Table of Contents

• Source Code Layout
• Syntax
• Naming
• Comments
  • Comment Annotations
  • Magic Comments
• Classes & Modules
• Exceptions
• Collections
• Numbers
• Strings
• Date & Time
• Regular Expressions
• Percent Literals
• Metaprogramming
• Misc
• Tools

Source Code Layout

• Use UTF-8 as the source file encoding. ^[link]

• Use two spaces per indentation level (aka soft tabs). No hard tabs. ^[link]

  # bad - four spaces
  def some_method
      do_something
  end
  
  # good
  def some_method
    do_something
  end

• Use Unix-style line endings. (*BSD/Solaris/Linux/macOS users are covered by default, Windows users have to be extra careful.) ^[link]

  • If you're using Git you might want to add the following configuration setting to protect your project from Windows line endings creeping in:

    $ git config --global core.autocrlf true

• Don't use ; to separate statements and expressions. As a corollary—use one expression per line. ^[link]

  # bad
  puts 'foobar'; # superfluous semicolon
  
  puts 'foo'; puts 'bar' # two expressions on the same line
  
  # good
  puts 'foobar'
  
  puts 'foo'
  puts 'bar'
  
  puts 'foo', 'bar' # this applies to puts in particular

• Prefer a single-line format for class definitions with no body. ^[link]

  # bad
  class FooError < StandardError
  end
  
  # good
  FooError = Class.new(StandardError)

• Avoid single-line methods. Although they are somewhat popular in the wild, there are a few peculiarities about their definition syntax that make their use undesirable. At any rate—there should be no more than one expression in a single-line method. ^[link]

  # bad
  def too_much; something; something_else; end
  
  # good
  def some_method
    body
  end

  One exception to the rule are empty-body methods.

  # good
  def no_op; end

• Use spaces around operators, after commas, colons and semicolons. Whitespace might be (mostly) irrelevant to the Ruby interpreter, but its proper use is the key to writing easily readable code. ^[link]

  sum = 1 + 2
  a, b = 1, 2
  class FooError < StandardError; end

  There are a few exceptions. One is the exponent operator:

  # bad
  e = M * c ** 2
  
  # good
  e = M * c**2

  Another exception is the slash in rational literals:

  # bad
  o_scale = 1 / 48r
  
  # good
  o_scale = 1/48r

  Another exception is the safe navigation operator:

  # bad
  foo &. bar
  foo &.bar
  foo&. bar
  
  # good
  foo&.bar

• No spaces after (, [ or before ], ). Use spaces around { and before }. ^[link]

  # bad
  some( arg ).other
  [ 1, 2, 3 ].each{|e| puts e}
  
  # good
  some(arg).other
  [1, 2, 3].each { |e| puts e }

  { and } deserve a bit of clarification, since they are used for block and hash literals, as well as string interpolation.

  # good - space after { and before }
  { one: 1, two: 2 }

  With interpolated expressions, there should be no padded-spacing inside the braces.

  # bad
  "From: #{ user.first_name }, #{ user.last_name }"
  
  # good
  "From: #{user.first_name}, #{user.last_name}"

• No space after !. ^[link]

  # bad
  ! something
  
  # good
  !something

• No space inside range literals. ^[link]

  # bad
  1 .. 3
  'a' ... 'z'
  
  # good
  1..3
  'a'...'z'

• Indent when as deep as case. This is the style established in both "The Ruby Programming Language" and "Programming Ruby". Historically it is derived from the fact that case and switch statements are not blocks, hence should not be indented, and the when and else keywords are labels (compiled in the C language, they are literally labels for JMP calls). ^[link]

  # bad
  case
    when song.name == 'Misty'
      puts 'Not again!'
    when song.duration > 120
      puts 'Too long!'
    when Time.now.hour > 21
      puts "It's too late"
    else
      song.play
  end
  
  # good
  case
  when song.name == 'Misty'
    puts 'Not again!'
  when song.duration > 120
    puts 'Too long!'
  when Time.now.hour > 21
    puts "It's too late"
  else
    song.play
  end

• When assigning the result of a conditional expression to a variable, preserve the usual alignment of its branches. ^[link]

  # bad - pretty convoluted
  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'
  when 1940..1950 then 'Bebop'
  else 'Jazz'
  end
  
  result = if some_cond
    calc_something
  else
    calc_something_else
  end
  
  # good - it's apparent what's going on
  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'
         when 1940..1950 then 'Bebop'
         else 'Jazz'
         end
  
  result = if some_cond
             calc_something
           else
             calc_something_else
           end

• Use empty lines between method definitions and also to break up methods into logical paragraphs internally. ^[link]

  def some_method
    data = initialize(options)
  
    data.manipulate!
  
    data.result
  end
  
  def some_method
    result
  end

• Don't use several empty lines in a row. ^[link]

  # bad - It has two empty lines.
  some_method
  
  
  some_method
  
  # good
  some_method
  
  some_method

• Use empty lines around access modifiers. ^[link]

  # bad
  class Foo
    attr_reader :foo
    def foo
      # do something...
    end
  end
  
  # good
  class Foo
    attr_reader :foo
  
    def foo
      # do something...
    end
  end

• Don't use empty lines around method, class, module, block bodies. ^[link]

  # bad
  class Foo
  
    def foo
  
      begin
  
        do_something do
  
          something
  
        end
  
      rescue
  
        something
  
      end
  
    end
  
  end
  
  # good
  class Foo
    def foo
      begin
        do_something do
          something
        end
      rescue
        something
      end
    end
  end

• Avoid comma after the last parameter in a method call, especially when the parameters are not on separate lines. ^[link]

  # bad - easier to move/add/remove parameters, but still not preferred
  some_method(
    size,
    count,
    color,
  )
  
  # bad
  some_method(size, count, color, )
  
  # good
  some_method(size, count, color)

• Use spaces around the = operator when assigning default values to method parameters: ^[link]

  # bad
  def some_method(arg1=:default, arg2=nil, arg3=[])
    # do something...
  end
  
  # good
  def some_method(arg1 = :default, arg2 = nil, arg3 = [])
    # do something...
  end

  While several Ruby books suggest the first style, the second is much more prominent in practice (and arguably a bit more readable).

• Avoid line continuation \ where not required. In practice, avoid using line continuations for anything but string concatenation. ^[link]

  # bad
  result = 1 - \
           2
  
  # good (but still ugly as hell)
  result = 1 \
           - 2
  
  long_string = 'First part of the long string' \
                ' and second part of the long string'

• Adopt a consistent multi-line method chaining style. ^[link]

  • When continuing a chained method invocation on another line, include the . on the first line to indicate that the expression continues.

    # bad - need to read ahead to the second line to know that the chain continues
    one.two.three
      .four
    
    # good - it's immediately clear that the expression continues beyond the first line
    one.two.three.
      four

• Align the parameters of a method call if they span more than one line. When aligning parameters is not appropriate due to line-length constraints, single indent for the lines after the first is also acceptable. ^[link]

  # starting point (line is too long)
  def send_mail(source)
    Mailer.deliver(to: '[email protected]', from: '[email protected]', subject: 'Important message', body: source.text)
  end
  
  # bad (double indent)
  def send_mail(source)
    Mailer.deliver(
        to: '[email protected]',
        from: '[email protected]',
        subject: 'Important message',
        body: source.text)
  end
  
  # good
  def send_mail(source)
    Mailer.deliver(to: '[email protected]',
                   from: '[email protected]',
                   subject: 'Important message',
                   body: source.text)
  end

• Align the elements of array literals spanning multiple lines. ^[link]

  # bad - single indent
  menu_item = %w[Spam Spam Spam Spam Spam Spam Spam Spam
    Baked beans Spam Spam Spam Spam Spam]
  
  # good
  menu_item = %w[
    Spam Spam Spam Spam Spam Spam Spam Spam
    Baked beans Spam Spam Spam Spam Spam
  ]

• Add underscores to large numeric literals to improve their readability. ^[link]

  # bad - how many 0s are there?
  num = 1000000
  
  # good - much easier to parse for the human brain
  num = 1_000_000

• Prefer lowercase letters for numeric literal prefixes. 0o for octal, 0x for hexadecimal and 0b for binary. Do not use 0d prefix for decimal literals. ^[link]

  # bad
  num = 01234
  num = 0O1234
  num = 0X12AB
  num = 0B10101
  num = 0D1234
  num = 0d1234
  
  # good - easier to separate digits from the prefix
  num = 0o1234
  num = 0x12AB
  num = 0b10101
  num = 1234

• Use [YARD][yard] and its conventions for API documentation. ^[link]

• Limit lines to 80 characters. ^[link]

• Avoid trailing whitespace. ^[link]

• End each file with a newline. ^[link]

• Don't use block comments. They cannot be preceded by whitespace and are not as easy to spot as regular comments. ^[link]

  # bad
  =begin
  comment line
  another comment line
  =end
  
  # good
  # comment line
  # another comment line

Syntax

• Use :: only to reference constants (this includes classes and modules) and constructors (like Array() or Nokogiri::HTML()).

  • Do not use :: for regular method invocation. ^[link]

    # bad
    SomeClass::some_method
    some_object::some_method
    
    # good
    SomeClass.some_method
    some_object.some_method
    SomeModule::SomeClass::SOME_CONST
    SomeModule::SomeClass()

  • Do not use :: to define class methods. ^[link]

    # bad
    class Foo
      def self::some_method
      end
    end
    
    # good
    class Foo
      def self.some_method
      end
    end

• Use def with parentheses when there are parameters. Omit the parentheses when the method doesn't accept any parameters. ^[link]

  # bad
  def some_method()
    # body omitted
  end
  
  # good
  def some_method
    # body omitted
  end
  
  # bad
  def some_method_with_parameters param1, param2
    # body omitted
  end
  
  # good
  def some_method_with_parameters(param1, param2)
    # body omitted
  end

• Use parentheses around the arguments of method invocations, especially if the first argument begins with an open parenthesis (, as in f((3 + 2) + 1). ^[link]

  # bad
  x = Math.sin y
  # good
  x = Math.sin(y)
  
  # bad
  array.delete e
  # good
  array.delete(e)
  
  # bad
  temperance = Person.new 'Temperance', 30
  # good
  temperance = Person.new('Temperance', 30)

  Always omit parentheses for

  • Method calls with no arguments:

    # bad
    Kernel.exit!()
    2.even?()
    fork()
    'test'.upcase()
    
    # good
    Kernel.exit!
    2.even?
    fork
    'test'.upcase

  • Methods that are part of an internal DSL (e.g., Rake, Rails, RSpec):

    # bad
    validates(:name, presence: true)
    # good
    validates :name, presence: true

  • Methods that have "keyword" status in Ruby:

    class Person
      # bad
      attr_reader(:name, :age)
      # good
      attr_reader :name, :age
    
      # body omitted
    end

  Can omit parentheses for

  • Methods that have "keyword" status in Ruby, but are not declarative:

    # good
    puts(temperance.age)
    system('ls')
    # also good
    puts temperance.age
    system 'ls'

• Define optional arguments at the end of the list of arguments. Ruby has some unexpected results when calling methods that have optional arguments at the front of the list. ^[link]

  # bad
  def some_method(a = 1, b = 2, c, d)
    puts "#{a}, #{b}, #{c}, #{d}"
  end
  
  some_method('w', 'x') # => '1, 2, w, x'
  some_method('w', 'x', 'y') # => 'w, 2, x, y'
  some_method('w', 'x', 'y', 'z') # => 'w, x, y, z'
  
  # good
  def some_method(c, d, a = 1, b = 2)
    puts "#{a}, #{b}, #{c}, #{d}"
  end
  
  some_method('w', 'x') # => '1, 2, w, x'
  some_method('w', 'x', 'y') # => 'y, 2, w, x'
  some_method('w', 'x', 'y', 'z') # => 'y, z, w, x'

• Use keyword arguments when passing boolean argument to a method. ^[link]

  # bad
  def some_method(bar = false)
    puts bar
  end
  
  # bad - common hack before keyword args were introduced
  def some_method(options = {})
    bar = options.fetch(:bar, false)
    puts bar
  end
  
  # good
  def some_method(bar: false)
    puts bar
  end
  
  some_method            # => false
  some_method(bar: true) # => true

• Prefer keyword arguments over optional arguments. ^[link]

  # bad
  def some_method(a, b = 5, c = 1)
    # body omitted
  end
  
  # good
  def some_method(a, b: 5, c: 1)
    # body omitted
  end

• Use keyword arguments instead of option hashes. ^[link]

  # bad
  def some_method(options = {})
    bar = options.fetch(:bar, false)
    puts bar
  end
  
  # good
  def some_method(bar: false)
    puts bar
  end

• Avoid the use of parallel assignment for defining variables. Parallel assignment is allowed when it is the return of a method call, used with the splat operator, or when used to swap variable assignment. Parallel assignment is less readable than separate assignment. ^[link]

  # bad
  a, b, c, d = 'foo', 'bar', 'baz', 'foobar'
  
  # good
  a = 'foo'
  b = 'bar'
  c = 'baz'
  d = 'foobar'
  
  # good - swapping variable assignment
  # Swapping variable assignment is a special case because it will allow you to
  # swap the values that are assigned to each variable.
  a = 'foo'
  b = 'bar'
  
  a, b = b, a
  puts a # => 'bar'
  puts b # => 'foo'
  
  # good - method return
  def multi_return
    [1, 2]
  end
  
  first, second = multi_return
  
  # good - use with splat
  first, *list = [1, 2, 3, 4] # first => 1, list => [2, 3, 4]
  
  hello_array = *'Hello' # => ["Hello"]
  
  a = *(1..3) # => [1, 2, 3]

• Avoid the use of unnecessary trailing underscore variables during parallel assignment. Named underscore variables are to be preferred over underscore variables because of the context that they provide. Trailing underscore variables are necessary when there is a splat variable defined on the left side of the assignment, and the splat variable is not an underscore. ^[link]

  # bad
  foo = 'one,two,three,four,five'
  # Unnecessary assignment that does not provide useful information
  first, second, _ = foo.split(',')
  first, _, _ = foo.split(',')
  first, *_ = foo.split(',')
  
  
  # good
  foo = 'one,two,three,four,five'
  # The underscores are needed to show that you want all elements
  # except for the last number of underscore elements
  *beginning, _ = foo.split(',')
  *beginning, something, _ = foo.split(',')
  
  a, = foo.split(',')
  a, b, = foo.split(',')
  # Unnecessary assignment to an unused variable, but the assignment
  # provides us with useful information.
  first, _second = foo.split(',')
  first, _second, = foo.split(',')
  first, *_ending = foo.split(',')

• Do not use for, unless you know exactly why. Most of the time iterators should be used instead. for is implemented in terms of each (so you're adding a level of indirection), but with a twist—for doesn't introduce a new scope (unlike each) and variables defined in its block will be visible outside it. ^[link]

  arr = [1, 2, 3]
  
  # bad
  for elem in arr do
    puts elem
  end
  
  # note that elem is accessible outside of the for loop
  elem # => 3
  
  # good
  arr.each { |elem| puts elem }
  
  # elem is not accessible outside each's block
  elem # => NameError: undefined local variable or method `elem'

• Do not use then for multi-line if/unless. ^[link]

  # bad
  if some_condition then
    # body omitted
  end
  
  # good
  if some_condition
    # body omitted
  end

• Always put the condition on the same line as the if/unless in a multi-line conditional. ^[link]

  # bad
  if
    some_condition
    do_something
    do_something_else
  end
  
  # good
  if some_condition
    do_something
    do_something_else
  end

• Favor the ternary operator(?:) over if/then/else/end constructs. It's more common and obviously more concise. ^[link]

  # bad
  result = if some_condition then something else something_else end
  
  # good
  result = some_condition ? something : something_else

• Use one expression per branch in a ternary operator. This also means that ternary operators must not be nested. Prefer if/else constructs in these cases. ^[link]

  # bad
  some_condition ? (nested_condition ? nested_something : nested_something_else) : something_else
  
  # good
  if some_condition
    nested_condition ? nested_something : nested_something_else
  else
    something_else
  end

• Do not use if x; .... Use the ternary operator instead. ^[link]

  # bad
  result = if some_condition; something else something_else end
  
  # good
  result = some_condition ? something : something_else

• Leverage the fact that if and case are expressions which return a result. ^[link]

  # bad
  if condition
    result = x
  else
    result = y
  end
  
  # good
  result =
    if condition
      x
    else
      y
    end

• Use when x then ... for one-line cases. The alternative syntax when x: ... has been removed as of Ruby 1.9. ^[link]

• Do not use when x; .... See the previous rule. ^[link]

• Use ! instead of not. ^[link]

  # bad - parentheses are required because of op precedence
  x = (not something)
  
  # good
  x = !something

• Avoid the use of !!. ^[link]

  !! converts a value to boolean, but you don't need this explicit conversion in the condition of a control expression; using it only obscures your intention. If you want to do a nil check, use nil? instead.

  # bad
  x = 'test'
  # obscure nil check
  if !!x
    # body omitted
  end
  
  # good
  x = 'test'
  if x
    # body omitted
  end

• The and and or keywords are banned. The minimal added readability is just not worth the high probability of introducing subtle bugs. For boolean expressions, always use && and || instead. For flow control, use if and unless; && and || are also acceptable but less clear. ^[link]

  # bad
  # boolean expression
  ok = got_needed_arguments and arguments_are_valid
  
  # control flow
  document.save or raise("Failed to save document!")
  
  # good
  # boolean expression
  ok = got_needed_arguments && arguments_are_valid
  
  # control flow
  raise("Failed to save document!") unless document.save
  
  # ok
  # control flow
  document.save || raise("Failed to save document!")

• Avoid multi-line ?: (the ternary operator); use if/unless instead. ^[link]

• Favor modifier if/unless usage when you have a single-line body. Another good alternative is the usage of control flow &&/||. ^[link]

  # bad
  if some_condition
    do_something
  end
  
  # good
  do_something if some_condition

• Avoid modifier if/unless usage at the end of a non-trivial multi-line block. ^[link]

  # bad
  10.times do
    # multi-line body omitted
  end if some_condition
  
  # good
  if some_condition
    10.times do
      # multi-line body omitted
    end
  end

• Avoid nested modifier if/unless/while/until usage. Favor &&/|| if appropriate. ^[link]

  # bad
  do_something if other_condition if some_condition
  
  # good
  do_something if some_condition && other_condition

• Do not use unless with else. Rewrite these with the positive case first. ^[link]

  # bad
  unless success?
    puts 'failure'
  else
    puts 'success'
  end
  
  # good
  if success?
    puts 'success'
  else
    puts 'failure'
  end

• Don't use parentheses around the condition of a control expression. ^[link]

  # bad
  if (x > 10)
    # body omitted
  end
  
  # good
  if x > 10
    # body omitted
  end

  Note that there is an exception to this rule, namely safe assignment in condition.

• Do not use while/until condition do for multi-line while/until. ^[link]

  # bad
  while x > 5 do
    # body omitted
  end
  
  until x > 5 do
    # body omitted
  end
  
  # good
  while x > 5
    # body omitted
  end
  
  until x > 5
    # body omitted
  end

• Favor modifier while/until usage when you have a single-line body. ^[link]

  # bad
  while some_condition
    do_something
  end
  
  # good
  do_something while some_condition

• Use Kernel#loop instead of while/until when you need an infinite loop. ^[link]

  # bad
  while true
    do_something
  end
  
  until false
    do_something
  end
  
  # good
  loop do
    do_something
  end

• Use Kernel#loop with break rather than begin/end/until or begin/end/while for post-loop tests. ^[link]

  # bad
  begin
    puts val
    val += 1
  end while val < 0
  
  # good
  loop do
    puts val
    val += 1
    break unless val < 0
  end

• Omit the outer braces around an implicit options hash. ^[link]

  # bad
  user.set({ name: 'John', age: 45, permissions: { read: true } })
  
  # good
  user.set(name: 'John', age: 45, permissions: { read: true })

• Omit both the outer braces and parentheses for methods that are part of an internal DSL. ^[link]

  class Person < ActiveRecord::Base
    # bad
    validates(:name, { presence: true, length: { within: 1..10 } })
  
    # good
    validates :name, presence: true, length: { within: 1..10 }
  end

• Use the proc invocation shorthand when the invoked method is the only operation of a block. ^[link]

  # bad
  names.map { |name| name.upcase }
  
  # good
  names.map(&:upcase)

• Prefer {...} over do...end for single-line blocks. Avoid using {...} for multi-line blocks (multi-line chaining is always ugly). Always use do...end for "control flow" and "method definitions" (e.g. in Rakefiles and certain DSLs). Avoid do...end when chaining. ^[link]

  names = %w[Bozhidar Steve Sarah]
  
  # bad
  names.each do |name|
    puts name
  end
  
  # good
  names.each { |name| puts name }
  
  # bad
  names.select do |name|
    name.start_with?('S')
  end.map { |name| name.upcase }
  
  # good
  names.select { |name| name.start_with?('S') }.map(&:upcase)

  Some will argue that multi-line chaining would look OK with the use of {...}, but they should ask themselves—is this code really readable and can the blocks' contents be extracted into nifty methods?

• Consider using explicit block argument to avoid writing block literal that just passes its arguments to another block. Beware of the performance impact, though, as the block gets converted to a Proc. ^[link]

  require 'tempfile'
  
  # bad
  def with_tmp_dir
    Dir.mktmpdir do |tmp_dir|
      Dir.chdir(tmp_dir) { |dir| yield dir }  # block just passes arguments
    end
  end
  
  # good
  def with_tmp_dir(&block)
    Dir.mktmpdir do |tmp_dir|
      Dir.chdir(tmp_dir, &block)
    end
  end
  
  with_tmp_dir do |dir|
    puts "dir is accessible as a parameter and pwd is set: #{dir}"
  end

• Avoid return where not required for flow of control. ^[link]

  # bad
  def some_method(some_arr)
    return some_arr.size
  end
  
  # good
  def some_method(some_arr)
    some_arr.size
  end

• Avoid self where not required. (It is only required when calling a self write accessor, methods named after reserved words, or overloadable operators.) ^[link]

  # bad
  def ready?
    if self.last_reviewed_at > self.last_updated_at
      self.worker.update(self.content, self.options)
      self.status = :in_progress
    end
    self.status == :verified
  end
  
  # good
  def ready?
    if last_reviewed_at > last_updated_at
      worker.update(content, options)
      self.status = :in_progress
    end
    status == :verified
  end

• As a corollary, avoid shadowing methods with local variables unless they are both equivalent. ^[link]

  class Foo
    attr_accessor :options
  
    # ok
    def initialize(options)
      self.options = options
      # both options and self.options are equivalent here
    end
  
    # bad
    def do_something(options = {})
      unless options[:when] == :later
        output(self.options[:message])
      end
    end
  
    # good
    def do_something(params = {})
      unless params[:when] == :later
        output(options[:message])
      end
    end
  end

• Don't use the return value of = (an assignment) in conditional expressions unless the assignment is wrapped in parentheses. This is a fairly popular idiom among Rubyists that's sometimes referred to as safe assignment in condition. ^[link]

  # bad (+ a warning)
  if v = array.grep(/foo/)
    do_something(v)
    # some code
  end
  
  # good
  v = array.grep(/foo/)
  if v
    do_something(v)
    # some code
  end

• Use shorthand self assignment operators whenever applicable. ^[link]

  # bad
  x = x + y
  x = x * y
  x = x**y
  x = x / y
  x = x || y
  x = x && y
  
  # good
  x += y
  x *= y
  x **= y
  x /= y
  x ||= y
  x &&= y

• Use ||= to initialize variables only if they're not already initialized. ^[link]

  # bad
  name = name ? name : 'Bozhidar'
  
  # bad
  name = 'Bozhidar' unless name
  
  # good - set name to 'Bozhidar', only if it's nil or false
  name ||= 'Bozhidar'

• Don't use ||= to initialize boolean variables. (Consider what would happen if the current value happened to be false.) ^[link]

  # bad - would set enabled to true even if it was false
  enabled ||= true
  
  # good
  enabled = true if enabled.nil?

• Use &&= to preprocess variables that may or may not exist. Using &&= will change the value only if it exists, removing the need to check its existence with if. ^[link]

  # bad
  if something
    something = something.downcase
  end
  
  # bad
  something = something ? something.downcase : nil
  
  # ok
  something = something.downcase if something
  
  # good
  something = something && something.downcase
  
  # better
  something &&= something.downcase

• Avoid explicit use of the case equality operator ===. As its name implies it is meant to be used implicitly by case expressions and outside of them it yields some pretty confusing code. ^[link]

  # bad
  Array === something
  (1..100) === 7
  /something/ === some_string
  
  # good
  something.is_a?(Array)
  (1..100).include?(7)
  some_string.match?(/something/)

• Do not use eql? when using == will do. The stricter comparison semantics provided by eql? are rarely needed in practice. ^[link]

  # bad - eql? is the same as == for strings
  'ruby'.eql? some_str
  
  # good
  'ruby' == some_str
  1.0.eql? x # eql? makes sense here if want to differentiate between Integer and Float 1

• Avoid using Perl-style special variables (like $:, $;, etc. ). They are quite cryptic and their use in anything but one-liner scripts is discouraged. Use the human-friendly aliases provided by the English library. ^[link]

  # bad
  $:.unshift File.dirname(__FILE__)
  
  # good
  require 'English'
  $LOAD_PATH.unshift File.dirname(__FILE__)

• Do not put a space between a method name and the opening parenthesis. ^[link]

  # bad
  f (3 + 2) + 1
  
  # good
  f(3 + 2) + 1

• Always run the Ruby interpreter with the -w option so it will warn you if you forget either of the rules above! ^[link]

• Do not use nested method definitions, use lambda instead. Nested method definitions actually produce methods in the same scope (e.g. class) as the outer method. Furthermore, the "nested method" will be redefined every time the method containing its definition is invoked. ^[link]

  # bad
  def foo(x)
    def bar(y)
      # body omitted
    end
  
    bar(x)
  end
  
  # good - the same as the previous, but no bar redefinition on every foo call
  def bar(y)
    # body omitted
  end
  
  def foo(x)
    bar(x)
  end
  
  # also good
  def foo(x)
    bar = ->(y) { ... }
    bar.call(x)
  end

• Use the new lambda literal syntax for single line body blocks. Use the lambda method for multi-line blocks. ^[link]

  # bad
  l = lambda { |a, b| a + b }
  l.call(1, 2)
  
  # correct, but looks extremely awkward
  l = ->(a, b) do
    tmp = a * 7
    tmp * b / 50
  end
  
  # good
  l = ->(a, b) { a + b }
  l.call(1, 2)
  
  l = lambda do |a, b|
    tmp = a * 7
    tmp * b / 50
  end

• Don't omit the parameter parentheses when defining a stabby lambda with parameters. ^[link]

  # bad
  l = ->x, y { something(x, y) }
  
  # good
  l = ->(x, y) { something(x, y) }

• Omit the parameter parentheses when defining a stabby lambda with no parameters. ^[link]

  # bad
  l = ->() { something }
  
  # good
  l = -> { something }

• Prefer proc over Proc.new. ^[link]

  # bad
  p = Proc.new { |n| puts n }
  
  # good
  p = proc { |n| puts n }

• Prefer proc.call() over proc[] or proc.() for both lambdas and procs. ^[link]

  # bad - looks similar to Enumeration access
  l = ->(v) { puts v }
  l[1]
  
  # also bad - uncommon syntax
  l = ->(v) { puts v }
  l.(1)
  
  # good
  l = ->(v) { puts v }
  l.call(1)

• Prefix with _ unused block parameters and local variables. It's also acceptable to use just _ (although it's a bit less descriptive). This convention is recognized by the Ruby interpreter and tools like RuboCop and will suppress their unused variable warnings. ^[link]

  # bad
  result = hash.map { |k, v| v + 1 }
  
  def something(x)
    unused_var, used_var = something_else(x)
    # some code
  end
  
  # good
  result = hash.map { |_, v| v + 1 }
  
  def something(x)
    _, used_var = something_else(x)
    # some code
  end

• Use $stdout/$stderr/$stdin instead of STDOUT/STDERR/STDIN. STDOUT/STDERR/STDIN are constants, and while you can actually reassign (possibly to redirect some stream) constants in Ruby, you'll get an interpreter warning if you do so. ^[link]

• Use warn instead of $stderr.puts. Apart from being more concise and clear, warn allows you to suppress warnings if you need to (by setting the warn level to 0 via -W0). ^[link]

• Favor the use of sprintf and its alias format over the fairly cryptic String#% method. ^[link]

  # bad
  '%d %d' % [20, 10]
  # => '20 10'
  
  # good
  sprintf('%d %d', 20, 10)
  # => '20 10'
  
  # good
  sprintf('%<first>d %<second>d', first: 20, second: 10)
  # => '20 10'
  
  format('%d %d', 20, 10)
  # => '20 10'
  
  # good
  format('%<first>d %<second>d', first: 20, second: 10)
  # => '20 10'

• Favor the use of Array#join over the fairly cryptic Array#* with a string argument. ^[link]

  # bad
  %w[one two three] * ', '
  # => 'one, two, three'
  
  # good
  %w[one two three].join(', ')
  # => 'one, two, three'

• Use ranges or Comparable#between? instead of complex comparison logic when possible. ^[link]

  # bad
  do_something if x >= 1000 && x <= 2000
  
  # good
  do_something if (1000..2000).include?(x)
  
  # good
  do_something if x.between?(1000, 2000)

• Favor the use of predicate methods to explicit comparisons with ==. Numeric comparisons are OK. ^[link]

  # bad
  if x % 2 == 0
  end
  
  if x % 2 == 1
  end
  
  if x == nil
  end
  
  # good
  if x.even?
  end
  
  if x.odd?
  end
  
  if x.nil?
  end
  
  if x.zero?
  end
  
  if x == 0
  end

• Don't do explicit non-nil checks unless you're dealing with boolean values. ^[link]

  # bad
  do_something if !something.nil?
  do_something if something != nil
  
  # good
  do_something if something
  
  # good - dealing with a boolean
  def value_set?
    !@some_boolean.nil?
  end

• Avoid the use of BEGIN blocks. ^[link]

• Do not use END blocks. Use Kernel#at_exit instead. ^[link]

  # bad
  END { puts 'Goodbye!' }
  
  # good
  at_exit { puts 'Goodbye!' }

• Avoid the use of flip-flops. ^[link]

• Avoid use of nested conditionals for flow of control. ^[link]

  Prefer a guard clause when you can assert invalid data. A guard clause is a conditional statement at the top of a function that bails out as soon as it can.

  # bad
  def compute_thing(thing)
    if thing[:foo]
      update_with_bar(thing[:foo])
      if thing[:foo][:bar]
        partial_compute(thing)
      else
        re_compute(thing)
      end
    end
  end
  
  # good
  def compute_thing(thing)
    return unless thing[:foo]
    update_with_bar(thing[:foo])
    return re_compute(thing) unless thing[:foo][:bar]
    partial_compute(thing)
  end

  Prefer next in loops instead of conditional blocks.

  # bad
  [0, 1, 2, 3].each do |item|
    if item > 1
      puts item
    end
  end
  
  # good
  [0, 1, 2, 3].each do |item|
    next unless item > 1
    puts item
  end

• Prefer map over collect, find over detect, select over find_all, reduce over inject and size over length. This is not a hard requirement; if the use of the alias enhances readability, it's ok to use it. The rhyming methods are inherited from Smalltalk and are not common in other programming languages. The reason the use of select is encouraged over find_all is that it goes together nicely with reject and its name is pretty self-explanatory. ^[link]

• Don't use count as a substitute for size. For Enumerable objects other than Array it will iterate the entire collection in order to determine its size. ^[link]

  # bad
  some_hash.count
  
  # good
  some_hash.size

• Use flat_map instead of map + flatten. This does not apply for arrays with a depth greater than 2, i.e. if users.first.songs == ['a', ['b','c']], then use map + flatten rather than flat_map. flat_map flattens the array by 1, whereas flatten flattens it all the way. ^[link]

  # bad
  all_songs = users.map(&:songs).flatten.uniq
  
  # good
  all_songs = users.flat_map(&:songs).uniq

• Prefer reverse_each to reverse.each because some classes that include Enumerable will provide an efficient implementation. Even in the worst case where a class does not provide a specialized implementation, the general implementation inherited from Enumerable will be at least as efficient as using reverse.each. ^[link]

  # bad
  array.reverse.each { ... }
  
  # good
  array.reverse_each { ... }

Naming

• Name identifiers in English. ^[link]

  # bad - identifier using non-ascii characters
  заплата = 1_000
  
  # bad - identifier is a Bulgarian word, written with Latin letters (instead of Cyrillic)
  zaplata = 1_000
  
  # good
  salary = 1_000

• Use snake_case for symbols, methods and variables. ^[link]

  # bad
  :'some symbol'
  :SomeSymbol
  :someSymbol
  
  someVar = 5
  
  def someMethod
    # some code
  end
  
  def SomeMethod
    # some code
  end
  
  # good
  :some_symbol
  
  some_var = 5
  
  def some_method
    # some code
  end

• Do not separate numbers from letters on symbols, methods and variables. ^[link]

  # bad
  :some_sym_1
  
  some_var_1 = 1
  
  var_10 = 10
  
  def some_method_1
    # some code
  end
  
  # good
  :some_sym1
  
  some_var1 = 1
  
  var10 = 10
  
  def some_method1
    # some code
  end

• Use CamelCase for classes and modules. (Keep acronyms like HTTP, RFC, XML uppercase.) ^[link]

  # bad
  class Someclass
    # some code
  end
  
  class Some_Class
    # some code
  end
  
  class SomeXml
    # some code
  end
  
  class XmlSomething
    # some code
  end
  
  # good
  class SomeClass
    # some code
  end
  
  class SomeXML
    # some code
  end
  
  class XMLSomething
    # some code
  end

• Use snake_case for naming files, e.g. hello_world.rb. ^[link]

• Use snake_case for naming directories, e.g. lib/hello_world/hello_world.rb. ^[link]

• Aim to have just a single class/module per source file. Name the file name as the class/module, but replacing CamelCase with snake_case. ^[link]

• Use SCREAMING_SNAKE_CASE for other constants. ^[link]

  # bad
  SomeConst = 5
  
  # good
  SOME_CONST = 5

• The names of predicate methods (methods that return a boolean value) should end in a question mark. (i.e. Array#empty?). Methods that don't return a boolean, shouldn't end in a question mark. ^[link]

• Avoid prefixing predicate methods with the auxiliary verbs such as is, does, or can. These words are redundant and inconsistent with the style of boolean methods in the Ruby core library, such as empty? and include?. ^[link]

  # bad
  class Person
    def is_tall?
      true
    end
  
    def can_play_basketball?
      false
    end
  
    def does_like_candy?
      true
    end
  end
  
  # good
  class Person
    def tall?
      true
    end
  
    def basketball_player?
      false
    end
  
    def likes_candy?
      true
    end
  end

• The names of potentially dangerous methods (i.e. methods that modify self or the arguments, exit! (doesn't run the finalizers like exit does), etc.) should end with an exclamation mark if there exists a safe version of that dangerous method. ^[link]

  # bad - there is no matching 'safe' method
  class Person
    def update!
    end
  end
  
  # good
  class Person
    def update
    end
  end
  
  # good
  class Person
    def update!
    end
  
    def update
    end
  end

• Define the non-bang (safe) method in terms of the bang (dangerous) one if possible. ^[link]

  class Array
    def flatten_once!
      res = []
  
      each do |e|
        [*e].each { |f| res << f }
      end
  
      replace(res)
    end
  
    def flatten_once
      dup.flatten_once!
    end
  end

• When defining binary operators and operator-alike methods, name the parameter other for operators with "symmetrical" semantics of operands. Symmetrical semantics means both sides of the operator are typically of same or coercible types. ^[link]

  • operators and operator-alike methods with symmetrical semantics (the parameter should be named other): +, -, *, /, %, **, ==, >, <, |, &, ^, eql?, equal?;
  • operators with non-symmetrical semantics (the parameter should not be named other): <<, [] (collection/item relations between operands), === (pattern/matchable relations);

  Note that the rule should be followed only if both sides of the operator have the same semantics. Prominent exception in Ruby core is, for example, Array#*(int).

  # good
  def +(other)
    # body omitted
  end
  
  # bad
  def <<(other)
    @internal << other
  end
  
  # good
  def <<(item)
    @internal << item
  end
  
  # bad
  # Returns some string multiplied `other` times
  def *(other)
    # body omitted
  end
  
  # good
  # Returns some string multiplied `num` times
  def *(num)
    # body omitted
  end

Comments

• Write self-documenting code and ignore the rest of this section. Seriously! ^[link]

• Write comments in English. ^[link]

• Use one space between the leading # character of the comment and the text of the comment. ^[link]

• Comments longer than a word are capitalized and use punctuation. Use one space after periods. ^[link]

• Avoid superfluous comments. ^[link]

  # bad
  counter += 1 # Increments counter by one.

• Keep existing comments up-to-date. An outdated comment is worse than no comment at all. ^[link]

• Write self-documenting code and ignore the rest of this section. Seriously! ^[link]

• Avoid writing comments to explain bad code. Refactor the code to make it self-explanatory. ^[link]

Comment Annotations

• Annotations should usually be written on the line immediately above the relevant code. ^[link]

• The annotation keyword is followed by a colon and a space, then a note describing the problem. ^[link]

• If multiple lines are required to describe the problem, subsequent lines should be indented three spaces after the # (one general plus two for indentation purpose). ^[link]

  def bar
    # FIXME: This has crashed occasionally since v3.2.1. It may
    #   be related to the BarBazUtil upgrade.
    baz(:quux)
  end

• In cases where the problem is so obvious that any documentation would be redundant, annotations may be left at the end of the offending line with no note. This usage should be the exception and not the rule. ^[link]

  def bar
    sleep 100 # OPTIMIZE
  end

• Use TODO to note missing features or functionality that should be added at a later date. ^[link]

• Use FIXME to note broken code that needs to be fixed. ^[link]

• Use OPTIMIZE to note slow or inefficient code that may cause performance problems. ^[link]

• Use HACK to note code smells where questionable coding practices were used and should be refactored away. ^[link]

• Use REVIEW to note anything that should be looked at to confirm it is working as intended. For example: REVIEW: Are we sure this is how the client does X currently? ^[link]

• Use other custom annotation keywords if it feels appropriate, but be sure to document them in your project's README or similar. ^[link]

Magic Comments

• Place magic comments above all code and documentation in a file (except shebangs, which are discussed next).

  # bad
  # Some documentation about Person
  
  # frozen_string_literal: true
  class Person
  end
  
  # good
  # frozen_string_literal: true
  
  # Some documentation about Person
  class Person
  end

• Place magic comments below shebangs when they are present in a file. ^[link]

  # bad
  # frozen_string_literal: true
  #!/usr/bin/env ruby
  
  App.parse(ARGV)
  
  # good
  #!/usr/bin/env ruby
  # frozen_string_literal: true
  
  App.parse(ARGV)

• Use one magic comment per line if you need multiple. ^[link]

  # bad
  # -*- frozen_string_literal: true; encoding: ascii-8bit -*-
  
  # good
  # frozen_string_literal: true
  # encoding: ascii-8bit

• Separate magic comments from code and documentation with a blank line. ^[link]

  # bad
  # frozen_string_literal: true
  # Some documentation for Person
  class Person
    # Some code
  end
  
  # good
  # frozen_string_literal: true
  
  # Some documentation for Person
  class Person
    # Some code
  end

Classes & Modules

• Use a consistent structure in your class definitions. ^[link]

  class Person
    # extend and include go first
    extend SomeModule
    include AnotherModule
  
    # inner classes
    CustomError = Class.new(StandardError)
  
    # constants are next
    SOME_CONSTANT = 20
  
    # afterwards we have attribute macros
    attr_reader :name
  
    # followed by other macros (if any)
    validates :name
  
    # public class methods are next in line
    def self.some_method
    end
  
    # initialization goes between class methods and other instance methods
    def initialize
    end
  
    # followed by other public instance methods
    def some_method
    end
  
    # protected and private methods are grouped near the end
    protected
  
    def some_protected_method
    end
  
    private
  
    def some_private_method
    end
  end

• Split multiple mixins into separate statements. ^[link]

  # bad
  class Person
    include Foo, Bar
  end
  
  # good
  class Person
    # multiple mixins go in separate statements
    include Foo
    include Bar
  end

• Don't nest multi-line classes within classes. Try to have such nested classes each in their own file in a folder named like the containing class. ^[link]

  # bad
  
  # foo.rb
  class Foo
    class Bar
      # 30 methods inside
    end
  
    class Car
      # 20 methods inside
    end
  
    # 30 methods inside
  end
  
  # good
  
  # foo.rb
  class Foo
    # 30 methods inside
  end
  
  # foo/bar.rb
  class Foo
    class Bar
      # 30 methods inside
    end
  end
  
  # foo/car.rb
  class Foo
    class Car
      # 20 methods inside
    end
  end

• Define (and reopen) namespaced classes and modules using explicit nesting. Using the scope resolution operator can lead to surprising constant lookups due to Ruby's lexical scoping, which depends on the module nesting at the point of definition. ^[link]

  module Utilities
    class Queue
    end
  end
  
  # bad
  class Utilities::Store
    Module.nesting # => [Utilities::Store]
  
    def initialize
      # Refers to the top level ::Queue class because Utilities isn't in the
      # current nesting chain.
      @queue = Queue.new
    end
  end
  
  # good
  module Utilities
    class WaitingList
      Module.nesting # => [Utilities::WaitingList, Utilities]
  
      def initialize
        @queue = Queue.new # Refers to Utilities::Queue
      end
    end
  end

• Prefer modules to classes with only class methods. Classes should be used only when it makes sense to create instances out of them. ^[link]

  # bad
  class SomeClass
    def self.some_method
      # body omitted
    end
  
    def self.some_other_method
      # body omitted
    end
  end
  
  # good
  module SomeModule
    module_function
  
    def some_method
      # body omitted
    end
  
    def some_other_method
      # body omitted
    end
  end

• Favor the use of module_function over extend self when you want to turn a module's instance methods into class methods. ^[link]

  # bad
  module Utilities
    extend self
  
    def parse_something(string)
      # do stuff here
    end
  
    def other_utility_method(number, string)
      # do some more stuff
    end
  end
  
  # good
  module Utilities
    module_function
  
    def parse_something(string)
      # do stuff here
    end
  
    def other_utility_method(number, string)
      # do some more stuff
    end
  end

• When designing class hierarchies make sure that they conform to the Liskov Substitution Principle. ^[link]

• Try to make your classes as SOLID as possible. ^[link]

• Always supply a proper to_s method for classes that represent domain objects. ^[link]

  class Person
    attr_reader :first_name, :last_name
  
    def initialize(first_name, last_name)
      @first_name = first_name
      @last_name = last_name
    end
  
    def to_s
      "#{first_name} #{last_name}"
    end
  end

• Use the attr family of functions to define trivial accessors or mutators. ^[link]

  # bad
  class Person
    def initialize(first_name, last_name)
      @first_name = first_name
      @last_name = last_name
    end
  
    def first_name
      @first_name
    end
  
    def last_name
      @last_name
    end
  end
  
  # good
  class Person
    attr_reader :first_name, :last_name
  
    def initialize(first_name, last_name)
      @first_name = first_name
      @last_name = last_name
    end
  end

• For accessors and mutators, avoid prefixing method names with get_ and set_. It is a Ruby convention to use attribute names for accessors (readers) and attr_name= for mutators (writers). ^[link]

  # bad
  class Person
    def get_name
      "#{@first_name} #{@last_name}"
    end
  
    def set_name(name)
      @first_name, @last_name = name.split(' ')
    end
  end
  
  # good
  class Person
    def name
      "#{@first_name} #{@last_name}"
    end
  
    def name=(name)
      @first_name, @last_name = name.split(' ')
    end
  end

• Avoid the use of attr. Use attr_reader and attr_accessor instead. ^[link]

  # bad - creates a single attribute accessor (deprecated in Ruby 1.9)
  attr :something, true
  attr :one, :two, :three # behaves as attr_reader
  
  # good
  attr_accessor :something
  attr_reader :one, :two, :three

• Consider using Struct.new, which defines the trivial accessors, constructor and comparison operators for you. ^[link]

  # good
  class Person
    attr_accessor :first_name, :last_name
  
    def initialize(first_name, last_name)
      @first_name = first_name
      @last_name = last_name
    end
  end
  
  # better
  Person = Struct.new(:first_name, :last_name) do
  end

• Don't extend an instance initialized by Struct.new. Extending it introduces a superfluous class level and may also introduce weird errors if the file is required multiple times. ^[link]

  # bad
  class Person < Struct.new(:first_name, :last_name)
  end
  
  # good
  Person = Struct.new(:first_name, :last_name)

• Consider adding factory methods to provide additional sensible ways to create instances of a particular class. ^[link]

  class Person
    def self.create(options_hash)
      # body omitted
    end
  end

• Prefer duck-typing over inheritance. ^[link]

  # bad
  class Animal
    # abstract method
    def speak
    end
  end
  
  # extend superclass
  class Duck < Animal
    def speak
      puts 'Quack! Quack'
    end
  end
  
  # extend superclass
  class Dog < Animal
    def speak
      puts 'Bau! Bau!'
    end
  end
  
  # good
  class Duck
    def speak
      puts 'Quack! Quack'
    end
  end
  
  class Dog
    def speak
      puts 'Bau! Bau!'
    end
  end

• Avoid the usage of class (@@) variables due to their "nasty" behavior in inheritance. ^[link]

  class Parent
    @@class_var = 'parent'
  
    def self.print_class_var
      puts @@class_var
    end
  end
  
  class Child < Parent
    @@class_var = 'child'
  end
  
  Parent.print_class_var # => will print 'child'

  As you can see all the classes in a class hierarchy actually share one class variable. Class instance variables should usually be preferred over class variables.

• Assign proper visibility levels to methods (private, protected) in accordance with their intended usage. Don't go off leaving everything public (which is the default). After all we're coding in Ruby now, not in Python. ^[link]

• Indent the public, protected, and private methods as much as the method definitions they apply to. Leave one blank line above the visibility modifier and one blank line below in order to emphasize that it applies to all methods below it. ^[link]

  class SomeClass
    def public_method
      # some code
    end
  
    private
  
    def private_method
      # some code
    end
  
    def another_private_method
      # some code
    end
  end

• Use def self.method to define class methods. This makes the code easier to refactor since the class name is not repeated. ^[link]

  class TestClass
    # bad
    def TestClass.some_method
      # body omitted
    end
  
    # good
    def self.some_other_method
      # body omitted
    end
  end

• Prefer alias when aliasing methods in lexical class scope as the resolution of self in this context is also lexical, and it communicates clearly to the user that the indirection of your alias will not be altered at runtime or by any subclass unless made explicit. ^[link]

  class Westerner
    def first_name
      @names.first
    end
  
    alias given_name first_name
  end

  Since alias, like def, is a keyword, prefer bareword arguments over symbols or strings. In other words, do alias foo bar, not alias :foo :bar.

  Also be aware of how Ruby handles aliases and inheritance: an alias references the method that was resolved at the time the alias was defined; it is not dispatched dynamically.

  class Fugitive < Westerner
    def first_name
      'Nobody'
    end
  end

  In this example, Fugitive#given_name would still call the original Westerner#first_name method, not Fugitive#first_name. To override the behavior of Fugitive#given_name as well, you'd have to redefine it in the derived class.

  class Fugitive < Westerner
    def first_name
      'Nobody'
    end
  
    alias given_name first_name
  end

• Always use alias_method when aliasing methods of modules, classes, or singleton classes at runtime, as the lexical scope of alias leads to unpredictability in these cases. ^[link]

  module Mononymous
    def self.included(other)
      other.class_eval { alias_method :full_name, :given_name }
    end
  end
  
  class Sting < Westerner
    include Mononymous
  end

  alias executed when source file read, alias_method executed at run time.

• When class (or module) methods call other such methods, omit the use of a leading self or own name followed by a . when calling other such methods. This is often seen in "service classes" or other similar concepts where a class is treated as though it were a function. This convention tends to reduce repetitive boilerplate in such classes. ^[link]

  class TestClass
    # bad -- more work when class renamed/method moved
    def self.call(param1, param2)
      TestClass.new(param1).call(param2)
    end
  
    # bad -- more verbose than necessary
    def self.call(param1, param2)
      self.new(param1).call(param2)
    end
  
    # good
    def self.call(param1, param2)
      new(param1).call(param2)
    end
  
    # ...other methods...
  end

Exceptions

• Prefer raise over fail for exceptions. ^[link]

  # bad
  fail SomeException, 'message'
  
  # good
  raise SomeException, 'message'

• Don't specify RuntimeError explicitly in the two argument version of raise. ^[link]

  # bad
  raise RuntimeError, 'message'
  
  # good - signals a RuntimeError by default
  raise 'message'

• Prefer supplying an exception class and a message as two separate arguments to raise, instead of an exception instance. ^[link]

  # bad
  raise SomeException.new('message')
  # Note that there is no way to do `raise SomeException.new('message'), backtrace`.
  
  # good
  raise SomeException, 'message'
  # Consistent with `raise SomeException, 'message', backtrace`.

• Do not return from an ensure block. If you explicitly return from a method inside an ensure block, the return will take precedence over any exception being raised, and the method will return as if no exception had been raised at all. In effect, the exception will be silently thrown away. ^[link]

  # bad
  def foo
    raise
  ensure
    return 'very bad idea'
  end

• Use implicit begin blocks where possible. ^[link]

  # bad
  def foo
    begin
      # main logic goes here
    rescue
      # failure handling goes here
    end
  end
  
  # good
  def foo
    # main logic goes here
  rescue
    # failure handling goes here
  end

• Mitigate the proliferation of begin blocks by using contingency methods (a term coined by Avdi Grimm). ^[link]

  # bad
  begin
    something_that_might_fail
  rescue IOError
    # handle IOError
  end
  
  begin
    something_else_that_might_fail
  rescue IOError
    # handle IOError
  end
  
  # good
  def with_io_error_handling
     yield
  rescue IOError
    # handle IOError
  end
  
  with_io_error_handling { something_that_might_fail }
  
  with_io_error_handling { something_else_that_might_fail }

• Don't suppress exceptions. ^[link]

  # bad
  begin
    # an exception occurs here
  rescue SomeError
    # the rescue clause does absolutely nothing
  end
  
  # bad
  do_something rescue nil

• Avoid using rescue in its modifier form. ^[link]

  # bad - this catches exceptions of StandardError class and its descendant classes
  read_file rescue handle_error($!)
  
  # good - this catches only the exceptions of Errno::ENOENT class and its descendant classes
  def foo
    read_file
  rescue Errno::ENOENT => ex
    handle_error(ex)
  end

• Don't use exceptions for flow of control. ^[link]

  # bad
  begin
    n / d
  rescue ZeroDivisionError
    puts 'Cannot divide by 0!'
  end
  
  # good
  if d.zero?
    puts 'Cannot divide by 0!'
  else
    n / d
  end

• Avoid rescuing the Exception class. This will trap signals and calls to exit, requiring you to kill -9 the process. ^[link]

  # bad
  begin
    # calls to exit and kill signals will be caught (except kill -9)
    exit
  rescue Exception
    puts "you didn't really want to exit, right?"
    # exception handling
  end
  
  # good
  begin
    # a blind rescue rescues from StandardError, not Exception as many
    # programmers assume.
  rescue => e
    # exception handling
  end
  
  # also good
  begin
    # an exception occurs here
  rescue StandardError => e
    # exception handling
  end

• Put more specific exceptions higher up the rescue chain, otherwise they'll never be rescued from. ^[link]

  # bad
  begin
    # some code
  rescue StandardError => e
    # some handling
  rescue IOError => e
    # some handling that will never be executed
  end
  
  # good
  begin
    # some code
  rescue IOError => e
    # some handling
  rescue StandardError => e
    # some handling
  end

• Release external resources obtained by your program in an ensure block. ^[link]

  f = File.open('testfile')
  begin
    # .. process
  rescue
    # .. handle error
  ensure
    f.close if f
  end

• Use versions of resource obtaining methods that do automatic resource cleanup when possible. ^[link]

  # bad - you need to close the file descriptor explicitly
  f = File.open('testfile')
  # some action on the file
  f.close
  
  # good - the file descriptor is closed automatically
  File.open('testfile') do |f|
    # some action on the file
  end

• Favor the use of exceptions from the standard library over introducing new exception classes. ^[link]

Collections

• Prefer literal array and hash creation notation (unless you need to pass parameters to their constructors, that is). ^[link]

  # bad
  arr = Array.new
  hash = Hash.new
  
  # good
  arr = []
  arr = Array.new(10)
  hash = {}
  hash = Hash.new(0)

• Avoid comma after the last item of an Array or Hash literal, especially when the items are not on separate lines. ^[link]

  # bad - easier to move/add/remove items, but still not preferred
  VALUES = [
             1001,
             2020,
             3333,
           ]
  
  # bad
  VALUES = [1001, 2020, 3333, ]
  
  # good
  VALUES = [1001, 2020, 3333]

• Avoid the creation of huge gaps in arrays. ^[link]

  arr = []
  arr[100] = 1 # now you have an array with lots of nils

• When accessing the first or last element from an array, prefer first or last over [0] or [-1]. ^[link]

• Use Set instead of Array when dealing with unique elements. Set implements a collection of unordered values with no duplicates. This is a hybrid of Array's intuitive inter-operation facilities and Hash's fast lookup. ^[link]

• Prefer symbols instead of strings as hash keys. ^[link]

  # bad
  hash = { 'one' => 1, 'two' => 2, 'three' => 3 }
  
  # good
  hash = { one: 1, two: 2, three: 3 }

• Avoid the use of mutable objects as hash keys. ^[link]

• Use the Ruby 1.9 hash literal syntax when your hash keys are symbols. ^[link]

  # bad
  hash = { :one => 1, :two => 2, :three => 3 }
  
  # good
  hash = { one: 1, two: 2, three: 3 }

• Don't mix the Ruby 1.9 hash syntax with hash rockets in the same hash literal. When you've got keys that are not symbols stick to the hash rockets syntax. ^[link]

  # bad
  { a: 1, 'b' => 2 }
  
  # good
  { :a => 1, 'b' => 2 }

• Use Hash#key? instead of Hash#has_key? and Hash#value? instead of Hash#has_value?. ^[link]

  # bad
  hash.has_key?(:test)
  hash.has_value?(value)
  
  # good
  hash.key?(:test)
  hash.value?(value)

• Use Hash#each_key instead of Hash#keys.each and Hash#each_value instead of Hash#values.each. ^[link]

  # bad
  hash.keys.each { |k| p k }
  hash.values.each { |v| p v }
  hash.each { |k, _v| p k }
  hash.each { |_k, v| p v }
  
  # good
  hash.each_key { |k| p k }
  hash.each_value { |v| p v }

• Use Hash#fetch when dealing with hash keys that should be present. ^[link]

  heroes = { batman: 'Bruce Wayne', superman: 'Clark Kent' }
  # bad - if we make a mistake we might not spot it right away
  heroes[:batman] # => 'Bruce Wayne'
  heroes[:supermann] # => nil
  
  # good - fetch raises a KeyError making the problem obvious
  heroes.fetch(:supermann)

• Introduce default values for hash keys via Hash#fetch as opposed to using custom logic. ^[link]

  batman = { name: 'Bruce Wayne', is_evil: false }
  
  # bad - if we just use || operator with falsy value we won't get the expected result
  batman[:is_evil] || true # => true
  
  # good - fetch works correctly with falsy values
  batman.fetch(:is_evil, true) # => false

• Prefer the use of the block instead of the default value in Hash#fetch if the code that has to be evaluated may have side effects or be expensive. ^[link]

  batman = { name: 'Bruce Wayne' }
  
  # bad - if we use the default value, we eager evaluate it
  # so it can slow the program down if done multiple times
  batman.fetch(:powers, obtain_batman_powers) # obtain_batman_powers is an expensive call
  
  # good - blocks are lazy evaluated, so only triggered in case of KeyError exception
  batman.fetch(:powers) { obtain_batman_powers }

• Use Hash#values_at when you need to retrieve several values consecutively from a hash. ^[link]

  # bad
  email = data['email']
  username = data['nickname']
  
  # good
  email, username = data.values_at('email', 'nickname')

• Rely on the fact that as of Ruby 1.9 hashes are ordered. ^[link]

• Do not modify a collection while traversing it. ^[link]

• When accessing elements of a collection, avoid direct access via [n] by using an alternate form of the reader method if it is supplied. This guards you from calling [] on nil. ^[link]

  # bad
  Regexp.last_match[1]
  
  # good
  Regexp.last_match(1)

• When providing an accessor for a collection, provide an alternate form to save users from checking for nil before accessing an element in the collection. ^[link]

  # bad
  def awesome_things
    @awesome_things
  end
  
  # good
  def awesome_things(index = nil)
    if index && @awesome_things
      @awesome_things[index]
    else
      @awesome_things
    end
  end
  
  # good
  def awesome_things(index: nil)
    return @awesome_things[index] if index && @awesome_things
    @awesome_things
  end

Numbers

• Use Integer to check type of an integer number. Since Fixnum is platform-dependent, checking against it will return different results on 32-bit and 64-bit machines. ^[link]

  timestamp = Time.now.to_i
  
  # bad
  timestamp.is_a? Fixnum
  timestamp.is_a? Bignum
  
  # good
  timestamp.is_a? Integer

• Prefer to use ranges when generating random numbers instead of integers with offsets, since it clearly states your intentions. Imagine simulating a roll of a dice: ^[link]

  # bad
  rand(6) + 1
  
  # good
  rand(1..6)

Strings

• Prefer string interpolation and string formatting instead of string concatenation: ^[link]

  # bad
  email_with_name = user.name + ' <' + user.email + '>'
  
  # good
  email_with_name = "#{user.name} <#{user.email}>"

• Prefer single-quoted strings when you don't need string interpolation or special symbols such as \t, \n, ', etc.

  # bad
  name = "Bozhidar"
  
  name = 'De\'Andre'
  
  # good
  name = 'Bozhidar'
  
  name = "De'Andre"

• Don't use the character literal syntax ?x. Since Ruby 1.9 it's basically redundant—?x would be interpreted as 'x' (a string with a single character in it). ^[link]

  # bad
  char = ?c
  
  # good
  char = 'c'

• Don't leave out {} around instance and global variables being interpolated into a string. ^[link]

  class Person
    attr_reader :first_name, :last_name
  
    def initialize(first_name, last_name)
      @first_name = first_name
      @last_name = last_name
    end
  
    # bad - valid, but awkward
    def to_s
      "#@first_name #@last_name"
    end
  
    # good
    def to_s
      "#{@first_name} #{@last_name}"
    end
  end
  
  $global = 0
  # bad
  puts "$global = #$global"
  
  # good
  puts "$global = #{$global}"

• Don't use Object#to_s on interpolated objects. It's invoked on them automatically. ^[link]

  # bad
  message = "This is the #{result.to_s}."
  
  # good
  message = "This is the #{result}."

• Avoid using String#+ when you need to construct large data chunks. Instead, use String#<<. Concatenation mutates the string instance in-place and is always faster than String#+, which creates a bunch of new string objects. ^[link]

  # bad
  html = ''
  html += '<h1>Page title</h1>'
  
  paragraphs.each do |paragraph|
    html += "<p>#{paragraph}</p>"
  end
  
  # good and also fast
  html = ''
  html << '<h1>Page title</h1>'
  
  paragraphs.each do |paragraph|
    html << "<p>#{paragraph}</p>"
  end

• Don't use String#gsub in scenarios in which you can use a faster and more specialized alternative. ^[link]

  url = 'http://example.com'
  str = 'lisp-case-rules'
  
  # bad
  url.gsub('http://', 'https://')
  str.gsub('-', '_')
  
  # good
  url.sub('http://', 'https://')
  str.tr('-', '_')

• When using heredocs for multi-line strings keep in mind the fact that they preserve leading whitespace. It's a good practice to employ some margin based on which to trim the excessive whitespace. ^[link]

  code = <<-END.gsub(/^\s+\|/, '')
    |def test
    |  some_method
    |  other_method
    |end
  END
  # => "def test\n  some_method\n  other_method\nend\n"

• Use Ruby 2.3's squiggly heredocs for nicely indented multi-line strings. ^[link]

  # bad - using Powerpack String#strip_margin
  code = <<-RUBY.strip_margin('|')
    |def test
    |  some_method
    |  other_method
    |end
  RUBY
  
  # also bad
  code = <<-RUBY
  def test
    some_method
    other_method
  end
  RUBY
  
  # good
  code = <<~RUBY
    def test
      some_method
      other_method
    end
  RUBY

• Use descriptive delimiters for heredocs. Delimiters add valuable information about the heredoc content, and as an added bonus some editors can highlight code within heredocs if the correct delimiter is used. ^[link]

  # bad
  code = <<~END
    def foo
      bar
    end
  END
  
  # good
  code = <<~RUBY
    def foo
      bar
    end
  RUBY
  
  # good
  code = <<~SUMMARY
    An imposing black structure provides a connection between the past and
    the future in this enigmatic adaptation of a short story by revered
    sci-fi author Arthur C. Clarke.
  SUMMARY

Date & Time

• Prefer Time.now over Time.new when retrieving the current system time. ^[link]

• Don't use DateTime unless you need to account for historical calendar reform—and if you do, explicitly specify the start argument to clearly state your intentions. ^[link]

  # bad - uses DateTime for current time
  DateTime.now
  
  # good - uses Time for current time
  Time.now
  
  # bad - uses DateTime for modern date
  DateTime.iso8601('2016-06-29')
  
  # good - uses Date for modern date
  Date.iso8601('2016-06-29')
  
  # good - uses DateTime with start argument for historical date
  DateTime.iso8601('1751-04-23', Date::ENGLAND)

Regular Expressions

• Don't use regular expressions if you just need plain text search in string: string['text'] ^[link]

• For simple constructions you can use regexp directly through string index. ^[link]

  match = string[/regexp/]             # get content of matched regexp
  first_group = string[/text(grp)/, 1] # get content of captured group
  string[/text (grp)/, 1] = 'replace'  # string => 'text replace'

• Use non-capturing groups when you don't use the captured result. ^[link]

  # bad
  /(first|second)/
  
  # good
  /(?:first|second)/

• Don't use the cryptic Perl-legacy variables denoting last regexp group matches ($1, $2, etc). Use Regexp.last_match(n) instead. ^[link]

  /(regexp)/ =~ string
  ...
  
  # bad
  process $1
  
  # good
  process Regexp.last_match(1)

• Avoid using numbered groups as it can be hard to track what they contain. Named groups can be used instead. ^[link]

  # bad
  /(regexp)/ =~ string
  # some code
  process Regexp.last_match(1)
  
  # good
  /(?<meaningful_var>regexp)/ =~ string
  # some code
  process meaningful_var

• Character classes have only a few special characters you should care about: ^, -, \, ], so don't escape . or brackets in []. ^[link]

• Be careful with ^ and $ as they match start/end of line, not string endings. If you want to match the whole string use: \A and \z (not to be confused with \Z which is the equivalent of /\n?\z/). ^[link]

  string = "some injection\nusername"
  string[/^username$/]   # matches
  string[/\Ausername\z/] # doesn't match

• Use x modifier for complex regexps. This makes them more readable and you can add some useful comments. Just be careful as spaces are ignored. ^[link]

  regexp = /
    start         # some text
    \s            # white space char
    (group)       # first group
    (?:alt1|alt2) # some alternation
    end
  /x

• For complex replacements sub/gsub can be used with a block or a hash. ^[link]

  words = 'foo bar'
  words.sub(/f/, 'f' => 'F') # => 'Foo bar'
  words.gsub(/\w+/) { |word| word.capitalize } # => 'Foo Bar'

Percent Literals

• Use %()(it's a shorthand for %Q) for single-line strings which require both interpolation and embedded double-quotes. For multi-line strings, prefer heredocs. ^[link]

  # bad (no interpolation needed)
  %(<div class="text">Some text</div>)
  # should be '<div class="text">Some text</div>'
  
  # bad (no double-quotes)
  %(This is #{quality} style)
  # should be "This is #{quality} style"
  
  # bad (multiple lines)
  %(<div>\n<span class="big">#{exclamation}</span>\n</div>)
  # should be a heredoc.
  
  # good (requires interpolation, has quotes, single line)
  %(<tr><td class="name">#{name}</td>)

• Avoid %() or the equivalent %q() unless you have a string with both ' and " in it. Regular string literals are more readable and should be preferred unless a lot of characters would have to be escaped in them. ^[link]

  # bad
  name = %q(Bruce Wayne)
  time = %q(8 o'clock)
  question = %q("What did you say?")
  
  # good
  name = 'Bruce Wayne'
  time = "8 o'clock"
  question = '"What did you say?"'
  quote = %q(<p class='quote'>"What did you say?"</p>)

• Use %r only for regular expressions matching at least one '/' character. ^[link]

  # bad
  %r{\s+}
  
  # good
  %r{^/(.*)$}
  %r{^/blog/2011/(.*)$}

• Avoid the use of %x unless you're going to invoke a command with backquotes in it (which is rather unlikely). ^[link]

  # bad
  date = %x(date)
  
  # good
  date = `date`
  echo = %x(echo `date`)

• Avoid the use of %s. It seems that the community has decided :"some string" is the preferred way to create a symbol with spaces in it. ^[link]

• Use the braces that are the most appropriate for the various kinds of percent literals. ^[link]

  • () for string literals(%q, %Q).
  • [] for array literals(%w, %i, %W, %I) as it is aligned with the standard array literals.
  • {} for regexp literals(%r) since parentheses often appear inside regular expressions. That's why a less common character with { is usually the best delimiter for %r literals.
  • () for all other literals (e.g. %s, %x)

  # bad
  %q{"Test's king!", John said.}
  
  # good
  %q("Test's king!", John said.)
  
  # bad
  %w(one two three)
  %i(one two three)
  
  # good
  %w[one two three]
  %i[one two three]
  
  # bad
  %r((\w+)-(\d+))
  %r{\w{1,2}\d{2,5}}
  
  # good
  %r{(\w+)-(\d+)}
  %r|\w{1,2}\d{2,5}|

Metaprogramming

• Avoid needless metaprogramming. ^[link]

• Do not mess around in core classes when writing libraries. (Do not monkey-patch them.) ^[link]

• The block form of class_eval is preferable to the string-interpolated form. ^[link]

  • when you use the string-interpolated form, always supply __FILE__ and __LINE__, so that your backtraces make sense:

    class_eval 'def use_relative_model_naming?; true; end', __FILE__, __LINE__

  • define_method is preferable to class_eval{ def ... }

• When using class_eval (or other eval) with string interpolation, add a comment block showing its appearance if interpolated (a practice used in Rails code): ^[link]

  # from activesupport/lib/active_support/core_ext/string/output_safety.rb
  UNSAFE_STRING_METHODS.each do |unsafe_method|
    if 'String'.respond_to?(unsafe_method)
      class_eval <<-EOT, __FILE__, __LINE__ + 1
        def #{unsafe_method}(*params, &block)       # def capitalize(*params, &block)
          to_str.#{unsafe_method}(*params, &block)  #   to_str.capitalize(*params, &block)
        end                                         # end
  
        def #{unsafe_method}!(*params)              # def capitalize!(*params)
          @dirty = true                             #   @dirty = true
          super                                     #   super
        end                                         # end
      EOT
    end
  end

• Avoid using method_missing for metaprogramming because backtraces become messy, the behavior is not listed in #methods, and misspelled method calls might silently work, e.g. nukes.launch_state = false. Consider using delegation, proxy, or define_method instead. If you must use method_missing: ^[link]

  • Be sure to also define respond_to_missing?

  • Only catch methods with a well-defined prefix, such as find_by_*—make your code as assertive as possible.

  • Call super at the end of your statement

  • Delegate to assertive, non-magical methods:

    # bad
    def method_missing(meth, *params, &block)
      if /^find_by_(?<prop>.*)/ =~ meth
        # ... lots of code to do a find_by
      else
        super
      end
    end
    
    # good
    def method_missing(meth, *params, &block)
      if /^find_by_(?<prop>.*)/ =~ meth
        find_by(prop, *params, &block)
      else
        super
      end
    end
    
    # best of all, though, would to define_method as each findable attribute is declared

• Prefer public_send over send so as not to circumvent private/protected visibility. ^[link]

  # We have an ActiveModel Organization that includes concern Activatable
  module Activatable
    extend ActiveSupport::Concern
  
    included do
      before_create :create_token
    end
  
    private
  
    def reset_token
      # some code
    end
  
    def create_token
      # some code
    end
  
    def activate!
      # some code
    end
  end
  
  class Organization < ActiveRecord::Base
    include Activatable
  end
  
  linux_organization = Organization.find(...)
  # BAD - violates privacy
  linux_organization.send(:reset_token)
  # GOOD - should throw an exception
  linux_organization.public_send(:reset_token)

• Prefer __send__ over send, as send may overlap with existing methods. ^[link]

  require 'socket'
  
  u1 = UDPSocket.new
  u1.bind('127.0.0.1', 4913)
  u2 = UDPSocket.new
  u2.connect('127.0.0.1', 4913)
  # Won't send a message to the receiver obj.
  # Instead it will send a message via UDP socket.
  u2.send :sleep, 0
  # Will actually send a message to the receiver obj.
  u2.__send__ ...

Misc

• Write ruby -w safe code. ^[link]

• Avoid hashes as optional parameters. Does the method do too much? (Object initializers are exceptions for this rule). ^[link]

• Avoid methods longer than 10 LOC (lines of code). Ideally, most methods will be shorter than 5 LOC. Empty lines do not contribute to the relevant LOC. ^[link]

• Avoid parameter lists longer than three or four parameters. ^[link]

• If you really need "global" methods, add them to Kernel and make them private. ^[link]

• Use module instance variables instead of global variables. ^[link]

  # bad
  $foo_bar = 1
  
  # good
  module Foo
    class << self
      attr_accessor :bar
    end
  end
  
  Foo.bar = 1

• Use OptionParser for parsing complex command line options and ruby -s for trivial command line options. ^[link]

• Code in a functional way, avoiding mutation when that makes sense. ^[link]

  a = []; [1, 2, 3].each { |i| a << i * 2 }   # bad
  a = [1, 2, 3].map { |i| i * 2 }             # good
  
  a = {}; [1, 2, 3].each { |i| a[i] = i * 17 }                # bad
  a = [1, 2, 3].reduce({}) { |h, i| h[i] = i * 17; h }        # good
  a = [1, 2, 3].each_with_object({}) { |i, h| h[i] = i * 17 } # good

• Do not mutate parameters unless that is the purpose of the method. ^[link]

• Avoid more than three levels of block nesting. ^[link]

• Be consistent. In an ideal world, be consistent with these guidelines. ^[link]

• Use common sense. ^[link]

Tools

Here are some tools to help you automatically check Ruby code against this guide.

RuboCop

[RuboCop][] is a Ruby code style checker based on this style guide. RuboCop already covers a significant portion of the Guide, supports both MRI 1.9 and MRI 2.0 and has good Emacs integration.

RubyMine

RubyMine's code inspections are partially based on this guide.

License

This work is licensed under a Creative Commons Attribution 3.0 Unported License