nokogiri 1.5.10 → 1.12.5

data/lib/nokogiri/css/parser.y

@@ -10,13 +10,12 @@ rule
         result = [val.first, val.last].flatten
       }
     | prefixless_combinator_selector { result = val.flatten }
-    | simple_selector_1toN { result = val.flatten }
+    | optional_S simple_selector_1toN { result = [val.last].flatten }
     ;
   combinator
     : PLUS { result = :DIRECT_ADJACENT_SELECTOR }
     | GREATER { result = :CHILD_SELECTOR }
     | TILDE { result = :FOLLOWING_SELECTOR }
-    | S { result = :DESCENDANT_SELECTOR }
     | DOUBLESLASH { result = :DESCENDANT_SELECTOR }
     | SLASH { result = :CHILD_SELECTOR }
     ;
@@ -28,17 +27,6 @@ rule
                     Node.new(:CONDITIONAL_SELECTOR, [val.first, val[1]])
                   end
       }
-    | element_name hcap_1toN negation {
-        result = Node.new(:CONDITIONAL_SELECTOR,
-          [
-            val.first,
-            Node.new(:COMBINATOR, [val[1], val.last])
-          ]
-        )
-      }
-    | element_name negation {
-        result = Node.new(:CONDITIONAL_SELECTOR, val)
-      }
     | function
     | function pseudo {
         result = Node.new(:CONDITIONAL_SELECTOR, val)
@@ -46,14 +34,6 @@ rule
     | function attrib {
         result = Node.new(:CONDITIONAL_SELECTOR, val)
       }
-    | hcap_1toN negation {
-        result = Node.new(:CONDITIONAL_SELECTOR,
-          [
-            Node.new(:ELEMENT_NAME, ['*']),
-            Node.new(:COMBINATOR, val)
-          ]
-        )
-      }
     | hcap_1toN {
         result = Node.new(:CONDITIONAL_SELECTOR,
           [Node.new(:ELEMENT_NAME, ['*']), val.first]
@@ -69,10 +49,13 @@ rule
     : simple_selector combinator simple_selector_1toN {
         result = Node.new(val[1], [val.first, val.last])
       }
+    | simple_selector S simple_selector_1toN {
+        result = Node.new(:DESCENDANT_SELECTOR, [val.first, val.last])
+      }
     | simple_selector
     ;
   class
-    : '.' IDENT { result = Node.new(:CLASS_CONDITION, [val[1]]) }
+    : '.' IDENT { result = Node.new(:CLASS_CONDITION, [unescape_css_identifier(val[1])]) }
     ;
   element_name
     : namespaced_ident
@@ -105,7 +88,7 @@ rule
         )
       }
     | LSQUARE NUMBER RSQUARE {
-        # Non standard, but hpricot supports it.
+        # non-standard, from hpricot
         result = Node.new(:PSEUDO_CLASS,
           [Node.new(:FUNCTION, ['nth-child(', val[1]])]
         )
@@ -130,7 +113,7 @@ rule
     | FUNCTION expr RPAREN {
         result = Node.new(:FUNCTION, [val.first.strip, val[1]].flatten)
       }
-    | FUNCTION an_plus_b RPAREN {
+    | FUNCTION nth RPAREN {
         result = Node.new(:FUNCTION, [val.first.strip, val[1]].flatten)
       }
     | NOT expr RPAREN {
@@ -148,14 +131,15 @@ rule
     | STRING
     | IDENT                             # even, odd
       {
-        if val[0] == 'even'
-          val = ["2","n","+","0"]
-          result = Node.new(:AN_PLUS_B, val)
-        elsif val[0] == 'odd'
-          val = ["2","n","+","1"]
-          result = Node.new(:AN_PLUS_B, val)
+        case val[0]
+        when 'even'
+          result = Node.new(:NTH, ['2','n','+','0'])
+        when 'odd'
+          result = Node.new(:NTH, ['2','n','+','1'])
+        when 'n'
+          result = Node.new(:NTH, ['1','n','+','0'])
         else
-          # This is not CSS standard.  It allows us to support this:
+          # non-standard to support custom functions:
           # assert_xpath("//a[foo(., @href)]", @parser.parse('a:foo(@href)'))
           # assert_xpath("//a[foo(., @a, b)]", @parser.parse('a:foo(@a, b)'))
           # assert_xpath("//a[foo(., a, 10)]", @parser.parse('a:foo(a, 10)'))
@@ -163,11 +147,11 @@ rule
         end
       }
     ;
-  an_plus_b
+  nth
     : NUMBER IDENT PLUS NUMBER          # 5n+3 -5n+3
       {
         if val[1] == 'n'
-          result = Node.new(:AN_PLUS_B, val)
+          result = Node.new(:NTH, val)
         else
           raise Racc::ParseError, "parse error on IDENT '#{val[1]}'"
         end
@@ -175,21 +159,27 @@ rule
     | IDENT PLUS NUMBER {               # n+3, -n+3
         if val[0] == 'n'
           val.unshift("1")
-          result = Node.new(:AN_PLUS_B, val)
+          result = Node.new(:NTH, val)
         elsif val[0] == '-n'
           val[0] = 'n'
           val.unshift("-1")
-          result = Node.new(:AN_PLUS_B, val)
+          result = Node.new(:NTH, val)
         else
           raise Racc::ParseError, "parse error on IDENT '#{val[1]}'"
         end
       }
-    | NUMBER IDENT                      # 5n, -5n
-      {
-        if val[1] == 'n'
+    | NUMBER IDENT {                    # 5n, -5n, 10n-1
+        n = val[1]
+        if n[0, 2] == 'n-'
+          val[1] = 'n'
+          val << "-"
+          # b is contained in n as n is the string "n-b"
+          val << n[2, n.size]
+          result = Node.new(:NTH, val)
+        elsif n == 'n'
           val << "+"
           val << "0"
-          result = Node.new(:AN_PLUS_B, val)
+          result = Node.new(:NTH, val)
         else
           raise Racc::ParseError, "parse error on IDENT '#{val[1]}'"
         end
@@ -218,17 +208,22 @@ rule
     | pseudo hcap_1toN {
         result = Node.new(:COMBINATOR, val)
       }
+    | negation hcap_1toN {
+        result = Node.new(:COMBINATOR, val)
+      }
     | attribute_id
     | class
     | attrib
     | pseudo
+    | negation
     ;
   attribute_id
-    : HASH { result = Node.new(:ID, val) }
+    : HASH { result = Node.new(:ID, [unescape_css_identifier(val.first)]) }
     ;
   attrib_val_0or1
-    : eql_incl_dash IDENT { result = [val.first, val[1]] }
-    | eql_incl_dash STRING { result = [val.first, val[1]] }
+    : eql_incl_dash IDENT { result = [val.first, unescape_css_identifier(val[1])] }
+    | eql_incl_dash STRING { result = [val.first, unescape_css_string(val[1])] }
+    | eql_incl_dash NUMBER { result = [val.first, val[1]] }
     |
     ;
   eql_incl_dash
@@ -250,9 +245,28 @@ rule
     | element_name hcap_1toN
     | hcap_1toN
     ;
+  optional_S
+    : S
+    |
+    ;
 end
 
 ---- header
 
-require 'nokogiri/css/parser_extras'
+require_relative "parser_extras"
 
+---- inner
+
+def unescape_css_identifier(identifier)
+  identifier.gsub(/\\(?:([^0-9a-fA-F])|([0-9a-fA-F]{1,6})\s?)/){ |m| $1 || [$2.hex].pack('U') }
+end
+
+def unescape_css_string(str)
+  str.gsub(/\\(?:([^0-9a-fA-F])|([0-9a-fA-F]{1,6})\s?)/) do |m|
+    if $1=="\n"
+      ''
+    else
+      $1 || [$2.hex].pack('U')
+    end
+  end
+end

data/lib/nokogiri/css/parser_extras.rb

@@ -1,63 +1,66 @@
-require 'thread'
+# frozen_string_literal: true
+require "thread"
 
 module Nokogiri
   module CSS
     class Parser < Racc::Parser
-      @cache_on = true
-      @cache    = {}
-      @mutex    = Mutex.new
+      CACHE_SWITCH_NAME = :nokogiri_css_parser_cache_is_off
+
+      @cache = {}
+      @mutex = Mutex.new
 
       class << self
-        # Turn on CSS parse caching
-        attr_accessor :cache_on
-        alias :cache_on? :cache_on
-        alias :set_cache :cache_on=
+        # Return a thread-local boolean indicating whether the CSS-to-XPath cache is active. (Default is `true`.)
+        def cache_on?
+          !Thread.current[CACHE_SWITCH_NAME]
+        end
+
+        # Set a thread-local boolean to turn cacheing on and off. Truthy values turn the cache on, falsey values turn the cache off.
+        def set_cache(value)
+          Thread.current[CACHE_SWITCH_NAME] = !value
+        end
 
         # Get the css selector in +string+ from the cache
-        def [] string
-          return unless @cache_on
+        def [](string)
+          return unless cache_on?
           @mutex.synchronize { @cache[string] }
         end
 
         # Set the css selector in +string+ in the cache to +value+
-        def []= string, value
-          return value unless @cache_on
+        def []=(string, value)
+          return value unless cache_on?
           @mutex.synchronize { @cache[string] = value }
         end
 
         # Clear the cache
-        def clear_cache
-          @mutex.synchronize { @cache = {} }
+        def clear_cache(create_new_object = false)
+          @mutex.synchronize do
+            if create_new_object
+              @cache = {}
+            else
+              @cache.clear
+            end
+          end
         end
 
         # Execute +block+ without cache
-        def without_cache &block
-          tmp = @cache_on
-          @cache_on = false
+        def without_cache(&block)
+          original_cache_setting = cache_on?
+          set_cache false
           block.call
-          @cache_on = tmp
-        end
-
-        ###
-        # Parse this CSS selector in +selector+.  Returns an AST.
-        def parse selector
-          @warned ||= false
-          unless @warned
-            $stderr.puts('Nokogiri::CSS::Parser.parse is deprecated, call Nokogiri::CSS.parse(), this will be removed August 1st or version 1.4.0 (whichever is first)')
-            @warned = true
-          end
-          new.parse selector
+        ensure
+          set_cache original_cache_setting
         end
       end
 
       # Create a new CSS parser with respect to +namespaces+
-      def initialize namespaces = {}
-        @tokenizer  = Tokenizer.new
+      def initialize(namespaces = {})
+        @tokenizer = Tokenizer.new
         @namespaces = namespaces
         super()
       end
 
-      def parse string
+      def parse(string)
         @tokenizer.scan_setup string
         do_parse
       end
@@ -67,14 +70,14 @@ module Nokogiri
       end
 
       # Get the xpath for +string+ using +options+
-      def xpath_for string, options={}
+      def xpath_for(string, options = {})
         key = "#{string}#{options[:ns]}#{options[:prefix]}"
         v = self.class[key]
         return v if v
 
         args = [
-          options[:prefix] || '//',
-          options[:visitor] || XPathVisitor.new
+          options[:prefix] || "//",
+          options[:visitor] || XPathVisitor.new,
         ]
         self.class[key] = parse(string).map { |ast|
           ast.to_xpath(*args)
@@ -82,7 +85,7 @@ module Nokogiri
       end
 
       # On CSS parser error, raise an exception
-      def on_error error_token_id, error_value, value_stack
+      def on_error(error_token_id, error_value, value_stack)
         after = value_stack.compact.last
         raise SyntaxError.new("unexpected '#{error_value}' after '#{after}'")
       end

data/lib/nokogiri/css/syntax_error.rb

@@ -1,4 +1,5 @@
-require 'nokogiri/syntax_error'
+# frozen_string_literal: true
+require_relative "../syntax_error"
 module Nokogiri
   module CSS
     class SyntaxError < ::Nokogiri::SyntaxError

data/lib/nokogiri/css/tokenizer.rb

@@ -1,151 +1,153 @@
+# frozen_string_literal: true
 #--
 # DO NOT MODIFY!!!!
-# This file is automatically generated by rex 1.0.5
+# This file is automatically generated by rex 1.0.7
 # from lexical definition file "lib/nokogiri/css/tokenizer.rex".
 #++
 
 module Nokogiri
 module CSS
 class Tokenizer # :nodoc:
-  require 'strscan'
+      require 'strscan'
 
-  class ScanError < StandardError ; end
+      class ScanError < StandardError ; end
 
-  attr_reader   :lineno
-  attr_reader   :filename
-  attr_accessor :state
+      attr_reader   :lineno
+      attr_reader   :filename
+      attr_accessor :state
 
-  def scan_setup(str)
-    @ss = StringScanner.new(str)
-    @lineno =  1
-    @state  = nil
-  end
+      def scan_setup(str)
+        @ss = StringScanner.new(str)
+        @lineno =  1
+        @state  = nil
+      end
 
-  def action
-    yield
-  end
+      def action
+        yield
+      end
 
-  def scan_str(str)
-    scan_setup(str)
-    do_parse
-  end
-  alias :scan :scan_str
+      def scan_str(str)
+        scan_setup(str)
+        do_parse
+      end
+      alias :scan :scan_str
 
-  def load_file( filename )
-    @filename = filename
-    open(filename, "r") do |f|
-      scan_setup(f.read)
-    end
-  end
+      def load_file( filename )
+        @filename = filename
+        File.open(filename, "r") do |f|
+          scan_setup(f.read)
+        end
+      end
 
-  def scan_file( filename )
-    load_file(filename)
-    do_parse
-  end
+      def scan_file( filename )
+        load_file(filename)
+        do_parse
+      end
 
 
-  def next_token
-    return if @ss.eos?
-    
-    # skips empty actions
-    until token = _next_token or @ss.eos?; end
-    token
-  end
+        def next_token
+          return if @ss.eos?
 
-  def _next_token
-    text = @ss.peek(1)
-    @lineno  +=  1  if text == "\n"
-    token = case @state
-    when nil
-      case
-      when (text = @ss.scan(/has\([\s]*/))
-         action { [:HAS, text] }
+          # skips empty actions
+          until token = _next_token or @ss.eos?; end
+          token
+        end
 
-      when (text = @ss.scan(/[-@]?([_A-Za-z]|[^\0-\177]|\\[0-9A-Fa-f]{1,6}(\r\n|[\s])?|\\[^\n\r\f0-9A-Fa-f])([_A-Za-z0-9-]|[^\0-\177]|\\[0-9A-Fa-f]{1,6}(\r\n|[\s])?|\\[^\n\r\f0-9A-Fa-f])*\([\s]*/))
-         action { [:FUNCTION, text] }
+        def _next_token
+          text = @ss.peek(1)
+          @lineno  +=  1  if text == "\n"
+          token = case @state
+            when nil
+          case
+                  when (text = @ss.scan(/has\([\s]*/))
+                     action { [:HAS, text] }
 
-      when (text = @ss.scan(/[-@]?([_A-Za-z]|[^\0-\177]|\\[0-9A-Fa-f]{1,6}(\r\n|[\s])?|\\[^\n\r\f0-9A-Fa-f])([_A-Za-z0-9-]|[^\0-\177]|\\[0-9A-Fa-f]{1,6}(\r\n|[\s])?|\\[^\n\r\f0-9A-Fa-f])*/))
-         action { [:IDENT, text] }
+                  when (text = @ss.scan(/[-@]?([_A-Za-z]|[^\0-\177]|\\[0-9A-Fa-f]{1,6}(\r\n|[\s])?|\\[^\n\r\f0-9A-Fa-f])([_A-Za-z0-9-]|[^\0-\177]|\\[0-9A-Fa-f]{1,6}(\r\n|[\s])?|\\[^\n\r\f0-9A-Fa-f])*\([\s]*/))
+                     action { [:FUNCTION, text] }
 
-      when (text = @ss.scan(/\#([_A-Za-z0-9-]|[^\0-\177]|\\[0-9A-Fa-f]{1,6}(\r\n|[\s])?|\\[^\n\r\f0-9A-Fa-f])+/))
-         action { [:HASH, text] }
+                  when (text = @ss.scan(/[-@]?([_A-Za-z]|[^\0-\177]|\\[0-9A-Fa-f]{1,6}(\r\n|[\s])?|\\[^\n\r\f0-9A-Fa-f])([_A-Za-z0-9-]|[^\0-\177]|\\[0-9A-Fa-f]{1,6}(\r\n|[\s])?|\\[^\n\r\f0-9A-Fa-f])*/))
+                     action { [:IDENT, text] }
 
-      when (text = @ss.scan(/[\s]*~=[\s]*/))
-         action { [:INCLUDES, text] }
+                  when (text = @ss.scan(/\#([_A-Za-z0-9-]|[^\0-\177]|\\[0-9A-Fa-f]{1,6}(\r\n|[\s])?|\\[^\n\r\f0-9A-Fa-f])+/))
+                     action { [:HASH, text] }
 
-      when (text = @ss.scan(/[\s]*\|=[\s]*/))
-         action { [:DASHMATCH, text] }
+                  when (text = @ss.scan(/[\s]*~=[\s]*/))
+                     action { [:INCLUDES, text] }
 
-      when (text = @ss.scan(/[\s]*\^=[\s]*/))
-         action { [:PREFIXMATCH, text] }
+                  when (text = @ss.scan(/[\s]*\|=[\s]*/))
+                     action { [:DASHMATCH, text] }
 
-      when (text = @ss.scan(/[\s]*\$=[\s]*/))
-         action { [:SUFFIXMATCH, text] }
+                  when (text = @ss.scan(/[\s]*\^=[\s]*/))
+                     action { [:PREFIXMATCH, text] }
 
-      when (text = @ss.scan(/[\s]*\*=[\s]*/))
-         action { [:SUBSTRINGMATCH, text] }
+                  when (text = @ss.scan(/[\s]*\$=[\s]*/))
+                     action { [:SUFFIXMATCH, text] }
 
-      when (text = @ss.scan(/[\s]*!=[\s]*/))
-         action { [:NOT_EQUAL, text] }
+                  when (text = @ss.scan(/[\s]*\*=[\s]*/))
+                     action { [:SUBSTRINGMATCH, text] }
 
-      when (text = @ss.scan(/[\s]*=[\s]*/))
-         action { [:EQUAL, text] }
+                  when (text = @ss.scan(/[\s]*!=[\s]*/))
+                     action { [:NOT_EQUAL, text] }
 
-      when (text = @ss.scan(/[\s]*\)/))
-         action { [:RPAREN, text] }
+                  when (text = @ss.scan(/[\s]*=[\s]*/))
+                     action { [:EQUAL, text] }
 
-      when (text = @ss.scan(/[\s]*\[[\s]*/))
-         action { [:LSQUARE, text] }
+                  when (text = @ss.scan(/[\s]*\)/))
+                     action { [:RPAREN, text] }
 
-      when (text = @ss.scan(/[\s]*\]/))
-         action { [:RSQUARE, text] }
+                  when (text = @ss.scan(/\[[\s]*/))
+                     action { [:LSQUARE, text] }
 
-      when (text = @ss.scan(/[\s]*\+[\s]*/))
-         action { [:PLUS, text] }
+                  when (text = @ss.scan(/[\s]*\]/))
+                     action { [:RSQUARE, text] }
 
-      when (text = @ss.scan(/[\s]*>[\s]*/))
-         action { [:GREATER, text] }
+                  when (text = @ss.scan(/[\s]*\+[\s]*/))
+                     action { [:PLUS, text] }
 
-      when (text = @ss.scan(/[\s]*,[\s]*/))
-         action { [:COMMA, text] }
+                  when (text = @ss.scan(/[\s]*>[\s]*/))
+                     action { [:GREATER, text] }
 
-      when (text = @ss.scan(/[\s]*~[\s]*/))
-         action { [:TILDE, text] }
+                  when (text = @ss.scan(/[\s]*,[\s]*/))
+                     action { [:COMMA, text] }
 
-      when (text = @ss.scan(/\:not\([\s]*/))
-         action { [:NOT, text] }
+                  when (text = @ss.scan(/[\s]*~[\s]*/))
+                     action { [:TILDE, text] }
 
-      when (text = @ss.scan(/-?([0-9]+|[0-9]*\.[0-9]+)/))
-         action { [:NUMBER, text] }
+                  when (text = @ss.scan(/\:not\([\s]*/))
+                     action { [:NOT, text] }
 
-      when (text = @ss.scan(/[\s]*\/\/[\s]*/))
-         action { [:DOUBLESLASH, text] }
+                  when (text = @ss.scan(/-?([0-9]+|[0-9]*\.[0-9]+)/))
+                     action { [:NUMBER, text] }
 
-      when (text = @ss.scan(/[\s]*\/[\s]*/))
-         action { [:SLASH, text] }
+                  when (text = @ss.scan(/[\s]*\/\/[\s]*/))
+                     action { [:DOUBLESLASH, text] }
 
-      when (text = @ss.scan(/U\+[0-9a-f?]{1,6}(-[0-9a-f]{1,6})?/))
-         action {[:UNICODE_RANGE, text] }
+                  when (text = @ss.scan(/[\s]*\/[\s]*/))
+                     action { [:SLASH, text] }
 
-      when (text = @ss.scan(/[\s]+/))
-         action { [:S, text] }
+                  when (text = @ss.scan(/U\+[0-9a-f?]{1,6}(-[0-9a-f]{1,6})?/))
+                     action {[:UNICODE_RANGE, text] }
 
-      when (text = @ss.scan(/"([^\n\r\f"]|\n|\r\n|\r|\f|[^\0-\177]|\\[0-9A-Fa-f]{1,6}(\r\n|[\s])?|\\[^\n\r\f0-9A-Fa-f])*"|'([^\n\r\f']|\n|\r\n|\r|\f|[^\0-\177]|\\[0-9A-Fa-f]{1,6}(\r\n|[\s])?|\\[^\n\r\f0-9A-Fa-f])*'/))
-         action { [:STRING, text] }
+                  when (text = @ss.scan(/[\s]+/))
+                     action { [:S, text] }
 
-      when (text = @ss.scan(/./))
-         action { [text, text] }
+                  when (text = @ss.scan(/"([^\n\r\f"]|\n|\r\n|\r|\f|[^\0-\177]|\\[0-9A-Fa-f]{1,6}(\r\n|[\s])?|\\[^\n\r\f0-9A-Fa-f])*(?<!\\)(?:\\{2})*"|'([^\n\r\f']|\n|\r\n|\r|\f|[^\0-\177]|\\[0-9A-Fa-f]{1,6}(\r\n|[\s])?|\\[^\n\r\f0-9A-Fa-f])*(?<!\\)(?:\\{2})*'/))
+                     action { [:STRING, text] }
 
-      else
-        text = @ss.string[@ss.pos .. -1]
-        raise  ScanError, "can not match: '" + text + "'"
-      end  # if
+                  when (text = @ss.scan(/./))
+                     action { [text, text] }
 
-    else
-      raise  ScanError, "undefined state: '" + state.to_s + "'"
-    end  # case state
-    token
-  end  # def _next_token
+          
+          else
+            text = @ss.string[@ss.pos .. -1]
+            raise  ScanError, "can not match: '" + text + "'"
+          end  # if
+
+        else
+          raise  ScanError, "undefined state: '" + state.to_s + "'"
+        end  # case state
+          token
+        end  # def _next_token
 
 end # class
 end

data/lib/nokogiri/css/tokenizer.rex

@@ -14,8 +14,8 @@ macro
   nmstart   [_A-Za-z]|{nonascii}|{escape}
   ident     [-@]?({nmstart})({nmchar})*
   name      ({nmchar})+
-  string1   "([^\n\r\f"]|{nl}|{nonascii}|{escape})*"
-  string2   '([^\n\r\f']|{nl}|{nonascii}|{escape})*'
+  string1   "([^\n\r\f"]|{nl}|{nonascii}|{escape})*(?<!\\)(?:\\{2})*"
+  string2   '([^\n\r\f']|{nl}|{nonascii}|{escape})*(?<!\\)(?:\\{2})*'
   string    {string1}|{string2}
 
 rule
@@ -34,7 +34,7 @@ rule
             {w}!={w}         { [:NOT_EQUAL, text] }
             {w}={w}          { [:EQUAL, text] }
             {w}\)            { [:RPAREN, text] }
-            {w}\[{w}         { [:LSQUARE, text] }
+            \[{w}            { [:LSQUARE, text] }
             {w}\]            { [:RSQUARE, text] }
             {w}\+{w}         { [:PLUS, text] }
             {w}>{w}          { [:GREATER, text] }
@@ -44,9 +44,9 @@ rule
             {num}            { [:NUMBER, text] }
             {w}\/\/{w}       { [:DOUBLESLASH, text] }
             {w}\/{w}         { [:SLASH, text] }
-            
+
             U\+[0-9a-f?]{1,6}(-[0-9a-f]{1,6})?  {[:UNICODE_RANGE, text] }
-            
+
             [\s]+            { [:S, text] }
             {string}         { [:STRING, text] }
             .                { [text, text] }