Ruby

The Rubo config repo is the authoritative source of Ruby style. This exists to cover cases not easily handled by Rubocop. There are likely other rules that are handled by Rubocop, but they are here for vestigial reasons, e.g. we haven’t removed the extraneous rules from this file after we pivoted to using the config repo/gem.

Source Code Layout
Syntax
Naming
Comments
Classes
Exceptions
Collections
Strings
Regular Expressions
Percent Literals
Metaprogramming
Misc
Notes

Source Code Layout

Use two spaces per indentation level (aka soft tabs). No hard tabs.

# bad - four spaces
def some_method
    do_something
end

# good
def some_method
  do_something
end

No blank after class or before last end

  # bad
  class MAndM

    attr_reader :color

    def eat
      do_something
    end

  end

  # good
  class MAndM
    attr_reader :color

    def eat
      do_something
    end
  end

Don’t align assignments/values/blocks

  # bad
  a      = 1
  banana = 2

  # good
  a = 1
  banana = 2

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.

# bad
def too_much; something; something_else; end

# okish - notice that the first ; is required
def no_braces_method; body end

# okish - notice that the second ; is optional
def no_braces_method; body; end

# okish - valid syntax, but no ; makes it kind of hard to read
def some_method() body 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, around { and before }. Whitespace might be (mostly) irrelevant to the Ruby interpreter, but its proper use is the key to writing easily readable code.
```
sum = 1 + 2
a, b = 1, 2
[1, 2, 3].each { |e| puts e }
class FooError < StandardError; end
```
{ and } deserve a bit of clarification, since they are used for block and hash literals, as well as embedded expressions in strings.
```
# good - space after { and before }
{ one: 1, two: 2 }
```
As far as embedded expressions go:
```
# bad
"string#{ expr }"

# good - no spaces
"string#{expr}"
```
No spaces after (, [ or before ], ).
```
some(arg).other
[1, 2, 3].size
```

Indent when as deep as case. I know that many would disagree with this one, but it’s the style established in both “The Ruby Programming Language” and “Programming Ruby”.

# 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.

# 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

# bad
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
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 a method into logical paragraphs internally.

def some_method
  data = initialize(options)

  data.manipulate!

  data.result
end

def some_method
  result
end

Use spaces around the = operator when assigning default values to method parameters:

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

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

Multi-line method chaining. When continuing a chained method invocation on another line keep the . on the first line.
```
  # bad
  one.two.three.
    four

  # good
  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.

# starting point (line is too long)
def send_mail(source)
  Mailer.deliver(to: 'bob@example.com', from: 'us@example.com', subject: 'Important message', body: source.text)
end

# bad (double indent)
def send_mail(source)
  Mailer.deliver(
      to: 'bob@example.com',
      from: 'us@example.com',
      subject: 'Important message',
      body: source.text)
end

# bad
def send_mail(source)
  Mailer.deliver(to: 'bob@example.com',
                 from: 'us@example.com',
                 subject: 'Important message',
                 body: source.text)
end

# good (normal indent)
def send_mail(source)
  Mailer.deliver(
    to: 'bob@example.com',
    from: 'us@example.com',
    subject: 'Important message',
    body: source.text
  )
end

Align the elements of array literals spanning multiple lines.

# bad
fruits = ['Apple', 'Banana',
  'Cantaloupe', 'Dragonfruit']

# bad
fruits =
  ['Apple', 'Banana',
   'Cantaloupe', 'Dragonfruit']

# good
menu_item = [
  'Apple', 'Banana',
  'Cantaloupe', 'Dragonfruit'
]

# good
menu_item = [
  'Apple',
  'Banana',
  'Cantaloupe',
  'Dragonfruit'
]

Add underscores to large numeric literals to improve their readability.
```
# bad
num = 1000000

# good
num = 1_000_000
```
Limit lines to 80 characters.
Avoid trailing whitespace.
- In Sublime, "trim_trailing_white_space_on_save": true
End each file with a newline.
- In Sublime, "ensure_newline_at_eof_on_save": true

Syntax

Use def with parentheses when there are arguments. Omit the parentheses when the method doesn’t accept any arguments.

 # bad
 def some_method()
   # body omitted
 end

 # good
 def some_method
   # body omitted
 end

 # bad
 def some_method_with_arguments arg1, arg2
   # body omitted
 end

 # good
 def some_method_with_arguments(arg1, arg2)
   # body omitted
 end

Favor the ternary operator(?:) over single-line if/then/else/end constructs. It’s more common and obviously more concise.

# 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.

# 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

Leverage the fact that if and case are expressions which return a result.

# 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.
Never use when x; .... See the previous rule.

Use ! instead of not.

# bad - braces are required because of op precedence
x = (not something)

# good
x = !something

Avoid the use of !!.

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

x = false
# double negation is useless on booleans
!!x # => false

# good
x = 'test'
unless x.nil?
  # body omitted
end

The and and or keywords are for control flow only. Otherwise use && and || instead.

# bad - boolean expression
if some_condition and some_other_condition
  do_something
end

# good - control flow
render json: {}, status: 400 and return

Avoid multi-line ?: (the ternary operator); use if/unless instead.
Put the double equals or double ampersand at the end of a multi-line conditional

  # bad
  if some_really_long_condition
    && another_really_really_long_condition
    do_something
  end

  # good
  if some_really_long_condition &&
    another_really_really_long_condition
    do_something
  end

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

# bad
if some_condition
  do_something
end

# good
do_something if some_condition

# another good option
some_condition && do_something

Avoid modifier if/unless usage at the end of a non-trivial multi-line block.

# 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

Favor unless over if for negative conditions (or control flow ||).

# bad
do_something if !some_condition

# bad
do_something if not some_condition

# good
do_something unless some_condition

# another good option
some_condition || do_something

Never use unless with else. Rewrite these with the positive case first.

# 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 an if/unless/while/until.

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

# good
if x > 10
  # body omitted
end

Never use while/until condition do for multi-line while/until.

# 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.

# bad
while some_condition
  do_something
end

# good
do_something while some_condition

Favor until over while for negative conditions.

# bad
do_something while !some_condition

# good
do_something until some_condition

Use Kernel#loop instead of while/until when you need an infinite loop.

  # 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.

# bad
begin
  puts val
  val += 1
end while val < 0

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

Omit parentheses around parameters for methods that are part of an internal DSL (e.g. Rake, Rails, RSpec), methods that have “keyword” status in Ruby (e.g. attr_reader, puts) and attribute access methods. Use parentheses around the arguments of all other method invocations.
```
class Person
  attr_reader :name, :age

  # omitted
end

temperance = Person.new('Temperance', 30)
temperance.name

puts temperance.age

x = Math.sin(y)
array.delete(e)

bowling.score.should == 0
```

Omit the outer braces around an implicit options hash.

# 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.

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

  # good
  validates :name, presence: true, length: { within: 1..10 }
end

Omit parentheses for method calls with no arguments.

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

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

Prefer {...} over do...end for single-line blocks. Avoid using {...} for multi-line blocks (multiline 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.
```
names = ['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 { |name| name.upcase }
```
Some will argue that multiline 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.

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.

# 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.)

# 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.

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

Wrap assignment in parentheses when using the return value of = in conditional expressions.

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

# good (MRI would still complain, but RuboCop won't)
if (v = array.grep(/foo/))
  do_something(v)
  ...
end

# good
v = array.grep(/foo/)
if v
  do_something(v)
  ...
end

Use shorthand self assignment operators whenever applicable.

# 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.

# 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.)
```
# 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.

# bad
if something
  something = something.downcase
end

# bad
something = something ? nil : something.downcase

# 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.
```
# bad
Array === something
(1..100) === 7
/something/ === some_string

# good
something.is_a?(Array)
(1..100).include?(7)
some_string =~ /something/
```
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.
```
# bad
$:.unshift File.dirname(__FILE__)

# good
require 'English'
$LOAD_PATH.unshift File.dirname(__FILE__)
```
Never put a space between a method name and the opening parenthesis.
```
# bad
f (3 + 2) + 1

# good
f(3 + 2) + 1
```
If the first argument to a method begins with an open parenthesis, always use parentheses in the method invocation. For example, write 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!

Use the new lambda literal syntax for single line body blocks. Use the lambda method for multi-line blocks.

# 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

Prefer proc over Proc.new.

# 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.

# 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.

# bad
result = hash.map { |k, v| v + 1 }

def something(x)
  unused_var, used_var = something_else(x)
  # ...
end

# good
result = hash.map { |_k, v| v + 1 }

def something(x)
  _unused_var, used_var = something_else(x)
  # ...
end

# good
result = hash.map { |_, v| v + 1 }

def something(x)
  _, used_var = something_else(x)
  # ...
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.
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).

Favor the use of sprintf and its alias format over the fairly cryptic String#% method.

# bad
'%d %d' % [20, 10]
# => '20 10'

# good
sprintf('%d %d', 20, 10)
# => '20 10'

# good
sprintf('%{first} %{second}', first: 20, second: 10)
# => '20 10'

format('%d %d', 20, 10)
# => '20 10'

# good
format('%{first} %{second}', first: 20, second: 10)
# => '20 10'

Favor the use of Array#join over the fairly cryptic Array#* with a string argument.

# bad
%w(one two three) * ', '
# => 'one, two, three'

# good
%w(one two three).join(', ')
# => 'one, two, three'

Use [*var] or Array() instead of explicit Array check, when dealing with a variable you want to treat as an Array, but you’re not certain it’s an array.

# bad
paths = [paths] unless paths.is_a? Array
paths.each { |path| do_something(path) }

# good
[*paths].each { |path| do_something(path) }

# good (and a bit more readable)
Array(paths).each { |path| do_something(path) }

Use ranges or Comparable#between? instead of complex comparison logic when possible.

# 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.

# 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.

  # 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.

Never use END blocks. Use Kernel#at_exit instead.

# bad
END { puts 'Goodbye!' }

# good
at_exit { puts 'Goodbye!' }

Avoid the use of flip-flops.

Avoid use of nested conditionals for flow of control.

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)
    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

Naming

Name identifiers in English.

# 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.

# bad
:'some symbol'
:SomeSymbol
:someSymbol

someVar = 5

def someMethod
  ...
end

def SomeMethod
 ...
end

# good
:some_symbol

def some_method
  ...
end

Use CamelCase for classes and modules. (Keep acronyms like HTTP, RFC, XML uppercase.)

# bad
class Someclass
  ...
end

class Some_Class
  ...
end

class SomeXml
  ...
end

# good
class SomeClass
  ...
end

class SomeXML
  ...
end

Use snake_case for naming files, e.g. hello_world.rb.
Use snake_case for naming directories, e.g. lib/hello_world/hello_world.rb.
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.

Use SCREAMING_SNAKE_CASE for other constants.

# 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.
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.
```
# 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.

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 using reduce with short blocks, name the arguments |a, e| (accumulator, element).
When defining binary operators, name the argument other(<< and [] are exceptions to the rule, since their semantics are different).
```
def +(other)
  # body omitted
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.
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.
```
# 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.
```
# bad
all_songs = users.map(&:songs).flatten.uniq

# good
all_songs = users.flat_map(&:songs).uniq
```
Use reverse_each instead of reverse.each. reverse_each doesn’t do a new array allocation and that’s a good thing.
```
# bad
array.reverse.each { ... }

# good
array.reverse_each { ... }
```

Comments

Write self-documenting code and ignore the rest of this section. Seriously!
Write comments in English.
Use one space between the leading # character of the comment and the text of the comment.
Comments longer than a word are capitalized and use punctuation. Use one space after periods.

Avoid superfluous comments.

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

Keep existing comments up-to-date. An outdated comment is worse than no comment at all.
Avoid writing comments to explain bad code. Refactor the code to make it self-explanatory. (Do or do not - there is no try. –Yoda)

Classes & Modules

Use a consistent structure in your class definitions.

class Person
  # extend and include go first
  extend SomeModule
  include AnotherModule

  # inner classes
  CustomErrorKlass = 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

  # followed by 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

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.

# 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

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

# bad
class SomeClass
  def self.some_method
    # body omitted
  end

  def self.some_other_method
  end
end

# good
module SomeClass
  module_function

  def some_method
    # body omitted
  end

  def some_other_method
  end
end

Favor the use of module_function over extend self when you want to turn a module’s instance methods into class methods.

# 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.
Try to make your classes as SOLID as possible.

Always supply a proper to_s method for classes that represent domain objects.

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.

# 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

Avoid the use of attr. Use attr_reader and attr_accessor instead.

# bad - creates a single attribute accessor (deprecated in 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.

# 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 a Struct.new - it already is a new class. Extending it introduces a superfluous class level and may also introduce weird errors if the file is required multiple times.
Consider adding factory methods to provide additional sensible ways to create instances of a particular class.
```
class Person
  def self.create(options_hash)
    # body omitted
  end
end
```

Prefer duck-typing over inheritance.

# 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.
```
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.
Indent the public, protected, and private methods as much 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.
```
class SomeClass
  def public_method
    # ...
  end

  private

  def private_method
    # ...
  end

  def another_private_method
    # ...
  end
end
```

Use def self.method to define singleton methods. This makes the code easier to refactor since the class name is not repeated.

class TestClass
  # bad
  def TestClass.some_method
    # body omitted
  end

  # good
  def self.some_other_method
    # body omitted
  end

  # Also possible and convenient when you
  # have to define many singleton methods.
  class << self
    def first_method
      # body omitted
    end

    def second_method_etc
      # body omitted
    end
  end
end

Exceptions

Signal exceptions using the fail method. Use raise only when catching an exception and re-raising it (because here you’re not failing, but explicitly and purposefully raising an exception).
```
begin
  fail 'Oops'
rescue => error
  raise if error.message != 'Oops'
end
```

Don’t specify RuntimeError explicitly in the two argument version of fail/raise.

# bad
fail RuntimeError, 'message'

# good - signals a RuntimeError by default
fail 'message'

Prefer supplying an exception class and a message as two separate arguments to fail/raise, instead of an exception instance.

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

# good
fail SomeException, 'message'
# Consistent with `fail SomeException, 'message', backtrace`.

Never 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.
```
def foo
  begin
    fail
  ensure
    return 'very bad idea'
  end
end
```

Use implicit begin blocks where possible.

# 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).

# 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.

# 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.

# 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.

# 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.

# 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.

# bad
begin
  # some code
rescue Exception => e
  # some handling
rescue StandardError => e
  # some handling
end

# good
begin
  # some code
rescue StandardError => e
  # some handling
rescue Exception => e
  # some handling
end

Release external resources obtained by your program in an ensure block.

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

Favor the use of exceptions for the standard library over introducing new exception classes.

Collections

Prefer literal array and hash creation notation (unless you need to pass parameters to their constructors, that is).
```
# bad
arr = Array.new
hash = Hash.new

# good
arr = []
hash = {}
```
Prefer %w to the literal array syntax when you need an array of words (non-empty strings without spaces and special characters in them). Apply this rule only to arrays with two or more elements.
```
# bad
STATES = ['draft', 'open', 'closed']

# good
STATES = %w(draft open closed)
```
Prefer %i to the literal array syntax when you need an array of symbols (and you don’t need to maintain Ruby 1.9 compatibility). Apply this rule only to arrays with two or more elements.
```
# bad
STATES = [:draft, :open, :closed]

# good
STATES = %i(draft open closed)
```

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

# 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.

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].
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.

Prefer symbols instead of strings as hash keys.

# 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.

Use the Ruby 1.9 hash literal syntax when your hash keys are symbols.

# 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.
```
# 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?. As noted here by Matz, the longer forms are considered deprecated.
```
# bad
hash.has_key?(:test)
hash.has_value?(value)

# good
hash.key?(:test)
hash.value?(value)
```

Use Hash#fetch when dealing with hash keys that should be present.

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.

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 work correctly with falsy values
batman.fetch(:is_evil, true) # => false

Prefer the use of the block instead of the default value in Hash#fetch.

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, get_batman_powers) # get_batman_powers is an expensive call

# good - blocks are lazy evaluated, so only triggered in case of KeyError exception
batman.fetch(:powers) { get_batman_powers }

Use Hash#values_at when you need to retrieve several values consecutively from a hash.

# bad
email = data['email']
nickname = data['nickname']

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

Rely on the fact that as of Ruby 1.9 hashes are ordered.
Never modify a collection while traversing it.

Strings

Prefer string interpolation and string formatting instead of string concatenation:

# bad
email_with_name = user.name + ' <' + user.email + '>'

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

# good
email_with_name = format('%s <%s>', user.name, user.email)

Prefer double-quotes unless your string literal contains " or escape characters you want to suppress.
```
  # bad
  name = 'Bozhidar'

  # good
  name = "Bozhidar"
```
Don’t use the character literal syntax ?x. Since Ruby 1.9 it’s basically redundant - ?x would interpreted as 'x' (a string with a single character in it).
```
# bad
char = ?c

# good
char = 'c'
```

Don’t leave out {} around instance and global variables being interpolated into a string.

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.

# 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.
```
# good and also fast
html = ''
html << '<h1>Page title</h1>'

paragraphs.each do |paragraph|
  html << "<p>#{paragraph}</p>"
end
```
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.
```
code = <<-END.gsub(/^\s+\|/, '')
  |def test
  |  some_method
  |  other_method
  |end
END
#=> "def test\n  some_method\n  other_method\nend\n"
```

Regular Expressions

Use \A and \z unless need to match against something other than the entire string

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

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

For simple constructions you can use regexp directly through string index.

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 captured result of parentheses.
```
/(first|second)/   # bad
/(?:first|second)/ # good
```
Don’t use the cryptic Perl-legacy variables denoting last regexp group matches ($1, $2, etc). Use Regexp.last_match[n] instead.
```
/(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.

# bad
/(regexp)/ =~ string
...
process Regexp.last_match[1]

# good
/(?<meaningful_var>regexp)/ =~ string
...
process meaningful_var

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

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.

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 block or hash.

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.

# 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 %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.
```
# 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?"'
```

Use %r only for regular expressions matching more than one ‘/’ character.

# bad
%r(\s+)

# still bad
%r(^/(.*)$)
# should be /^\/(.*)$/

# good
%r(^/blog/2011/(.*)$)

Avoid the use of %x unless you’re going to invoke a command with backquotes in it(which is rather unlikely).
```
# 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.
Prefer () as delimiters for all % literals, except %r. Since braces often appear inside regular expressions in many scenarios a less common character like { might be a better choice for a delimiter, depending on the regexp’s content.
```
# bad
%w[one two three]
%q{"Test's king!", John said.}

# good
%w(one two three)
%q("Test's king!", John said.)
```

Metaprogramming

Avoid needless metaprogramming.
Do not mess around in core classes when writing libraries. (Do not monkey-patch them.)

Misc

Avoid warnings whenever possible.
Avoid boolean parameters.
Avoid hashes as optional parameters. Does the method do too much? (Object initializers are exceptions for this rule).
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.
Avoid parameter lists longer than three or four parameters.
If you really need “global” methods, add them to Kernel and make them private.

Use module instance variables instead of global variables.

# bad
$foo_bar = 1

# good
module Foo
  class << self
    attr_accessor :bar
  end
end

Foo.bar = 1

Avoid alias when alias_method will do.
Use OptionParser for parsing complex command line options and ruby -s for trivial command line options.
Prefer Time.now over Time.new when retrieving the current system time.
Code in a functional way, avoiding mutation when that makes sense.
Do not mutate arguments unless that is the purpose of the method.
Avoid more than three levels of block nesting.
Be consistent. In an ideal world, be consistent with these guidelines.
Use common sense.

Notes

Forked from https://github.com/bbatsov/ruby-style-guide/blob/master/README.md

Rails

In application code, do not use ActiveRecord::Base#update_column, ActiveRecord::Base#update_columns, or ActiveRecord::Relation#update_all without a comment explaining why it is necessary. Those methods do not hit validations or run callbacks. [reference link]

working-agreements

CoverHound's guide to working together