lib/rubocop/cop/util.rb



# frozen_string_literal: true

# rubocop:disable Metrics/ModuleLength
module RuboCop
  module Cop
    # This module contains a collection of useful utility methods.
    module Util
      include PathUtil
      extend RuboCop::AST::Sexp

      BYTE_ORDER_MARK = 0xfeff # The Unicode codepoint

      EQUALS_ASGN_NODES = %i[lvasgn ivasgn cvasgn gvasgn
                             casgn masgn].freeze
      SHORTHAND_ASGN_NODES = %i[op_asgn or_asgn and_asgn].freeze
      ASGN_NODES = (EQUALS_ASGN_NODES + SHORTHAND_ASGN_NODES).freeze

      MODIFIER_NODES = %i[if while until].freeze
      CONDITIONAL_NODES = (MODIFIER_NODES + [:case]).freeze
      LOGICAL_OPERATOR_NODES = %i[and or].freeze

      # http://phrogz.net/programmingruby/language.html#table_18.4
      # Backtick is added last just to help editors parse this code.
      OPERATOR_METHODS = %w(
        | ^ & <=> == === =~ > >= < <= << >>
        + - * / % ** ~ +@ -@ [] []= ! != !~
      ).map(&:to_sym).push(:'`').freeze

      # Match literal regex characters, not including anchors, character
      # classes, alternatives, groups, repetitions, references, etc
      LITERAL_REGEX = /[\w\s\-,"'!#%&<>=;:`~]|\\[^AbBdDgGhHkpPRwWXsSzZ0-9]/

      module_function

      def operator?(symbol)
        OPERATOR_METHODS.include?(symbol)
      end

      def strip_quotes(str)
        if str[0] == '"' || str[0] == "'"
          str[0] = ''
        else
          # we're dealing with %q or %Q
          str[0, 3] = ''
        end
        str[-1] = ''

        str
      end

      def block_length(block_node)
        block_node.loc.end.line - block_node.loc.begin.line
      end

      def comment_line?(line_source)
        line_source =~ /^\s*#/
      end

      def line_range(arg)
        source_range = case arg
                       when Parser::Source::Range
                         arg
                       when Parser::AST::Node
                         arg.source_range
                       else
                         raise ArgumentError, "Invalid argument #{arg}"
                       end

        source_range.begin.line..source_range.end.line
      end

      def parentheses?(node)
        node.loc.respond_to?(:end) && node.loc.end &&
          node.loc.end.is?(')'.freeze)
      end

      def parenthesized_call?(send)
        send.loc.begin && send.loc.begin.is?('(')
      end

      def on_node(syms, sexp, excludes = [], &block)
        return to_enum(:on_node, syms, sexp, excludes) unless block_given?

        yield sexp if Array(syms).include?(sexp.type)
        return if Array(excludes).include?(sexp.type)

        sexp.each_child_node { |elem| on_node(syms, elem, excludes, &block) }
      end

      def source_range(source_buffer, line_number, column, length = 1)
        if column.is_a?(Range)
          column_index = column.begin
          length = column.size
        else
          column_index = column
        end

        line_begin_pos = if line_number.zero?
                           0
                         else
                           source_buffer.line_range(line_number).begin_pos
                         end
        begin_pos = line_begin_pos + column_index
        end_pos = begin_pos + length

        Parser::Source::Range.new(source_buffer, begin_pos, end_pos)
      end

      # Returns the column attribute of the range, except if the range is on
      # the first line and there's a byte order mark at the beginning of that
      # line, in which case 1 is subtracted from the column value. This gives
      # the column as it appears when viewing the file in an editor.
      def effective_column(range)
        if range.line == 1 &&
           @processed_source.raw_source.codepoints.first == BYTE_ORDER_MARK
          range.column - 1
        else
          range.column
        end
      end

      def range_between(start_pos, end_pos)
        Parser::Source::Range.new(processed_source.buffer, start_pos, end_pos)
      end

      def range_with_surrounding_comma(range, side = :both)
        buffer = @processed_source.buffer
        src = buffer.source

        go_left, go_right = directions(side)

        begin_pos = range.begin_pos
        end_pos = range.end_pos
        begin_pos = move_pos(src, begin_pos, -1, go_left, /,/)
        end_pos = move_pos(src, end_pos, 1, go_right, /,/)

        Parser::Source::Range.new(buffer, begin_pos, end_pos)
      end

      def range_with_surrounding_space(range, side = :both, with_newline = true)
        buffer = @processed_source.buffer
        src = buffer.source

        go_left, go_right = directions(side)

        begin_pos = range.begin_pos
        end_pos = range.end_pos
        begin_pos = move_pos(src, begin_pos, -1, go_left, /[ \t]/)
        begin_pos = move_pos(src, begin_pos, -1, go_left && with_newline, /\n/)
        end_pos = move_pos(src, end_pos, 1, go_right, /[ \t]/)
        end_pos = move_pos(src, end_pos, 1, go_right && with_newline, /\n/)
        Parser::Source::Range.new(buffer, begin_pos, end_pos)
      end

      def range_by_whole_lines(range, include_final_newline: false)
        buffer = @processed_source.buffer

        begin_pos = range.begin_pos
        begin_offset = range.column
        begin_of_first_line = begin_pos - begin_offset

        last_line = buffer.source_line(range.last_line)
        end_pos = range.end_pos
        end_offset = last_line.length - range.last_column
        end_offset += 1 if include_final_newline
        end_of_last_line = end_pos + end_offset

        Parser::Source::Range.new(buffer, begin_of_first_line, end_of_last_line)
      end

      def move_pos(src, pos, step, condition, regexp)
        offset = step == -1 ? -1 : 0
        pos += step while condition && src[pos + offset] =~ regexp
        pos < 0 ? 0 : pos
      end

      def directions(side)
        if side == :both
          [true, true]
        else
          [side == :left, side == :right]
        end
      end

      def begins_its_line?(range)
        (range.source_line =~ /\S/) == range.column
      end

      def ends_its_line?(range)
        line = range.source_buffer.source_line(range.last_line)
        (line =~ /\s*\z/) == range.last_column
      end

      def within_node?(inner, outer)
        o = outer.is_a?(AST::Node) ? outer.source_range : outer
        i = inner.is_a?(AST::Node) ? inner.source_range : inner
        i.begin_pos >= o.begin_pos && i.end_pos <= o.end_pos
      end

      # Returns, for example, a bare `if` node if the given node is an `if`
      # with calls chained to the end of it.
      def first_part_of_call_chain(node)
        while node
          case node.type
          when :send
            receiver, _method_name, _args = *node
            node = receiver
          when :block
            method, _args, _body = *node
            node = method
          else
            break
          end
        end
        node
      end

      # If converting a string to Ruby string literal source code, must
      # double quotes be used?
      def double_quotes_required?(string)
        # Double quotes are required for strings which either:
        # - Contain single quotes
        # - Contain non-printable characters, which must use an escape

        # Regex matches IF there is a ' or there is a \\ in the string that is
        # not preceded/followed by another \\ (e.g. "\\x34") but not "\\\\".
        string =~ /'|(?<! \\) \\{2}* \\ (?![\\"])/x
      end

      def needs_escaping?(string)
        double_quotes_required?(escape_string(string))
      end

      def escape_string(string)
        string.inspect[1..-2].tap { |s| s.gsub!(/\\"/, '"') }
      end

      def to_string_literal(string)
        if needs_escaping?(string) && compatible_external_encoding_for?(string)
          string.inspect
        else
          "'#{string.gsub('\\') { '\\\\' }}'"
        end
      end

      def to_symbol_literal(string)
        if symbol_without_quote?(string)
          ":#{string}"
        else
          ":#{to_string_literal(string)}"
        end
      end

      def symbol_without_quote?(string)
        special_gvars = %w[
          $! $" $$ $& $' $* $+ $, $/ $; $: $. $< $= $> $? $@ $\\ $_ $` $~ $0
          $-0 $-F $-I $-K $-W $-a $-d $-i $-l $-p $-v $-w
        ]
        redefinable_operators = %w(
          | ^ & <=> == === =~ > >= < <= << >>
          + - * / % ** ~ +@ -@ [] []= ` ! != !~
        )

        # method name
        string =~ /\A[a-zA-Z_]\w*[!?]?\z/ ||
          # instance / class variable
          string =~ /\A\@\@?[a-zA-Z_]\w*\z/ ||
          # global variable
          string =~ /\A\$[1-9]\d*\z/ ||
          string =~ /\A\$[a-zA-Z_]\w*\z/ ||
          special_gvars.include?(string) ||
          redefinable_operators.include?(string)
      end

      def interpret_string_escapes(string)
        StringInterpreter.interpret(string)
      end

      def same_line?(n1, n2)
        n1.respond_to?(:loc) &&
          n2.respond_to?(:loc) &&
          n1.loc.line == n2.loc.line
      end

      def line_distance(n1, n2)
        n2.loc.line - n1.loc.line
      end

      def preceed?(n1, n2)
        line_distance(n1, n2) == 1
      end

      def stripped_source_upto(line)
        processed_source[0..line].map(&:strip)
      end

      def compatible_external_encoding_for?(src)
        src = src.dup if RUBY_VERSION < '2.3'
        src.force_encoding(Encoding.default_external).valid_encoding?
      end

      def to_supported_styles(enforced_style)
        enforced_style
          .sub(/^Enforced/, 'Supported')
          .sub('Style', 'Styles')
      end
    end
  end
end