# frozen_string_literal: true module RuboCop module Cop module Style # This cop can check for array literals made up of word-like # strings, that are not using the %w() syntax. # # Alternatively, it can check for uses of the %w() syntax, in projects # which do not want to include that syntax. # # NOTE: When using the `percent` style, %w() arrays containing a space # will be registered as offenses. # # Configuration option: MinSize # If set, arrays with fewer elements than this value will not trigger the # cop. For example, a `MinSize` of `3` will not enforce a style on an # array of 2 or fewer elements. # # @example EnforcedStyle: percent (default) # # good # %w[foo bar baz] # # # bad # ['foo', 'bar', 'baz'] # # # bad (contains spaces) # %w[foo\ bar baz\ quux] # # @example EnforcedStyle: brackets # # good # ['foo', 'bar', 'baz'] # # # bad # %w[foo bar baz] # # # good (contains spaces) # ['foo bar', 'baz quux'] class WordArray < Base include ArrayMinSize include ArraySyntax include ConfigurableEnforcedStyle include PercentArray extend AutoCorrector PERCENT_MSG = 'Use `%w` or `%W` for an array of words.' ARRAY_MSG = 'Use `%s` for an array of words.' class << self attr_accessor :largest_brackets end def on_array(node) if bracketed_array_of?(:str, node) return if complex_content?(node.values) check_bracketed_array(node, 'w') elsif node.percent_literal?(:string) check_percent_array(node) end end private def complex_content?(strings, complex_regex: word_regex) strings.any? do |s| next unless s.str_content string = s.str_content.dup.force_encoding(::Encoding::UTF_8) !string.valid_encoding? || (complex_regex && !complex_regex.match?(string)) || / /.match?(string) end end def invalid_percent_array_contents?(node) # Disallow %w() arrays that contain invalid encoding or spaces complex_content?(node.values, complex_regex: false) end def word_regex Regexp.new(cop_config['WordRegex']) end def build_bracketed_array(node) words = node.children.map do |word| if word.dstr_type? string_literal = to_string_literal(word.source) trim_string_interporation_escape_character(string_literal) else to_string_literal(word.children[0]) end end "[#{words.join(', ')}]" end end end end end