parselly 0.1.0 → 1.0.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 4c687b7a170537d06d5a7fa1bbcc374c0826b73c167917cf42a44edd7eea9928
-  data.tar.gz: 2dcce5b4d335f893ad3a30587e264c1c0dd6d0177d2f1a4790b23096b8f0a257
+  metadata.gz: 117ef0c09557018d7129fa29c61565b90432dfeeec3a6b7b8bc9df6f06dcd06a
+  data.tar.gz: edfa4d22bbc8ffe26e9b6118993b0af67cbb4578471f8b2c62a807dcac554e95
 SHA512:
-  metadata.gz: 12982469258ee3f3ce04343948d6a605cbcdb91c87e1af25b3dd2233c1cb29512e89716dcf765c6680909096e78f23a40041ff5b25d753c058b2e34afac82fed
-  data.tar.gz: 7cc9778bd541fc4a4c160d407ce3ee85399efadd7532220053b6e5700e9ae962e95657856cb1ba6c4a12cb955a931446b9dd98918e8c4d39455f0ce2b0b0af8c
+  metadata.gz: 72e2ffff39cc66e2fb68da3d8e69eba7e2435d380d6246f6deb2d2800e75ca64e459cd8e3a029a9f4c30f6548125fbcf7bbafd3ed2efd6e31766052de3aceebb
+  data.tar.gz: bf902d1d2cfcc2f88c5824b8448a8354dfffac62dcdd36e4c33bcff84f2e32f72b9120f674a581d22cab0bcb12e559691a3c507236f67786b94eb34430e255d6

data/.rspec

@@ -0,0 +1 @@
+--require spec_helper

data/README.md

@@ -1,4 +1,4 @@
-# Parselly
+# Parselly [![Gem Version](https://badge.fury.io/rb/parselly.svg)](https://badge.fury.io/rb/parselly) [![CI](https://github.com/ydah/parselly/actions/workflows/test.yml/badge.svg)](https://github.com/ydah/parselly/actions/workflows/test.yml)
 
 Parselly is a module providing a simple way to parse and extract data from a css selector.
 

data/Rakefile

@@ -1,11 +1,18 @@
 # frozen_string_literal: true
 
-require "bundler/gem_tasks"
-require "rake/testtask"
+require 'bundler/gem_tasks'
 
-Rake::TestTask.new(:test) do |t|
-  t.libs << "test/lib"
-  t.test_files = FileList["test/**/test_*.rb"]
+namespace 'build' do
+  desc 'build parser from parser.y'
+  task :parser do
+    sh 'bundle exec racc parser.y --embedded --frozen -o lib/parselly/parser.rb -t --log-file=parser.output'
+  end
 end
 
-task default: :test
+require 'rspec/core/rake_task'
+RSpec::Core::RakeTask.new(:spec) do |spec|
+  spec.pattern = FileList['spec/**/*_spec.rb']
+end
+task spec: 'build:parser'
+
+task default: :spec

data/lib/parselly/lexer.rb

@@ -0,0 +1,148 @@
+# frozen_string_literal: true
+
+require 'strscan'
+
+module Parselly
+  class Lexer
+    TOKENS = {
+      # Combinators
+      '>' => :CHILD,
+      '+' => :ADJACENT,
+      '~' => :SIBLING,
+
+      # Delimiters
+      '[' => :LBRACKET,
+      ']' => :RBRACKET,
+      '(' => :LPAREN,
+      ')' => :RPAREN,
+      ':' => :COLON,
+      ',' => :COMMA,
+      '.' => :DOT,
+      '#' => :HASH,
+      '*' => :STAR,
+      '=' => :EQUAL,
+      '-' => :MINUS,
+
+      # Attribute operators
+      '~=' => :INCLUDES,
+      '|=' => :DASHMATCH,
+      '^=' => :PREFIXMATCH,
+      '$=' => :SUFFIXMATCH,
+      '*=' => :SUBSTRINGMATCH
+    }.freeze
+
+    attr_reader :line, :column
+
+    def initialize(input)
+      @scanner = StringScanner.new(input)
+      @line = 1
+      @column = 1
+      @tokens = []
+    end
+
+    def tokenize
+      until @scanner.eos?
+        skip_whitespace
+        break if @scanner.eos?
+
+        pos = { line: @line, column: @column }
+
+        if (token = scan_string)
+          @tokens << [:STRING, token, pos]
+        elsif (token = scan_number)
+          @tokens << [:NUMBER, token, pos]
+        elsif (token = scan_operator)
+          @tokens << [token, @scanner.matched, pos]
+        elsif (token = scan_identifier)
+          @tokens << [:IDENT, token, pos]
+        else
+          char = @scanner.getch
+          raise "Unexpected character: #{char} at #{pos[:line]}:#{pos[:column]}"
+        end
+      end
+
+      @tokens << [false, nil, { line: @line, column: @column }]
+      @tokens
+    end
+
+    private
+
+    def skip_whitespace
+      while @scanner.scan(/[ \t\n\r]+/)
+        @scanner.matched.each_char do |char|
+          update_position(char)
+        end
+      end
+    end
+
+    def scan_operator
+      # Check multi-character operators first
+      ['~=', '|=', '^=', '$=', '*='].each do |op|
+        if @scanner.scan(/#{Regexp.escape(op)}/)
+          update_position(@scanner.matched)
+          return TOKENS[op]
+        end
+      end
+
+      # Single character operators
+      return unless @scanner.scan(/[>+~\[\]():,.#*=-]/)
+
+      char = @scanner.matched
+      update_position(char)
+      TOKENS[char]
+    end
+
+    # NOTE: Unlike identifiers (where backslash escapes are processed),
+    # escape sequences inside strings (e.g., \n, \", \', \\) are NOT processed.
+    # The raw string content is returned as-is after removing outer quotes.
+    # This is a known limitation for attribute values, as strings are treated
+    # as raw text for simplicity. Identifiers process escapes to support patterns
+    # like .hover\:bg-blue-500, but strings in attributes don't require this.
+    def scan_string
+      if @scanner.scan(/"([^"\\]|\\.)*"/)
+        str = @scanner.matched
+        update_position(str)
+        str[1..-2] # Remove quotes
+      elsif @scanner.scan(/'([^'\\]|\\.)*'/)
+        str = @scanner.matched
+        update_position(str)
+        str[1..-2] # Remove quotes
+      end
+    end
+
+    def scan_identifier
+      # Match identifiers with optional escape sequences
+      # CSS allows \<any-char> as escape in identifiers (e.g., .hover\:bg-blue-500)
+      #
+      # NOTE: This also accepts CSS custom properties starting with -- (e.g., --my-variable).
+      # While custom properties are technically only valid in property contexts (not selectors),
+      # this parser accepts them as a superset of valid CSS for flexibility. In practice,
+      # selectors like .--invalid-class would parse but aren't valid CSS selectors.
+      return unless @scanner.scan(/(?:--|-?[a-zA-Z_])(?:[\w-]|\\[^\n\r\f])*/)
+
+      ident = @scanner.matched
+      update_position(ident)
+      # Remove backslashes from escaped characters
+      ident.gsub(/\\(.)/, '\1')
+    end
+
+    def scan_number
+      return unless @scanner.scan(/\d+(\.\d+)?/)
+
+      num = @scanner.matched
+      update_position(num)
+      num
+    end
+
+    def update_position(text)
+      text.each_char do |char|
+        if char == "\n"
+          @line += 1
+          @column = 1
+        else
+          @column += 1
+        end
+      end
+    end
+  end
+end

data/lib/parselly/node.rb

@@ -0,0 +1,311 @@
+# frozen_string_literal: true
+
+module Parselly
+  # Represents a node in the Abstract Syntax Tree (AST) for CSS selectors.
+  #
+  # Each Node represents a parsed CSS selector component (e.g., type selector,
+  # class selector, combinator, or selector list) with its type, optional value,
+  # child nodes, parent reference, and source position.
+  #
+  # @example Creating a simple AST node
+  #   node = Parselly::Node.new(:type_selector, 'div', { line: 1, column: 1 })
+  #   node.add_child(Parselly::Node.new(:class_selector, 'container'))
+  #
+  # @example Traversing the AST
+  #   node.ancestors    # Returns array of ancestor nodes
+  #   node.descendants  # Returns array of all descendant nodes
+  #   node.siblings     # Returns array of sibling nodes
+  class Node
+    attr_accessor :type, :value, :children, :parent, :position
+
+    # Creates a new AST node.
+    #
+    # @param type [Symbol] the type of the node (e.g., :type_selector, :class_selector)
+    # @param value [String, nil] optional value associated with the node
+    # @param position [Hash] source position with :line and :column keys
+    def initialize(type, value = nil, position = {})
+      @type = type
+      @value = value
+      @children = []
+      @parent = nil
+      @position = position
+    end
+
+    # Adds a child node to this node.
+    #
+    # @param node [Node, nil] the child node to add
+    # @return [Node, nil] the added node, or nil if the input was nil
+    def add_child(node)
+      return nil if node.nil?
+
+      node.parent = self
+      @children << node
+      node
+    end
+
+    # Returns an array of all ancestor nodes from parent to root.
+    #
+    # @return [Array<Node>] array of ancestor nodes
+    def ancestors
+      result = []
+      node = parent
+      while node
+        result << node
+        node = node.parent
+      end
+      result
+    end
+
+    # Returns an array of all descendant nodes (children, grandchildren, etc.).
+    #
+    # @return [Array<Node>] array of all descendant nodes
+    def descendants
+      result = []
+      @children.each do |child|
+        result << child
+        result.concat(child.descendants)
+      end
+      result
+    end
+
+    # Returns an array of sibling nodes (excluding self).
+    #
+    # @return [Array<Node>] array of sibling nodes, or empty array if no parent
+    def siblings
+      return [] unless parent
+
+      parent.children.reject { |child| child == self }
+    end
+
+    # Returns a tree representation of this node and its descendants.
+    #
+    # @param indent [Integer] indentation level for the tree display
+    # @return [String] formatted tree string
+    def to_tree(indent = 0)
+      lines = []
+      prefix = '  ' * indent
+      pos_info = position.empty? ? '' : " [#{position[:line]}:#{position[:column]}]"
+
+      lines << "#{prefix}#{type}#{"(#{value.inspect})" if value}#{pos_info}"
+
+      children.each do |child|
+        lines << child.to_tree(indent + 1)
+      end
+
+      lines.join("\n")
+    end
+
+    def inspect
+      "#<#{self.class.name} type=#{type} value=#{value.inspect} children=#{children.size}>"
+    end
+
+    # Converts the AST node back to a CSS selector string.
+    #
+    # @return [String] the CSS selector string representation of this node
+    def to_selector
+      case type
+      when :selector_list
+        children.map(&:to_selector).join(', ')
+      when :selector
+        children.map(&:to_selector).join
+      when :simple_selector_sequence
+        children.map(&:to_selector).join
+      when :type_selector
+        value
+      when :universal_selector
+        value
+      when :id_selector
+        "##{value}"
+      when :class_selector
+        ".#{value}"
+      when :attribute_selector
+        build_attribute_selector
+      when :pseudo_class
+        ":#{value}"
+      when :pseudo_element
+        "::#{value}"
+      when :pseudo_function
+        ":#{value}(#{children.map(&:to_selector).join})"
+      when :child_combinator
+        ' > '
+      when :adjacent_combinator
+        ' + '
+      when :sibling_combinator
+        ' ~ '
+      when :descendant_combinator
+        ' '
+      when :an_plus_b, :argument
+        value
+      when :attribute, :value
+        value
+      when :equal_operator, :includes_operator, :dashmatch_operator,
+           :prefixmatch_operator, :suffixmatch_operator, :substringmatch_operator
+        value
+      else
+        children.map(&:to_selector).join
+      end
+    end
+
+    # Checks if this node or any descendant contains an ID selector.
+    #
+    # @return [Boolean] true if an ID selector is present
+    def id?
+      type == :id_selector || descendants.any? { |node| node.type == :id_selector }
+    end
+
+    # Extracts the ID value from this node or its descendants.
+    #
+    # @return [String, nil] the ID value without the '#' prefix, or nil if no ID selector is found
+    def id
+      return value if type == :id_selector
+
+      id_node = descendants.find { |node| node.type == :id_selector }
+      id_node&.value
+    end
+
+    # Extracts all class names from this node and its descendants.
+    #
+    # @return [Array<String>] array of class names without the '.' prefix
+    def classes
+      result = []
+      result << value if type == :class_selector
+      descendants.each do |node|
+        result << node.value if node.type == :class_selector
+      end
+      result
+    end
+
+    # Checks if this node or any descendant contains an attribute selector.
+    #
+    # @return [Boolean] true if an attribute selector is present
+    def attribute?
+      type == :attribute_selector || descendants.any? { |node| node.type == :attribute_selector }
+    end
+
+    # Extracts all attribute selectors from this node and its descendants.
+    #
+    # @return [Array<Hash>] array of attribute information hashes
+    #   Each hash contains :name, :operator (optional), and :value (optional) keys
+    def attributes
+      result = []
+
+      if type == :attribute_selector
+        result << extract_attribute_info(self)
+      end
+
+      descendants.each do |node|
+        if node.type == :attribute_selector
+          result << extract_attribute_info(node)
+        end
+      end
+
+      result
+    end
+
+    # Extracts all pseudo-classes and pseudo-elements from this node and its descendants.
+    #
+    # @return [Array<String>] array of pseudo-class and pseudo-element names
+    def pseudo_classes
+      result = []
+
+      if [:pseudo_class, :pseudo_element, :pseudo_function].include?(type)
+        result << value
+      end
+
+      descendants.each do |node|
+        if [:pseudo_class, :pseudo_element, :pseudo_function].include?(node.type)
+          result << node.value
+        end
+      end
+
+      result
+    end
+
+    # Checks if this selector is a compound selector, as defined by CSS.
+    # A compound selector combines multiple simple selectors (type, class, id,
+    # attribute, pseudo-class) without combinators (e.g., `div.class#id[attr]:hover`).
+    # Returns true if more than one simple selector type is present.
+    #
+    # @return [Boolean] true if this node represents a compound selector
+    def compound_selector?
+      types = []
+
+      types << :id if id?
+      types << :class unless classes.empty?
+      types << :attribute if attribute?
+      types << :pseudo unless pseudo_classes.empty?
+      types << :type if type_selector?
+
+      types.size > 1
+    end
+
+    # Checks if this node or any descendant contains a type selector.
+    #
+    # @return [Boolean] true if a type selector is present
+    def type_selector?
+      type == :type_selector || descendants.any? { |node| node.type == :type_selector }
+    end
+
+    private
+
+    # Helper method to extract attribute information from an attribute_selector node.
+    #
+    # @param node [Node] an attribute_selector node
+    # @return [Hash] attribute information hash
+    def extract_attribute_info(node)
+      info = {}
+
+      # Simple attribute selector like [disabled]
+      if node.value
+        info[:name] = node.value
+        return info
+      end
+
+      # Attribute selector with operator and value like [type="text"]
+      node.children.each do |child|
+        case child.type
+        when :attribute
+          info[:name] = child.value
+        when :equal_operator, :includes_operator, :dashmatch_operator,
+             :prefixmatch_operator, :suffixmatch_operator, :substringmatch_operator
+          info[:operator] = child.value
+        when :value
+          info[:value] = child.value
+        end
+      end
+
+      info
+    end
+
+    # Helper method to build an attribute selector string.
+    #
+    # @return [String] the attribute selector string
+    def build_attribute_selector
+      # Simple attribute selector like [disabled]
+      return "[#{value}]" if value
+
+      # Attribute selector with operator and value like [type="text"]
+      attr_name = nil
+      operator = nil
+      attr_value = nil
+
+      children.each do |child|
+        case child.type
+        when :attribute
+          attr_name = child.value
+        when :equal_operator, :includes_operator, :dashmatch_operator,
+             :prefixmatch_operator, :suffixmatch_operator, :substringmatch_operator
+          operator = child.value
+        when :value
+          attr_value = child.value
+        end
+      end
+
+      if operator && attr_value
+        "[#{attr_name}#{operator}\"#{attr_value}\"]"
+      else
+        "[#{attr_name}]"
+      end
+    end
+  end
+end