don't even comprehend the parameter to new Object(): delegates the constructor?

var o = new Object();
log(o.constructor) //=> [Function: Object]

var o = new Object(1);
log(o.constructor) //=> [Function: Number]

var o = new Object('hi');
log(o.constructor) //=> [Function: String]

Use literal or Object.create (first param is either null or a prototype)

var o = Object.create(null);
log(o.constructor) //=> [Function: Object]

When a constructor is called with new, something like this happens behind the scenes:

// var this = Object.create(Person.prototype);

Example

function MyConstructor(){}

MyConstructor.prototype.hello = function(){
  log('hello');
}

function MyConstructorImposter(){
  var obj = Object.create(MyConstructor.prototype);
  return obj;
}

(new MyConstructor).hello(); //=> hello
(new MyConstructorImposter).hello(); //=> hello

Example guards to prevent calling constructor as function

// check this (this variant only works with ES5) (handles prototype assignment correctly)
// No danger of checking a mispelled Constructor
function Constructor() {
  if ( !this ) throw "forgot consturctor with new"
}

// ignore this (handles prototype assignment correctly)
function Constructor() {
  var that = Object.create( Constructor.prototype );
  return that;
}

// return object literal, no prototype assignment possible
function Constructor() {
  return {
    hello: function() {
      log( 'hello' )
    }
  }
}

// self correcting
function Constructor() {

  // ES3
  if ( !( this instanceof Constructor ) ) {
    return new Constructor;
  }

  // ES5
  if ( ! this  ) {
    return new Constructor;
  }

}

Only reliable way to check for arrayness

Array.isArray([]); // true

Two ways to create a regex

var re = /.*/gmi
var re = RegExp(".*");

As you can see, the regular expression literal notation is shorter and doesn't force you to think in terms of class-like constructors. Therefore it's preferable to use the literal... when using the RegExp() constructor, you also need to escape quotes and often you need to double-escape backslashes

The reason to use new RegExp() is that the pattern is not known in advance but is created as a string at runtime.

Calling RegExp() without new (as a function, not as a constructor) behaves the same as with new.

Function that redefines itself;

var changing = function() {

  log( 'do something' );

  changing = function() {
    log( 'do something else' )
  }

}

changing(); //=> do something
changing(); //=> do something else

Variations of immediate functions

var result = ( function() {
  return 2 + 2;
}() );

// remove outer () because when assignment not needed
var result = function() {
  return 2 + 2;
}();

var result = ( function() {
  return 2 + 2;
} )();


// outer () required when there is no outer assignment
( function() {
  log( 'here' )
} )()

Immediate object initialization pattern

both of these work:

({...}).init();
({...}.init());

Example from book (couldn't you do this with a constructor?)

( {
  // here you can define setting values
  // a.k.a. configuration constants
  maxwidth: 600,
  maxheight: 400,
  // you can also define utility methods
  gimmeMax: function() {
    return this.maxwidth + "x" + this.maxheight;
  },
  // initialize
  init: function() {
    console.log( this.gimmeMax() );
    // more init tasks...
  }
} ).init();

Memoization using a functions property as a cache

function my_add( a, b ) {

  if(my_add.cache === undefined){
    my_add.cache = [];
  }

  var cache_key = a.toString() + ',' + b.toString();
  if ( my_add.cache[ cache_key ] ) {
    return my_add.cache[cache_key]
  }

  var result = a + b;
  my_add.cache[cache_key] = result;

  return result;

}

The above could have used the json string of arguments as a cache key

 var cachekey = JSON.stringify(Array.prototype.slice.call(arguments)),
Configuration Objects (umm.. hash passed as argument?)
var conf = {
 username: "batman",
 first: "Bruce",
 last: "Wayne"
};
addPerson(conf);

The pros of the configuration objects are:

  • No need to remember the parameters and their order
  • You can safely skip optional parameters
  • Easier to read and maintain
  • Easier to add and remove parameters The cons of the configuration objects are:
  • You need to remember the names of the parameters
  • Property names cannot be minified
Namespace Pattern
var MYAPP = {};
MYAPP.Parent = function () {};
MYAPP.some_var = 1;
MYAPP.modules = {};
MYAPP.modules.module1 = {};

Checking if namespace exists (maybe at time of library file)

var MYAPP = MYAPP || {};

Build a namespace function that will autovivify, or maybe one exists already

Declaring Dependencies

It's a good idea to declare the modules your code relies on at the top of your function or module. The declaration involves creating only a local variable and pointing to the desired module: _(I don't know about doing this at function level)

var myFunction = function () {
  var event = YAHOO.util.Event,
  dom = YAHOO.util.Dom;
  //... 
};

Private Members are easy, use closure

function MyConstructor( data ) {

  var data = data;

  this.get_data = function() {
    return data
  }

}

var obj = new MyConstructor(123);
log(obj.get_data()); //=> 123

Warning: if you return a reference (like an array, object) then that can be modified outside the object - instead return a clone.

To create private data with an object literal you need to wrap with immediate function.

var obj = ( function() {

  var data = 123;

  return {
    get_data: function(){ return data }
  }

} )();

log(obj.get_data()); //=> 123

My example of implementing class variable

function MyConstructor() {}

MyConstructor.prototype = ( function() {

  var count = 0;

  return {
    increment: function() {
      log("incrementing " + count);
      count++
    },
    get_count: function() {
      return count
    }
  }

} )()


for ( var i = 0, a = []; i < 5; i++ ) {
  ( new MyConstructor ).increment();
}

log( ( new MyConstructor ).get_count() )
Revelation Pattern (paranoid much?)
var myParanoidObj = ( function() {

  function func1() { log( 'in func1' ) }

  function func2() { func1() } // calls func1

  return {
    func1: func1,
    func2: func2,
  }

} )()

myParanoidObj.func1(); //=> in func1
myParanoidObj.func2(); //=> in func1

myParanoidObj.func1 = null
myParanoidObj.func1(); //=> TypeError not a function
myParanoidObj.func2(); //=> in func1 (still works!)
Module Pattern
  • Namespaces
  • Immediate functions
  • Private and privileged members
  • Declaring dependencies

Create namespace

MYAPP = {};
MYAPP.utilities = {}
MYAPP.utilities.array = {}

Define module and public methods

MYAPP.utilities.array = ( function() {

  // methods
  inArray: function( needle, haystack ) { };
  isArray: function( a ) { };

  // reveal public methods
  return {
    inArray: inArray,
    isArray: isArray,
  };

}() );

Constructor version:

var MYAPP = {};
MYAPP.utilities = {};

MYAPP.utilities.Array = ( function() {

  // ... private methods and data

  function Constructor() {};

  Constructor.prototype = {
    inArray: function( needle, haystack ) {},
    isArray: function( a ) {},
  }

  return Constructor;

}() );


var obj = new MYAPP.utilities.Array
Public Static Method
var MyConstructor = function(){}
MyConstructor.static = function(){ log("I'm static") }
MyConstructor.static()
Private static data
var MyConstructor = ( function() {

  var counter = 0;

  return function(){ // this is what gets 'new'
    log(counter++)
  }

} )()

new MyConstructor; //=> 1
new MyConstructor; //=> 2
new MyConstructor; //=> 3
Private static method
var MyConstructor = ( function() {

  var counter = 0;
  var InternalConstructor = function(){
    counter = counter + 1;
    log(counter) 
  };

  InternalConstructor.prototype.getCount = function() {
    return counter
  };

  return InternalConstructor;

} )()

new MyConstructor; //=> 1
new MyConstructor; //=> 2
new MyConstructor; //=> 3

var obj = new MyConstructor; //=> 4
log( obj.getCount() ) //=> 4
Chaining methods pattern

When you create methods that have no meaningful return value, you can have them return this, the instance of the object they are working with. This will enable consumers of that object to call the next method chained to the previous:

var Constructor = function() {

  this.count = 0;

  this.add = function( x ) {
    this.count = this.count + x;
    return this;
  };

  this.get = function() {
    return this.count
  }

}

var o = new Constructor;
log( o.add( 1 ).add( 5 ).get() ) //=> 6

In ECMAScript 5, the prototypal inheritance pattern becomes officially a part of the language. This pattern is implemented through the method Object.create().

var child = Object.create(parent);

Design Patterns

Singleton

My attempt at a singleton:

var Universe = ( function() {

  var cache;

  return function() { // actual constructor

    if ( cache ) {
      return cache;
    } else {
      cache = this;
    }

  }

} )()

log( new Universe === new Universe ) //=> true
Factory

too easy

Built in object factor
var o = new Object(),
    n = new Object(1),
    s = Object('1'),
    b = Object(true);

o.constructor === Object; // true
n.constructor === Number; // true
s.constructor === String; // true
b.constructor === Boolean; // true

The fact that Object() is also a factory is of little practical use, just something worth mentioning as an example that the factory pattern is all around us

Revisit the following patterns

Figure out simplier implementations then the tedious ones offered in this book.

Decorator Strategy Facade Proxy Mediator Observer

Skipping Dom Stuff, better left for other books

Ember Guides and Tutorials

Ember's switch to ember-cli has broken pretty much all documentation and tutorials on the entire internet, trying to slog through this mess

Setup

much of the configs comes from here

http://ember.vicramon.com/creating-the-rails-api

Rails Backend Steps

gem update
rails new backend --database=mysql --skip-turbolinks 
cd backend

# in Gemfile
gem 'active_model_serializers'
gem 'ffaker' # in development group
bundle update

rails g model lead
rails g serializer lead
rails g controller api/v1/leads

Ember Frontend Steps

npm update -g
ember new frontend --skip-git
cd frontend
ember install:addon ember-cli-emblem-hbs-printer

# in .ember-cli (I don't know why this wasn't necessary this last time)
"liveReload": true,
"watcher": "polling"

The Object Model

Classes and Instances

To define a new Ember class, use extend(), to create object from that class use create()

var app_person = Ember.Object.extend( {
  say: function( thing ) {
    alert( thing );
  }
} );

app_person.create().say( 'something' ); //=> box 'something'

You can create a subclass of an existing class with extend

var myClass = Ember.Object.extend( {
  say: function( thing ) {
    alert( thing );
  }
} );

var mySubClass = myClass.extend();
mySubClass.create().say('something else'); //=> box 'something else'

Can access the superclass with this._super

var myClass = Ember.Object.extend( {
  say: function( thing ) {
    alert( 'myClass' );
  }
} );

var mySubClass = myClass.extend({
  say: function( thing ) {
    this._super(); //<---- HERE
    alert( 'mySubClass' );
  }
});

mySubClass.create().say('something else'); //=> boxs 'myClass'...'mySubClass'

Initialize new object by passing in a 'hash'

var myClass = Ember.Object.extend( {
  name: '',
  say: function( thing ) {
    alert( this.get('name') );
  }
} );

var myObj = myClass.create({name:'kevin'}); 
myObj.say();

For performance reasons, note that you cannot redefine an instance's computed properties or methods when calling create(), nor can you define new ones. You should only set simple properties when calling create(). If you need to define or redefine methods or computed properties, create a new subclass and instantiate that.

If you are subclassing a framework class, like Ember.View or Ember.ArrayController, and you override the init method, make sure you call this._super()! If you don't, the system may not have an opportunity to do important setup work, and you'll see strange behavior in your application.

var myClass = Ember.Object.extend( {
  init: function(){
    alert('init');
  },
  say: function( thing ) {
    alert( this.get('name') );
  }
} );

var myObj = myClass.create({name:'kevin'}); 
myObj.say();

When accessing the properties of an object, use the get and set accessor methods. Make sure to use these accessor methods; otherwise, computed properties won't recalculate, observers won't fire, and templates won't update.

Computed Properties

This example isn't that great because it doesn't show templates and other computed properties being updated. This example also shows chaining. Note that .property() method on function prototype makes function accessible with get('myfunction') and also wires up functin to be dynamic when properties it depends on change.

var myClass = Ember.Object.extend( {
  firstName: 'bob',
  lastName: 'smith',
  age: 45,
  fullName: function() {
    return this.get( 'firstName' ) + ' ' + this.get( 'lastName' );
  }.property( 'firstName', 'lastName' ),
  description: function() {
    return [ this.get( 'fullName' ), this.get( 'age' ) ].join( ' ' );
  }.property( 'firstName', 'lastName', 'age' ),
} );

var obj = myClass.create();
log( obj.get( 'description' ) ); //=> bob smith 45

Ember will call the computed property for both setters and getters, so if you want to use a computed property as a setter, you'll need to check the number of arguments to determine whether it is being called as a getter or a setter. Note that if a value is returned from the setter, it will be cached as the property's value.

App.Person = Ember.Object.extend({
  firstName: null,
  lastName: null,

  fullName: function(key, value) {
    // setter
    if (arguments.length > 1) { //<--- HERE
      var nameParts = value.split(/\s+/);
      this.set('firstName', nameParts[0]);
      this.set('lastName',  nameParts[1]);
    }

    // getter
    return this.get('firstName') + ' ' + this.get('lastName');
  }.property('firstName', 'lastName')
});


var captainAmerica = App.Person.create();
captainAmerica.set('fullName', "William Burnside");
captainAmerica.get('firstName'); // William
captainAmerica.get('lastName'); // Burnside

Observers

You Don't Know JS

this is actually a binding that is made when a function is invoked, and what it references is determined entirely by the call-site where the function is called.

  • Called with new? Use the newly constructed object.
  • Called with call or apply (or bind)? Use the specified object.
  • Called with a context object owning the call? Use that context object.
  • Default: undefined in strict mode, global object otherwise.

Explicit this binding

With respect to this binding, call(..) and apply(..) are identical. They do behave differently with their additional parameters, but that’s not something we care about presently.

Hard this binding in ES5

var bar = foo.bind( obj );

Many libraries’ functions, and indeed many new built-in functions in the JavaScript language and host environment, provide an optional parameter, usually called “context,” which is designed as a work- around for you not having to use bind(..) to ensure your callback function uses a particular this.

[1, 2, 3].forEach( foo, obj );

New this binding

When a function is invoked with new in front of it, otherwise known as a constructor call, the following things are done automatically:

  1. A brand new object is created (aka constructed) out of thin air.
  2. The newly constructed object is [[Prototype]]-linked.
  3. The newly constructed object is set as the this binding for that function call.
  4. Unless the function returns its own alternate object, the new invoked function call will automatically return the newly constructed object.

Create DMZ object for passing 'null' to call, apply, bind

This can prevent libraries from accidentally messing with global this. ø is typed with opt-o on mac

var ø = Object.create( null );
var bar = foo.bind( ø, 2 );

get and set

var obj = {
  get x() { return this._x },
  set x( x ) { this._x = x }
}