class RuboCop::Cop::Style::SpaceInsideHashLiteralBraces

surrounding space depending on configuration.
Checks that braces used for hash literals have or don’t have

def autocorrect(range)

def autocorrect(range)
  lambda do |corrector|
    # It is possible that BracesAroundHashParameters will remove the
    # braces while this cop inserts spaces. This can lead to unwanted
    # changes to the inspected code. If we replace the brace with a
    # brace plus space (rather than just inserting a space), then any
    # removal of the same brace will give us a clobbering error. This
    # in turn will make RuboCop fall back on cop-by-cop
    # auto-correction.  Problem solved.
    case range.source
    when /\s/ then corrector.remove(range)
    when '{' then corrector.replace(range, '{ ')
    else corrector.replace(range, ' }')
    end
  end
end

def check(t1, t2)

def check(t1, t2)
  # No offense if line break inside.
  return if t1.pos.line < t2.pos.line
  return if t2.type == :tCOMMENT # Also indicates there's a line break.
  is_empty_braces = t1.text == '{' && t2.text == '}'
  expect_space = if is_empty_braces
                   cop_config['EnforcedStyleForEmptyBraces'] == 'space'
                 else
                   style == :space
                 end
  if offense?(t1, t2, expect_space)
    incorrect_style_detected(t1, t2, expect_space, is_empty_braces)
  else
    correct_style_detected
  end
end

def incorrect_style_detected(t1, t2, expect_space, is_empty_braces)

def incorrect_style_detected(t1, t2, expect_space, is_empty_braces)
  brace = (t1.text == '{' ? t1 : t2).pos
  range = expect_space ? brace : space_range(brace)
  add_offense(range, range,
              message(brace, is_empty_braces, expect_space)) do
    opposite_style_detected
  end
end

def message(brace, is_empty_braces, expect_space)

def message(brace, is_empty_braces, expect_space)
  inside_what = if is_empty_braces
                  'empty hash literal braces'
                else
                  brace.source
                end
  problem = expect_space ? 'missing' : 'detected'
  format(MSG, "#{inside_what} #{problem}")
end

def offense?(t1, t2, expect_space)

def offense?(t1, t2, expect_space)
  has_space = space_between?(t1, t2)
  expect_space ? !has_space : has_space
end

def on_hash(node)

def on_hash(node)
  b_ix = index_of_first_token(node)
  tokens = processed_source.tokens
  # Hash literal with braces?
  return unless tokens[b_ix].type == :tLBRACE
  e_ix = index_of_last_token(node)
  check(tokens[b_ix], tokens[b_ix + 1])
  check(tokens[e_ix - 1], tokens[e_ix]) unless b_ix == e_ix - 1
end

def range_of_space_to_the_left(range)

def range_of_space_to_the_left(range)
  src = range.source_buffer.source
  begin_pos = range.begin_pos
  begin_pos -= 1 while src[begin_pos - 1] =~ /[ \t]/
  Parser::Source::Range.new(range.source_buffer, begin_pos,
                            range.end_pos - 1)
end

def range_of_space_to_the_right(range)

def range_of_space_to_the_right(range)
  src = range.source_buffer.source
  end_pos = range.end_pos
  end_pos += 1 while src[end_pos] =~ /[ \t]/
  Parser::Source::Range.new(range.source_buffer,
                            range.begin_pos + 1, end_pos)
end

def space_range(token_range)

def space_range(token_range)
  if token_range.source == '{'
    range_of_space_to_the_right(token_range)
  else
    range_of_space_to_the_left(token_range)
  end
end