Cognitive Complexity

Cognitive Complexity is a measure of how difficult a unit of code is to intuitively understand. Unlike Cyclomatic Complexity, which determines how difficult your code will be to test, Cognitive Complexity tells you how difficult your code will be to read and comprehend.

A method's cognitive complexity is based on a few simple rules:

• Code is not considered more complex when it uses shorthand that the language provides for collapsing multiple statements into one
• Code is considered more complex for each "break in the linear flow of the code"
• Code is considered more complex when "flow breaking structures are nested"

Further reading

• Cognitive Complexity docs
• Cognitive Complexity: A new way of measuring understandability

Checks that the cyclomatic complexity of methods is not higher than the configured maximum. The cyclomatic complexity is the number of linearly independent paths through a method. The algorithm counts decision points and adds one.

An if statement (or unless or ?:) increases the complexity by one. An else branch does not, since it doesn't add a decision point. The && operator (or keyword and) can be converted to a nested if statement, and ||/or is shorthand for a sequence of ifs, so they also add one. Loops can be said to have an exit condition, so they add one. Blocks that are calls to builtin iteration methods (e.g. `ary.map{...}) also add one, others are ignored.

def each_child_node(*types)               # count begins: 1
  unless block_given?                     # unless: +1
    return to_enum(__method__, *types)

  children.each do |child|                # each{}: +1
    next unless child.is_a?(Node)         # unless: +1

    yield child if types.empty? ||        # if: +1, ||: +1
                   types.include?(child.type)
  end

  self
end                                       # total: 6

Tries to produce a complexity score that's a measure of the complexity the reader experiences when looking at a method. For that reason it considers when nodes as something that doesn't add as much complexity as an if or a &&. Except if it's one of those special case/when constructs where there's no expression after case. Then the cop treats it as an if/elsif/elsif... and lets all the when nodes count. In contrast to the CyclomaticComplexity cop, this cop considers else nodes as adding complexity.

Example:

def my_method                   # 1
  if cond                       # 1
    case var                    # 2 (0.8 + 4 * 0.2, rounded)
    when 1 then func_one
    when 2 then func_two
    when 3 then func_three
    when 4..10 then func_other
    end
  else                          # 1
    do_something until a && b   # 2
  end                           # ===
end                             # 7 complexity points

Checks that the ABC size of methods is not higher than the configured maximum. The ABC size is based on assignments, branches (method calls), and conditions. See http://c2.com/cgi/wiki?AbcMetric and https://en.wikipedia.org/wiki/ABC_Software_Metric.

Interpreting ABC size:

• <= 17 satisfactory
• 18..30 unsatisfactory
• > 30 dangerous

You can have repeated "attributes" calls count as a single "branch". For this purpose, attributes are any method with no argument; no attempt is meant to distinguish actual attr_reader from other methods.

Example: CountRepeatedAttributes: false (default is true)

# `model` and `current_user`, referenced 3 times each,
 # are each counted as only 1 branch each if
 # `CountRepeatedAttributes` is set to 'false'

 def search
   @posts = model.active.visible_by(current_user)
             .search(params[:q])
   @posts = model.some_process(@posts, current_user)
   @posts = model.another_process(@posts, current_user)

   render 'pages/search/page'
 end

This cop also takes into account AllowedMethods (defaults to []) And AllowedPatterns (defaults to [])

In Ruby 2.4, String#match?, Regexp#match? and Symbol#match? have been added. The methods are faster than match. Because the methods avoid creating a MatchData object or saving backref. So, when MatchData is not used, use match? instead of match.

Example:

# bad
def foo
  if x =~ /re/
    do_something
  end
end

# bad
def foo
  if x.match(/re/)
    do_something
  end
end

# bad
def foo
  if /re/ === x
    do_something
  end
end

# good
def foo
  if x.match?(/re/)
    do_something
  end
end

# good
def foo
  if x =~ /re/
    do_something(Regexp.last_match)
  end
end

# good
def foo
  if x.match(/re/)
    do_something($~)
  end
end

# good
def foo
  if /re/ === x
    do_something($~)
  end
end

In Ruby 2.4, String#match?, Regexp#match? and Symbol#match? have been added. The methods are faster than match. Because the methods avoid creating a MatchData object or saving backref. So, when MatchData is not used, use match? instead of match.

Example:

# bad
def foo
  if x =~ /re/
    do_something
  end
end

# bad
def foo
  if x.match(/re/)
    do_something
  end
end

# bad
def foo
  if /re/ === x
    do_something
  end
end

# good
def foo
  if x.match?(/re/)
    do_something
  end
end

# good
def foo
  if x =~ /re/
    do_something(Regexp.last_match)
  end
end

# good
def foo
  if x.match(/re/)
    do_something($~)
  end
end

# good
def foo
  if /re/ === x
    do_something($~)
  end
end

This cop identifies the use of Regexp#match or String#match, which returns #<MatchData>/nil. The return value of =~ is an integral index/nil and is more performant.

Example:

# bad
do_something if str.match(/regex/)
while regex.match('str')
  do_something
end

# good
method(str =~ /regex/)
return value unless regex =~ 'str'

In Ruby 2.4, String#match?, Regexp#match? and Symbol#match? have been added. The methods are faster than match. Because the methods avoid creating a MatchData object or saving backref. So, when MatchData is not used, use match? instead of match.

Example:

# bad
def foo
  if x =~ /re/
    do_something
  end
end

# bad
def foo
  if x.match(/re/)
    do_something
  end
end

# bad
def foo
  if /re/ === x
    do_something
  end
end

# good
def foo
  if x.match?(/re/)
    do_something
  end
end

# good
def foo
  if x =~ /re/
    do_something(Regexp.last_match)
  end
end

# good
def foo
  if x.match(/re/)
    do_something($~)
  end
end

# good
def foo
  if /re/ === x
    do_something($~)
  end
end

Checks for uses of each_key and each_value Hash methods.

NOTE: If you have an array of two-element arrays, you can put parentheses around the block arguments to indicate that you're not working with a hash, and suppress RuboCop offenses.

Safety:

This cop is unsafe because it cannot be guaranteed that the receiver is a Hash. The AllowedReceivers configuration can mitigate, but not fully resolve, this safety issue.

Example:

# bad
hash.keys.each { |k| p k }
hash.values.each { |v| p v }

# good
hash.each_key { |k| p k }
hash.each_value { |v| p v }

Example: AllowedReceivers: ['execute']

# good
execute(sql).keys.each { |v| p v }
execute(sql).values.each { |v| p v }

Checks for redundant begin blocks.

Currently it checks for code like this:

Example:

# bad
def redundant
  begin
    ala
    bala
  rescue StandardError => e
    something
  end
end

# good
def preferred
  ala
  bala
rescue StandardError => e
  something
end

# bad
begin
  do_something
end

# good
do_something

# bad
# When using Ruby 2.5 or later.
do_something do
  begin
    something
  rescue => ex
    anything
  end
end

# good
# In Ruby 2.5 or later, you can omit `begin` in `do-end` block.
do_something do
  something
rescue => ex
  anything
end

# good
# Stabby lambdas don't support implicit `begin` in `do-end` blocks.
-> do
  begin
    foo
  rescue Bar
    baz
  end
end

Checks for uses of the case equality operator(===).

If AllowOnConstant option is enabled, the cop will ignore violations when the receiver of the case equality operator is a constant. If AllowOnSelfClass option is enabled, the cop will ignore violations when the receiver of the case equality operator is self.class. Note intermediate variables are not accepted.

Example:

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

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

Example: AllowOnConstant: false (default)

# bad
Array === something

Example: AllowOnConstant: true

# good
Array === something

Example: AllowOnSelfClass: false (default)

# bad
self.class === something

Example: AllowOnSelfClass: true

# good
self.class === something

Checks for redundant escapes in string literals.

Example:

# bad - no need to escape # without following {/$/@
"\#foo"

# bad - no need to escape single quotes inside double quoted string
"\'foo\'"

# bad - heredocs are also checked for unnecessary escapes
<<~STR
  \#foo \"foo\"
STR

# good
"#foo"

# good
"\#{no_interpolation}"

# good
"'foo'"

# good
"foo\
bar"

# good
<<~STR
  #foo "foo"
STR

Checks for redundant escapes in string literals.

Example:

# bad - no need to escape # without following {/$/@
"\#foo"

# bad - no need to escape single quotes inside double quoted string
"\'foo\'"

# bad - heredocs are also checked for unnecessary escapes
<<~STR
  \#foo \"foo\"
STR

# good
"#foo"

# good
"\#{no_interpolation}"

# good
"'foo'"

# good
"foo\
bar"

# good
<<~STR
  #foo "foo"
STR

Checks for uses of the case equality operator(===).

If AllowOnConstant option is enabled, the cop will ignore violations when the receiver of the case equality operator is a constant. If AllowOnSelfClass option is enabled, the cop will ignore violations when the receiver of the case equality operator is self.class. Note intermediate variables are not accepted.

Example:

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

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

Example: AllowOnConstant: false (default)

# bad
Array === something

Example: AllowOnConstant: true

# good
Array === something

Example: AllowOnSelfClass: false (default)

# bad
self.class === something

Example: AllowOnSelfClass: true

# good
self.class === something

Checks for a block that is known to need more positional block arguments than are given (by default this is configured for Enumerable methods needing 2 arguments). Optional arguments are allowed, although they don't generally make sense as the default value will be used. Blocks that have no receiver, or take splatted arguments (ie. *args) are always accepted.

Keyword arguments (including **kwargs) do not get counted towards this, as they are not used by the methods in question.

Method names and their expected arity can be configured like this:

Methods:
  inject: 2
  reduce: 2

Safety:

This cop matches for method names only and hence cannot tell apart methods with same name in different classes, which may lead to a false positive.

Example:

# bad
values.reduce {}
values.min { |a| a }
values.sort { |a; b| a + b }

# good
values.reduce { |memo, obj| memo << obj }
values.min { |a, b| a <=> b }
values.sort { |*x| x[0] <=> x[1] }

Checks hash literal syntax.

It can enforce either the use of the class hash rocket syntax or the use of the newer Ruby 1.9 syntax (when applicable).

A separate offense is registered for each problematic pair.

The supported styles are:

• ruby19 - forces use of the 1.9 syntax (e.g. {a: 1}) when hashes have all symbols for keys
• hash_rockets - forces use of hash rockets for all hashes
• nomixedkeys - simply checks for hashes with mixed syntaxes
• ruby19nomixed_keys - forces use of ruby 1.9 syntax and forbids mixed syntax hashes

This cop has EnforcedShorthandSyntax option. It can enforce either the use of the explicit hash value syntax or the use of Ruby 3.1's hash value shorthand syntax.

The supported styles are:

• always - forces use of the 3.1 syntax (e.g. {foo:})
• never - forces use of explicit hash literal value
• either - accepts both shorthand and explicit use of hash literal value
• consistent - forces use of the 3.1 syntax only if all values can be omitted in the hash

Example: EnforcedStyle: ruby19 (default)

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

# good
{a: 2, b: 1}
{:c => 2, 'd' => 2} # acceptable since 'd' isn't a symbol
{d: 1, 'e' => 2} # technically not forbidden

Example: EnforcedStyle: hash_rockets

# bad
{a: 1, b: 2}
{c: 1, 'd' => 5}

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

Example: EnforcedStyle: nomixedkeys

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

# good
{:a => 1, :b => 2}
{c: 1, d: 2}

Example: EnforcedStyle: ruby19nomixed_keys

# bad
{:a => 1, :b => 2}
{c: 2, 'd' => 3} # should just use hash rockets

# good
{a: 1, b: 2}
{:c => 3, 'd' => 4}

Example: EnforcedShorthandSyntax: always (default)

# bad
{foo: foo, bar: bar}

# good
{foo:, bar:}

Example: EnforcedShorthandSyntax: never

# bad
{foo:, bar:}

# good
{foo: foo, bar: bar}

Example: EnforcedShorthandSyntax: either

# good
{foo: foo, bar: bar}

# good
{foo:, bar:}

Example: EnforcedShorthandSyntax: consistent

# bad - `foo` and `bar` values can be omitted
{foo: foo, bar: bar}

# bad - `bar` value can be omitted
{foo:, bar: bar}

# bad - mixed syntaxes
{foo:, bar: baz}

# good
{foo:, bar:}

# good - can't omit `baz`
{foo: foo, bar: baz}