regexp_parser 2.2.1 → 2.3.0

data/lib/regexp_parser/syntax/any.rb

@@ -3,11 +3,8 @@ module Regexp::Syntax
   # is useful during development, testing, and should be useful for some types
   # of transformations as well.
   class Any < Base
-    def initialize # rubocop:disable Lint/MissingSuper
-      @implements = { :* => %i[*] }
-    end
+    implements :*, [:*]
 
-    def implements?(_type, _token) true end
-    def implements!(_type, _token) true end
+    def self.implements?(_type, _token) true end
   end
 end

data/lib/regexp_parser/syntax/base.rb

@@ -1,7 +1,7 @@
 module Regexp::Syntax
   class NotImplementedError < Regexp::Syntax::SyntaxError
     def initialize(syntax, type, token)
-      super "#{syntax.class.name} does not implement: [#{type}:#{token}]"
+      super "#{syntax} does not implement: [#{type}:#{token}]"
     end
   end
 
@@ -9,85 +9,112 @@ module Regexp::Syntax
   class Base
     include Regexp::Syntax::Token
 
-    def initialize
-      @implements = {}
+    class << self
+      attr_accessor :features
 
-      implements Token::Literal::Type,   Token::Literal::All
-      implements Token::FreeSpace::Type, Token::FreeSpace::All
-    end
+      # automatically inherit features through the syntax class hierarchy
+      def inherited(subclass)
+        super
+        subclass.features = features.to_h.map { |k, v| [k, v.dup] }.to_h
+      end
 
-    def features
-      @implements
-    end
+      def implements(type, tokens)
+        (features[type] ||= []).concat(tokens)
+        added_features[type] = tokens
+      end
 
-    def implementations(type)
-      @implements[type] ||= []
-    end
+      def excludes(type, tokens)
+        tokens.each { |tok| features[type].delete(tok) }
+        removed_features[type] = tokens
+      end
 
-    def implements(type, tokens)
-      implementations(type).concat(Array(tokens))
-    end
+      def implements?(type, token)
+        implementations(type).include?(token)
+      end
+      alias :check? :implements?
 
-    def excludes(type, tokens)
-      Array(tokens).each { |tok| implementations(type).delete(tok) }
-    end
+      def implementations(type)
+        features[type] || []
+      end
 
-    def implements?(type, token)
-      implementations(type).include?(token)
-    end
-    alias :check? :implements?
+      def implements!(type, token)
+        raise NotImplementedError.new(self, type, token) unless
+          implements?(type, token)
+      end
+      alias :check! :implements!
 
-    def implements!(type, token)
-      raise NotImplementedError.new(self, type, token) unless
-        implements?(type, token)
-    end
-    alias :check! :implements!
-
-    def normalize(type, token)
-      case type
-      when :group
-        normalize_group(type, token)
-      when :backref
-        normalize_backref(type, token)
-      else
-        [type, token]
+      def added_features
+        @added_features ||= {}
       end
-    end
 
-    def normalize_group(type, token)
-      case token
-      when :named_ab, :named_sq
-        %i[group named]
-      else
-        [type, token]
+      def removed_features
+        @removed_features ||= {}
       end
+
+      def normalize(type, token)
+        case type
+        when :group
+          normalize_group(type, token)
+        when :backref
+          normalize_backref(type, token)
+        else
+          [type, token]
+        end
+      end
+
+      def normalize_group(type, token)
+        case token
+        when :named_ab, :named_sq
+          %i[group named]
+        else
+          [type, token]
+        end
+      end
+
+      def normalize_backref(type, token)
+        case token
+        when :name_ref_ab, :name_ref_sq
+          %i[backref name_ref]
+        when :name_call_ab, :name_call_sq
+          %i[backref name_call]
+        when :name_recursion_ref_ab, :name_recursion_ref_sq
+          %i[backref name_recursion_ref]
+        when :number_ref_ab, :number_ref_sq
+          %i[backref number_ref]
+        when :number_call_ab, :number_call_sq
+          %i[backref number_call]
+        when :number_rel_ref_ab, :number_rel_ref_sq
+          %i[backref number_rel_ref]
+        when :number_rel_call_ab, :number_rel_call_sq
+          %i[backref number_rel_call]
+        when :number_recursion_ref_ab, :number_recursion_ref_sq
+          %i[backref number_recursion_ref]
+        else
+          [type, token]
+        end
+      end
+    end
+
+    # TODO: drop this backwards compatibility code in v3.0.0, do `private :new`
+    def initialize
+      warn 'Using instances of Regexp::Parser::Syntax is deprecated ' \
+           "and will no longer be supported in v3.0.0."
     end
 
-    def normalize_backref(type, token)
-      case token
-      when :name_ref_ab, :name_ref_sq
-        %i[backref name_ref]
-      when :name_call_ab, :name_call_sq
-        %i[backref name_call]
-      when :name_recursion_ref_ab, :name_recursion_ref_sq
-        %i[backref name_recursion_ref]
-      when :number_ref_ab, :number_ref_sq
-        %i[backref number_ref]
-      when :number_call_ab, :number_call_sq
-        %i[backref number_call]
-      when :number_rel_ref_ab, :number_rel_ref_sq
-        %i[backref number_rel_ref]
-      when :number_rel_call_ab, :number_rel_call_sq
-        %i[backref number_rel_call]
-      when :number_recursion_ref_ab, :number_recursion_ref_sq
-        %i[backref number_recursion_ref]
+    def method_missing(name, *args)
+      if self.class.respond_to?(name)
+        warn 'Using instances of Regexp::Parser::Syntax is deprecated ' \
+             "and will no longer be supported in v3.0.0. Please call "\
+             "methods on the class directly, e.g.: #{self.class}.#{name}"
+        self.class.send(name, *args)
       else
-        [type, token]
+        super
       end
     end
 
-    def self.inspect
-      "#{super} (feature set of #{ancestors[1].to_s.split('::').last})"
+    def respond_to_missing?(name, include_private = false)
+      self.class.respond_to?(name) || super
     end
+    # end of backwards compatibility code
   end
 end

data/lib/regexp_parser/syntax/token/quantifier.rb

@@ -1,7 +1,7 @@
 module Regexp::Syntax
   module Token
     module Quantifier
-      Greedy    = %i[
+      Greedy = %i[
         zero_or_one
         zero_or_more
         one_or_more
@@ -13,7 +13,7 @@ module Regexp::Syntax
         one_or_more_reluctant
       ]
 
-      Possessive  = %i[
+      Possessive = %i[
         zero_or_one_possessive
         zero_or_more_possessive
         one_or_more_possessive
@@ -23,9 +23,9 @@ module Regexp::Syntax
       IntervalReluctant    = %i[interval_reluctant]
       IntervalPossessive   = %i[interval_possessive]
 
-      IntervalAll = Interval + IntervalReluctant +
-                    IntervalPossessive
+      IntervalAll = Interval + IntervalReluctant + IntervalPossessive
 
+      V1_8_6 = Greedy + Reluctant + Interval + IntervalReluctant
       All = Greedy + Reluctant + Possessive + IntervalAll
       Type = :quantifier
     end

data/lib/regexp_parser/syntax/token/unicode_property.rb

@@ -59,6 +59,8 @@ module Regexp::Syntax
 
       Age_V3_1_0 = %i[age=13.0]
 
+      Age_V3_2_0 = %i[age=14.0]
+
       Age = all[:Age_V]
 
       Derived_V1_9_0 = %i[
@@ -322,6 +324,14 @@ module Regexp::Syntax
         yezidi
       ]
 
+      Script_V3_2_0 = %i[
+        cypro_minoan
+        old_uyghur
+        tangsa
+        toto
+        vithkuqi
+      ]
+
       Script = all[:Script_V]
 
       UnicodeBlock_V1_9_0 = %i[
@@ -660,6 +670,21 @@ module Regexp::Syntax
         in_yezidi
       ]
 
+      UnicodeBlock_V3_2_0 = %i[
+        in_arabic_extended_b
+        in_cypro_minoan
+        in_ethiopic_extended_b
+        in_kana_extended_b
+        in_latin_extended_f
+        in_latin_extended_g
+        in_old_uyghur
+        in_tangsa
+        in_toto
+        in_unified_canadian_aboriginal_syllabics_extended_a
+        in_vithkuqi
+        in_znamenny_musical_notation
+      ]
+
       UnicodeBlock = all[:UnicodeBlock_V]
 
       Emoji_V2_5_0 = %i[
@@ -683,6 +708,7 @@ module Regexp::Syntax
       V2_6_2 = all[:V2_6_2]
       V2_6_3 = all[:V2_6_3]
       V3_1_0 = all[:V3_1_0]
+      V3_2_0 = all[:V3_2_0]
 
       All = all[/^V\d+_\d+_\d+$/]
 

data/lib/regexp_parser/syntax/version_lookup.rb

@@ -17,29 +17,31 @@ module Regexp::Syntax
 
   module_function
 
-  # Loads and instantiates an instance of the syntax specification class for
-  # the given syntax version name. The special names 'any' and '*' return an
-  # instance of Syntax::Any.
+  # Returns the syntax specification class for the given syntax
+  # version name. The special names 'any' and '*' return Syntax::Any.
+  def for(name)
+    (@alias_map ||= {})[name] ||= version_class(name)
+  end
+
   def new(name)
-    return Regexp::Syntax::Any.new if ['*', 'any'].include?(name.to_s)
-    version_class(name).new
+    warn 'Regexp::Syntax.new is deprecated in favor of Regexp::Syntax.for. '\
+         'It does not return distinct instances and will be removed in v3.0.0.'
+    self.for(name)
   end
 
   def supported?(name)
-    name =~ VERSION_REGEXP &&
-      comparable_version(name) >= comparable_version('1.8.6')
+    name =~ VERSION_REGEXP && comparable(name) >= comparable('1.8.6')
   end
 
   def version_class(version)
+    return Regexp::Syntax::Any if ['*', 'any'].include?(version.to_s)
+
     version =~ VERSION_REGEXP || raise(InvalidVersionNameError, version)
-    version_const_name = version_const_name(version)
+    warn_if_future_version(version)
+    version_const_name = "V#{version.to_s.scan(/\d+/).join('_')}"
     const_get(version_const_name) || raise(UnknownSyntaxNameError, version)
   end
 
-  def version_const_name(version_string)
-    "V#{version_string.to_s.scan(/\d+/).join('_')}"
-  end
-
   def const_missing(const_name)
     if const_name =~ VERSION_CONST_REGEXP
       return fallback_version_class(const_name)
@@ -48,35 +50,24 @@ module Regexp::Syntax
   end
 
   def fallback_version_class(version)
-    sorted_versions = (specified_versions + [version])
-                      .sort_by { |name| comparable_version(name) }
-    return if (version_index = sorted_versions.index(version)) < 1
-
-    next_lower_version = sorted_versions[version_index - 1]
-    inherit_from_version(next_lower_version, version)
-  end
-
-  def inherit_from_version(parent_version, new_version)
-    new_const = version_const_name(new_version)
-    parent = const_get(version_const_name(parent_version))
-    const_defined?(new_const) || const_set(new_const, Class.new(parent))
-    warn_if_future_version(new_const)
-    const_get(new_const)
+    sorted = (specified_versions + [version]).sort_by { |ver| comparable(ver) }
+    index = sorted.index(version)
+    index > 0 && const_get(sorted[index - 1])
   end
 
   def specified_versions
     constants.select { |const_name| const_name =~ VERSION_CONST_REGEXP }
   end
 
-  def comparable_version(name)
+  def comparable(name)
     # add .99 to treat versions without a patch value as latest patch version
     Gem::Version.new((name.to_s.scan(/\d+/) << 99).join('.'))
   end
 
   def warn_if_future_version(const_name)
-    return if comparable_version(const_name) < comparable_version('4.0.0')
+    return if comparable(const_name) < comparable('4.0.0')
 
     warn('This library has only been tested up to Ruby 3.x, '\
-         "but you are running with #{const_get(const_name).inspect}")
+         "but you are running with #{const_name}")
   end
 end

data/lib/regexp_parser/syntax/versions/1.8.6.rb

@@ -1,21 +1,14 @@
-module Regexp::Syntax
-  class V1_8_6 < Regexp::Syntax::Base
-    def initialize
-      super
-
-      implements :anchor, Anchor::All
-      implements :assertion, Assertion::Lookahead
-      implements :backref, Backreference::V1_8_6
-      implements :posixclass, PosixClass::Standard
-      implements :group, Group::V1_8_6
-      implements :meta, Meta::Extended
-      implements :set, CharacterSet::All
-      implements :type, CharacterType::Extended
-      implements :escape,
-        Escape::Basic + Escape::ASCII + Escape::Meta + Escape::Control
-      implements :quantifier,
-        Quantifier::Greedy + Quantifier::Reluctant +
-        Quantifier::Interval + Quantifier::IntervalReluctant
-    end
-  end
+class Regexp::Syntax::V1_8_6 < Regexp::Syntax::Base
+  implements :anchor,     Anchor::All
+  implements :assertion,  Assertion::Lookahead
+  implements :backref,    Backreference::V1_8_6
+  implements :escape,     Escape::Basic + Escape::ASCII + Escape::Meta + Escape::Control
+  implements :free_space, FreeSpace::All
+  implements :group,      Group::V1_8_6
+  implements :literal,    Literal::All
+  implements :meta,       Meta::Extended
+  implements :posixclass, PosixClass::Standard
+  implements :quantifier, Quantifier::V1_8_6
+  implements :set,        CharacterSet::All
+  implements :type,       CharacterType::Extended
 end

data/lib/regexp_parser/syntax/versions/1.9.1.rb

@@ -1,18 +1,11 @@
-module Regexp::Syntax
-  class V1_9_1 < Regexp::Syntax::V1_8_6
-    def initialize
-      super
-
-      implements :assertion, Assertion::Lookbehind
-      implements :backref, Backreference::V1_9_1 + SubexpressionCall::All
-      implements :posixclass, PosixClass::Extensions
-      implements :nonposixclass, PosixClass::All
-      implements :escape, Escape::Unicode + Escape::Hex + Escape::Octal
-      implements :type, CharacterType::Hex
-      implements :property, UnicodeProperty::V1_9_0
-      implements :nonproperty, UnicodeProperty::V1_9_0
-      implements :quantifier,
-        Quantifier::Possessive + Quantifier::IntervalPossessive
-    end
-  end
+class Regexp::Syntax::V1_9_1 < Regexp::Syntax::V1_8_6
+  implements :assertion,     Assertion::Lookbehind
+  implements :backref,       Backreference::V1_9_1 + SubexpressionCall::All
+  implements :escape,        Escape::Unicode + Escape::Hex + Escape::Octal
+  implements :posixclass,    PosixClass::Extensions
+  implements :nonposixclass, PosixClass::All
+  implements :property,      UnicodeProperty::V1_9_0
+  implements :nonproperty,   UnicodeProperty::V1_9_0
+  implements :quantifier,    Quantifier::Possessive + Quantifier::IntervalPossessive
+  implements :type,          CharacterType::Hex
 end

data/lib/regexp_parser/syntax/versions/1.9.3.rb

@@ -1,11 +1,4 @@
-module Regexp::Syntax
-  class V1_9_3 < Regexp::Syntax::V1_9_1
-    def initialize
-      super
-
-      # these were added with update of Oniguruma to Unicode 6.0
-      implements :property,    UnicodeProperty::V1_9_3
-      implements :nonproperty, UnicodeProperty::V1_9_3
-    end
-  end
+class Regexp::Syntax::V1_9_3 < Regexp::Syntax::V1_9_1
+  implements :property,    UnicodeProperty::V1_9_3
+  implements :nonproperty, UnicodeProperty::V1_9_3
 end

data/lib/regexp_parser/syntax/versions/2.0.0.rb

@@ -1,17 +1,10 @@
-module Regexp::Syntax
-  # use the last 1.9 release as the base
-  class V2_0_0 < Regexp::Syntax::V1_9
-    def initialize
-      super
+class Regexp::Syntax::V2_0_0 < Regexp::Syntax::V1_9_3
+  implements :keep,        Keep::All
+  implements :conditional, Conditional::All
+  implements :property,    UnicodeProperty::V2_0_0
+  implements :nonproperty, UnicodeProperty::V2_0_0
+  implements :type,        CharacterType::Clustered
 
-      implements :keep,        Keep::All
-      implements :conditional, Conditional::All
-      implements :property,    UnicodeProperty::V2_0_0
-      implements :nonproperty, UnicodeProperty::V2_0_0
-      implements :type,        CharacterType::Clustered
-
-      excludes   :property,    :newline
-      excludes   :nonproperty, :newline
-    end
-  end
+  excludes   :property,    %i[newline]
+  excludes   :nonproperty, %i[newline]
 end

data/lib/regexp_parser/syntax/versions/2.2.0.rb

@@ -1,10 +1,4 @@
-module Regexp::Syntax
-  class V2_2_0 < Regexp::Syntax::V2_1
-    def initialize
-      super
-
-      implements :property,    UnicodeProperty::V2_2_0
-      implements :nonproperty, UnicodeProperty::V2_2_0
-    end
-  end
+class Regexp::Syntax::V2_2_0 < Regexp::Syntax::V2_0_0
+  implements :property,    UnicodeProperty::V2_2_0
+  implements :nonproperty, UnicodeProperty::V2_2_0
 end

data/lib/regexp_parser/syntax/versions/2.3.0.rb

@@ -1,10 +1,4 @@
-module Regexp::Syntax
-  class V2_3_0 < Regexp::Syntax::V2_2
-    def initialize
-      super
-
-      implements :property,    UnicodeProperty::V2_3_0
-      implements :nonproperty, UnicodeProperty::V2_3_0
-    end
-  end
+class Regexp::Syntax::V2_3_0 < Regexp::Syntax::V2_2_0
+  implements :property,    UnicodeProperty::V2_3_0
+  implements :nonproperty, UnicodeProperty::V2_3_0
 end

data/lib/regexp_parser/syntax/versions/2.4.0.rb

@@ -1,10 +1,4 @@
-module Regexp::Syntax
-  class V2_4_0 < Regexp::Syntax::V2_3
-    def initialize
-      super
-
-      implements :property,    UnicodeProperty::V2_4_0
-      implements :nonproperty, UnicodeProperty::V2_4_0
-    end
-  end
+class Regexp::Syntax::V2_4_0 < Regexp::Syntax::V2_3_0
+  implements :property,    UnicodeProperty::V2_4_0
+  implements :nonproperty, UnicodeProperty::V2_4_0
 end

data/lib/regexp_parser/syntax/versions/2.4.1.rb

@@ -1,9 +1,3 @@
-module Regexp::Syntax
-  class V2_4_1 < Regexp::Syntax::V2_4_0
-    def initialize
-      super
-
-      implements :group, Group::V2_4_1
-    end
-  end
+class Regexp::Syntax::V2_4_1 < Regexp::Syntax::V2_4_0
+  implements :group, Group::V2_4_1
 end

data/lib/regexp_parser/syntax/versions/2.5.0.rb

@@ -1,10 +1,4 @@
-module Regexp::Syntax
-  class V2_5_0 < Regexp::Syntax::V2_4
-    def initialize
-      super
-
-      implements :property,    UnicodeProperty::V2_5_0
-      implements :nonproperty, UnicodeProperty::V2_5_0
-    end
-  end
+class Regexp::Syntax::V2_5_0 < Regexp::Syntax::V2_4_1
+  implements :property,    UnicodeProperty::V2_5_0
+  implements :nonproperty, UnicodeProperty::V2_5_0
 end

data/lib/regexp_parser/syntax/versions/2.6.0.rb

@@ -1,10 +1,4 @@
-module Regexp::Syntax
-  class V2_6_0 < Regexp::Syntax::V2_5
-    def initialize
-      super
-
-      implements :property,    UnicodeProperty::V2_6_0
-      implements :nonproperty, UnicodeProperty::V2_6_0
-    end
-  end
+class Regexp::Syntax::V2_6_0 < Regexp::Syntax::V2_5_0
+  implements :property,    UnicodeProperty::V2_6_0
+  implements :nonproperty, UnicodeProperty::V2_6_0
 end

data/lib/regexp_parser/syntax/versions/2.6.2.rb

@@ -1,10 +1,4 @@
-module Regexp::Syntax
-  class V2_6_2 < Regexp::Syntax::V2_6_0
-    def initialize
-      super
-
-      implements :property,    UnicodeProperty::V2_6_2
-      implements :nonproperty, UnicodeProperty::V2_6_2
-    end
-  end
+class Regexp::Syntax::V2_6_2 < Regexp::Syntax::V2_6_0
+  implements :property,    UnicodeProperty::V2_6_2
+  implements :nonproperty, UnicodeProperty::V2_6_2
 end

data/lib/regexp_parser/syntax/versions/2.6.3.rb

@@ -1,10 +1,4 @@
-module Regexp::Syntax
-  class V2_6_3 < Regexp::Syntax::V2_6_2
-    def initialize
-      super
-
-      implements :property,    UnicodeProperty::V2_6_3
-      implements :nonproperty, UnicodeProperty::V2_6_3
-    end
-  end
+class Regexp::Syntax::V2_6_3 < Regexp::Syntax::V2_6_2
+  implements :property,    UnicodeProperty::V2_6_3
+  implements :nonproperty, UnicodeProperty::V2_6_3
 end

data/lib/regexp_parser/syntax/versions/3.1.0.rb

@@ -1,10 +1,4 @@
-module Regexp::Syntax
-  class V3_1_0 < Regexp::Syntax::V2_6_3
-    def initialize
-      super
-
-      implements :property,    UnicodeProperty::V3_1_0
-      implements :nonproperty, UnicodeProperty::V3_1_0
-    end
-  end
+class Regexp::Syntax::V3_1_0 < Regexp::Syntax::V2_6_3
+  implements :property,    UnicodeProperty::V3_1_0
+  implements :nonproperty, UnicodeProperty::V3_1_0
 end

data/lib/regexp_parser/syntax/versions/3.2.0.rb

@@ -0,0 +1,4 @@
+class Regexp::Syntax::V3_2_0 < Regexp::Syntax::V3_1_0
+  implements :property,    UnicodeProperty::V3_2_0
+  implements :nonproperty, UnicodeProperty::V3_2_0
+end

data/lib/regexp_parser/syntax/versions.rb

@@ -1,4 +1,4 @@
-# Ruby 1.8.x is no longer a supported runtime,
+# Ruby 1.x is no longer a supported runtime,
 # but its regex features are still recognized.
 #
 # Aliases for the latest patch version are provided as 'ruby/n.n',

data/lib/regexp_parser/version.rb

@@ -1,5 +1,5 @@
 class Regexp
   class Parser
-    VERSION = '2.2.1'
+    VERSION = '2.3.0'
   end
 end