regexp_parser 1.7.0 → 2.9.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: d7b93dde993f6fe427ff43755738bf7de50f8613cf6e8097c9d791646d803e4c
-  data.tar.gz: 993a88720a4ee1d8a34f4c95e167089adc6455289bfeb356de8c028a9bbee63d
+  metadata.gz: 8cc1826647cde51d6d1b5a5a58fb005efd2a38a85fa0e817616591ee2fad7862
+  data.tar.gz: 572a6203741b9970bcedc1ace243ea0b9c300ca60b71ac263036eb0f4222dd50
 SHA512:
-  metadata.gz: 0bf5c142591b2d5a65023c53f76a64a13106074050042d24614963cc14dabda197ea9140fccd93f26ad06885293369b076bb5e9198967a6e3762654df8033455
-  data.tar.gz: 1311b3dfa90633ef456edc12abf6ace2d7311c7be8450f3768a436f9c8491c3a87987f3d6ac24c6966b6f4de5363e0f6f874bfe9e1b038a6cf5d9c043553b58e
+  metadata.gz: 3521fe6dab4be0c0db3c37f3f8d196fc754ff72937336a73ef5547a15ae4f2d366aa28e73d6e5756920d610b943ee51cb2db8e51e53ccb19c1c235a8c45da708
+  data.tar.gz: d05b7babb79c118bdc36ae168d8199ee3500b0cff33cb00ed46d51a4a88725130e931c588146a3f989dd87778b1f39684b2c8a5541c9ac8f91427fc31b1ec97a

data/Gemfile

@@ -3,7 +3,13 @@ source 'https://rubygems.org'
 gemspec
 
 group :development, :test do
-  gem 'rake', '~> 13.0'
-  gem 'regexp_property_values', '~> 1.0'
-  gem 'rspec', '~> 3.8'
+  gem 'leto', '~> 2.1'
+  gem 'rake', '~> 13.1'
+  gem 'regexp_property_values', '~> 1.5'
+  gem 'rspec', '~> 3.10'
+  if RUBY_VERSION.to_f >= 2.7
+    gem 'benchmark-ips', '~> 2.1'
+    gem 'gouteur', '~> 1.1'
+    gem 'rubocop', '~> 1.59'
+  end
 end

data/LICENSE

@@ -1,4 +1,4 @@
-Copyright (c) 2010, 2012-2015,  Ammar Ali
+Copyright (c) 2010, 2012-2024,  Ammar Ali
 
 Permission is hereby granted, free of charge, to any person
 obtaining a copy of this software and associated documentation

data/Rakefile

@@ -1,87 +1,23 @@
+require 'bundler'
 require 'rubygems'
-
+require 'rubygems/package_task'
 require 'rake'
 require 'rake/testtask'
+require 'rspec/core/rake_task'
 
-require 'bundler'
-require 'rubygems/package_task'
-
-
-RAGEL_SOURCE_DIR = File.expand_path '../lib/regexp_parser/scanner', __FILE__
-RAGEL_OUTPUT_DIR = File.expand_path '../lib/regexp_parser', __FILE__
-RAGEL_SOURCE_FILES = %w{scanner} # scanner.rl includes property.rl
-
+Dir['tasks/**/*.rake'].each { |file| load(file) }
 
 Bundler::GemHelper.install_tasks
 
+RSpec::Core::RakeTask.new(:spec)
 
 task :default => [:'test:full']
 
 namespace :test do
-  task full: :'ragel:rb' do
-    sh 'bin/test'
-  end
+  task full: [:'ragel:rb', :spec]
 end
 
-namespace :ragel do
-  desc "Process the ragel source files and output ruby code"
-  task :rb do |t|
-    RAGEL_SOURCE_FILES.each do |file|
-      output_file = "#{RAGEL_OUTPUT_DIR}/#{file}.rb"
-      # using faster flat table driven FSM, about 25% larger code, but about 30% faster
-      sh "ragel -F1 -R #{RAGEL_SOURCE_DIR}/#{file}.rl -o #{output_file}"
-
-      contents = File.read(output_file)
-
-      File.open(output_file, 'r+') do |file|
-        contents = "# -*- warn-indent:false;  -*-\n" + contents
-
-        file.write(contents)
-      end
-    end
-  end
-
-  desc "Delete the ragel generated source file(s)"
-  task :clean do |t|
-    RAGEL_SOURCE_FILES.each do |file|
-      sh "rm -f #{RAGEL_OUTPUT_DIR}/#{file}.rb"
-    end
-  end
-end
-
-
 # Add ragel task as a prerequisite for building the gem to ensure that the
 # latest scanner code is generated and included in the build.
 desc "Runs ragel:rb before building the gem"
 task :build => ['ragel:rb']
-
-
-namespace :props do
-  desc 'Write new property value hashes for the properties scanner'
-  task :update do
-    require 'regexp_property_values'
-    RegexpPropertyValues.update
-    dir = File.expand_path('../lib/regexp_parser/scanner/properties', __FILE__)
-
-    require 'psych'
-    write_hash_to_file = ->(hash, path) do
-      File.open(path, 'w') do |f|
-        f.puts '#',
-               "# THIS FILE IS AUTO-GENERATED BY `rake props:update`, DO NOT EDIT",
-               '#',
-               hash.sort.to_h.to_yaml
-      end
-      puts "Wrote #{hash.count} aliases to `#{path}`"
-    end
-
-    long_names_to_tokens = RegexpPropertyValues.all.map do |val|
-      [val.identifier, val.full_name.downcase]
-    end
-    write_hash_to_file.call(long_names_to_tokens, "#{dir}/long.yml")
-
-    short_names_to_tokens = RegexpPropertyValues.alias_hash.map do |k, v|
-      [k.identifier, v.full_name.downcase]
-    end
-    write_hash_to_file.call(short_names_to_tokens, "#{dir}/short.yml")
-  end
-end

data/lib/regexp_parser/error.rb

@@ -0,0 +1,4 @@
+class Regexp::Parser
+  # base class for all gem-specific errors
+  class Error < StandardError; end
+end

data/lib/regexp_parser/expression/base.rb

@@ -0,0 +1,76 @@
+module Regexp::Expression
+  class Base
+    include Regexp::Expression::Shared
+
+    def initialize(token, options = {})
+      init_from_token_and_options(token, options)
+    end
+
+    def to_re(format = :full)
+      if set_level > 0
+        warn "Calling #to_re on character set members is deprecated - "\
+             "their behavior might not be equivalent outside of the set."
+      end
+      ::Regexp.new(to_s(format))
+    end
+
+    def quantify(*args)
+      self.quantifier = Quantifier.new(*args)
+    end
+
+    def unquantified_clone
+      clone.tap { |exp| exp.quantifier = nil }
+    end
+
+    # Deprecated. Prefer `#repetitions` which has a more uniform interface.
+    def quantity
+      return [nil,nil] unless quantified?
+      [quantifier.min, quantifier.max]
+    end
+
+    def repetitions
+      @repetitions ||=
+        if quantified?
+          min = quantifier.min
+          max = quantifier.max < 0 ? Float::INFINITY : quantifier.max
+          range = min..max
+          # fix Range#minmax on old Rubies - https://bugs.ruby-lang.org/issues/15807
+          if RUBY_VERSION.to_f < 2.7
+            range.define_singleton_method(:minmax) { [min, max] }
+          end
+          range
+        else
+          1..1
+        end
+    end
+
+    def greedy?
+      quantified? and quantifier.greedy?
+    end
+
+    def reluctant?
+      quantified? and quantifier.reluctant?
+    end
+    alias :lazy? :reluctant?
+
+    def possessive?
+      quantified? and quantifier.possessive?
+    end
+
+    def to_h
+      {
+        type:              type,
+        token:             token,
+        text:              to_s(:base),
+        starts_at:         ts,
+        length:            full_length,
+        level:             level,
+        set_level:         set_level,
+        conditional_level: conditional_level,
+        options:           options,
+        quantifier:        quantified? ? quantifier.to_h : nil,
+      }
+    end
+    alias :attributes :to_h
+  end
+end

data/lib/regexp_parser/expression/classes/alternation.rb

@@ -1,5 +1,5 @@
 module Regexp::Expression
-  # A sequence of expressions, used by Alternation as one of its alternative.
+  # A sequence of expressions, used by Alternation as one of its alternatives.
   class Alternative < Regexp::Expression::Sequence; end
 
   class Alternation < Regexp::Expression::SequenceOperation

data/lib/regexp_parser/expression/classes/anchor.rb

@@ -1,5 +1,4 @@
 module Regexp::Expression
-
   module Anchor
     class Base < Regexp::Expression::Base; end
 
@@ -22,5 +21,4 @@ module Regexp::Expression
     EOS      = EndOfString
     EOSobEOL = EndOfStringOrBeforeEndOfLine
   end
-
 end

data/lib/regexp_parser/expression/classes/{backref.rb → backreference.rb}

@@ -2,6 +2,23 @@ module Regexp::Expression
   module Backreference
     class Base < Regexp::Expression::Base
       attr_accessor :referenced_expression
+
+      def initialize_copy(orig)
+        exp_id = [self.class, self.starts_at]
+
+        # prevent infinite recursion for recursive subexp calls
+        copied = @@copied ||= {}
+        self.referenced_expression =
+          if copied[exp_id]
+            orig.referenced_expression
+          else
+            copied[exp_id] = true
+            orig.referenced_expression.dup
+          end
+        copied.clear
+
+        super
+      end
     end
 
     class Number < Backreference::Base
@@ -9,7 +26,7 @@ module Regexp::Expression
       alias reference number
 
       def initialize(token, options = {})
-        @number = token.text[token.token.equal?(:number) ? 1..-1 : 3..-2].to_i
+        @number = token.text[/-?\d+/].to_i
         super
       end
     end
@@ -33,7 +50,7 @@ module Regexp::Expression
     class NameCall           < Backreference::Name; end
     class NumberCallRelative < Backreference::NumberRelative; end
 
-    class NumberRecursionLevel < Backreference::Number
+    class NumberRecursionLevel < Backreference::NumberRelative
       attr_reader :recursion_level
 
       def initialize(token, options = {})
@@ -52,4 +69,7 @@ module Regexp::Expression
       end
     end
   end
+
+  # alias for symmetry between token symbol and Expression class name
+  Backref = Backreference
 end

data/lib/regexp_parser/expression/classes/{set → character_set}/range.rb

@@ -1,23 +1,19 @@
 module Regexp::Expression
   class CharacterSet < Regexp::Expression::Subexpression
     class Range < Regexp::Expression::Subexpression
-      def starts_at
-        expressions.first.starts_at
+      def ts
+        (head = expressions.first) ? head.ts : @ts
       end
-      alias :ts :starts_at
 
       def <<(exp)
-        complete? && raise("Can't add more than 2 expressions to a Range")
+        complete? and raise Regexp::Parser::Error,
+          "Can't add more than 2 expressions to a Range"
         super
       end
 
       def complete?
         count == 2
       end
-
-      def to_s(_format = :full)
-        expressions.join(text)
-      end
     end
   end
 end

data/lib/regexp_parser/expression/classes/{set.rb → character_set.rb}

@@ -1,10 +1,7 @@
 module Regexp::Expression
   class CharacterSet < Regexp::Expression::Subexpression
     attr_accessor :closed, :negative
-
-    alias :negative? :negative
-    alias :negated?  :negative
-    alias :closed?   :closed
+    alias :closed? :closed
 
     def initialize(token, options = {})
       self.negative = false
@@ -19,9 +16,8 @@ module Regexp::Expression
     def close
       self.closed = true
     end
-
-    def to_s(format = :full)
-      "#{text}#{'^' if negated?}#{expressions.join}]#{quantifier_affix(format)}"
-    end
   end
+
+  # alias for symmetry between token symbol and Expression class name
+  Set = CharacterSet
 end # module Regexp::Expression

data/lib/regexp_parser/expression/classes/{type.rb → character_type.rb}

@@ -1,5 +1,4 @@
 module Regexp::Expression
-
   module CharacterType
     class Base < Regexp::Expression::Base; end
 
@@ -15,5 +14,4 @@ module Regexp::Expression
     class Linebreak        < CharacterType::Base; end
     class ExtendedGrapheme < CharacterType::Base; end
   end
-
 end

data/lib/regexp_parser/expression/classes/conditional.rb

@@ -1,6 +1,6 @@
 module Regexp::Expression
   module Conditional
-    class TooManyBranches < StandardError
+    class TooManyBranches < Regexp::Parser::Error
       def initialize
         super('The conditional expression has more than 2 branches')
       end
@@ -15,6 +15,11 @@ module Regexp::Expression
         ref = text.tr("'<>()", "")
         ref =~ /\D/ ? ref : Integer(ref)
       end
+
+      def initialize_copy(orig)
+        self.referenced_expression = orig.referenced_expression.dup
+        super
+      end
     end
 
     class Branch < Regexp::Expression::Sequence; end
@@ -26,9 +31,9 @@ module Regexp::Expression
         expressions.last << exp
       end
 
-      def add_sequence(active_opts = {})
+      def add_sequence(active_opts = {}, params = { ts: 0 })
         raise TooManyBranches.new if branches.length == 2
-        params = { conditional_level: conditional_level + 1 }
+        params = params.merge({ conditional_level: conditional_level + 1 })
         Branch.add_to(self, params, active_opts)
       end
       alias :branch :add_sequence
@@ -50,8 +55,9 @@ module Regexp::Expression
         condition.reference
       end
 
-      def to_s(format = :full)
-        "#{text}#{condition}#{branches.join('|')})#{quantifier_affix(format)}"
+      def initialize_copy(orig)
+        self.referenced_expression = orig.referenced_expression.dup
+        super
       end
     end
   end

data/lib/regexp_parser/expression/classes/{escape.rb → escape_sequence.rb}

@@ -1,16 +1,21 @@
 module Regexp::Expression
   module EscapeSequence
     class Base < Regexp::Expression::Base
-      require 'yaml'
-
-      def char
-        # poor man's unescape without using eval
-        YAML.load(%Q(---\n"#{text}"\n))
-      end
-
       def codepoint
         char.ord
       end
+
+      if ''.respond_to?(:undump)
+        def char
+          %("#{text}").undump
+        end
+      else
+        # poor man's unescape without using eval
+        require 'yaml'
+        def char
+          YAML.load(%Q(---\n"#{text}"\n))
+        end
+      end
     end
 
     class Literal < EscapeSequence::Base
@@ -91,4 +96,7 @@ module Regexp::Expression
       end
     end
   end
+
+  # alias for symmetry between Token::* and Expression::*
+  Escape = EscapeSequence
 end

data/lib/regexp_parser/expression/classes/free_space.rb

@@ -1,17 +1,17 @@
 module Regexp::Expression
-
   class FreeSpace < Regexp::Expression::Base
-    def quantify(token, text, min = nil, max = nil, mode = :greedy)
-      raise "Can not quantify a free space object"
+    def quantify(*_args)
+      raise Regexp::Parser::Error, 'Can not quantify a free space object'
     end
   end
 
-  class Comment < Regexp::Expression::FreeSpace; end
+  class Comment < Regexp::Expression::FreeSpace
+  end
 
   class WhiteSpace < Regexp::Expression::FreeSpace
     def merge(exp)
+      warn("#{self.class}##{__method__} is deprecated and will be removed in v3.0.0.")
       text << exp.text
     end
   end
-
 end

data/lib/regexp_parser/expression/classes/group.rb

@@ -1,27 +1,45 @@
 module Regexp::Expression
   module Group
     class Base < Regexp::Expression::Subexpression
-      def to_s(format = :full)
-        "#{text}#{expressions.join})#{quantifier_affix(format)}"
-      end
+    end
 
-      def capturing?; false end
+    class Passive < Group::Base
+      attr_writer :implicit
+
+      def initialize(*)
+        @implicit = false
+        super
+      end
 
-      def comment?; false end
+      def implicit?
+        @implicit
+      end
     end
 
-    class Atomic  < Group::Base; end
-    class Passive < Group::Base; end
     class Absence < Group::Base; end
+    class Atomic  < Group::Base; end
+    # TODO: should split off OptionsSwitch in v3.0.0. Maybe even make it no
+    # longer inherit from Group because it is effectively a terminal expression.
     class Options < Group::Base
       attr_accessor :option_changes
+
+      def initialize_copy(orig)
+        self.option_changes = orig.option_changes.dup
+        super
+      end
+
+      def quantify(*args)
+        if token == :options_switch
+          raise Regexp::Parser::Error, 'Can not quantify an option switch'
+        else
+          super
+        end
+      end
     end
 
     class Capture < Group::Base
       attr_accessor :number, :number_at_level
       alias identifier number
-
-      def capturing?; true end
     end
 
     class Named < Group::Capture
@@ -33,18 +51,13 @@ module Regexp::Expression
         super
       end
 
-      def initialize_clone(orig)
+      def initialize_copy(orig)
         @name = orig.name.dup
         super
       end
     end
 
     class Comment < Group::Base
-      def to_s(_format = :full)
-        text.dup
-      end
-
-      def comment?; true end
     end
   end
 

data/lib/regexp_parser/expression/classes/keep.rb

@@ -1,5 +1,7 @@
 module Regexp::Expression
   module Keep
+    # TOOD: in regexp_parser v3.0.0 this should possibly be a Subexpression
+    #       that contains all expressions to its left.
     class Mark < Regexp::Expression::Base; end
   end
 end

data/lib/regexp_parser/expression/classes/literal.rb

@@ -1,7 +1,3 @@
 module Regexp::Expression
-
-  class Literal < Regexp::Expression::Base
-    # Obviously nothing special here, yet.
-  end
-
+  class Literal < Regexp::Expression::Base; end
 end

data/lib/regexp_parser/expression/classes/posix_class.rb

@@ -1,11 +1,11 @@
 module Regexp::Expression
   class PosixClass < Regexp::Expression::Base
-    def negative?
-      type == :nonposixclass
-    end
-
     def name
-      token.to_s
+      text[/\w+/]
     end
   end
+
+  # alias for symmetry between token symbol and Expression class name
+  Posixclass    = PosixClass
+  Nonposixclass = PosixClass
 end

data/lib/regexp_parser/expression/classes/root.rb

@@ -1,24 +1,9 @@
 module Regexp::Expression
-
   class Root < Regexp::Expression::Subexpression
-    # TODO: this override is here for backwards compatibility, remove in 2.0.0
-    def initialize(*args)
-      unless args.first.is_a?(Regexp::Token)
-        warn('WARNING: Root.new without a Token argument is deprecated and '\
-             'will be removed in 2.0.0. Use Root.build for the old behavior.')
-        return super(self.class.build_token, *args)
-      end
-      super
-    end
-
-    class << self
-      def build(options = {})
-        new(build_token, options)
-      end
-
-      def build_token
-        Regexp::Token.new(:expression, :root, '', 0)
-      end
+    def self.build(options = {})
+      warn "`#{self.class}.build(options)` is deprecated and will raise in "\
+           "regexp_parser v3.0.0. Please use `.construct(options: options)`."
+      construct(options: options)
     end
   end
 end

data/lib/regexp_parser/expression/classes/{property.rb → unicode_property.rb}

@@ -1,17 +1,12 @@
 module Regexp::Expression
-
   module UnicodeProperty
     class Base < Regexp::Expression::Base
-      def negative?
-        type == :nonproperty
-      end
-
       def name
-        text =~ /\A\\[pP]\{([^}]+)\}\z/; $1
+        text[/\A\\[pP]\{([^}]+)\}\z/, 1]
       end
 
       def shortcut
-        (Regexp::Scanner.short_prop_map.rassoc(token.to_s) || []).first
+        Regexp::Scanner.short_prop_map.key(token.to_s)
       end
     end
 
@@ -110,11 +105,15 @@ module Regexp::Expression
       class Unassigned  < Codepoint::Base; end
     end
 
-    class Age     < UnicodeProperty::Base; end
-    class Derived < UnicodeProperty::Base; end
-    class Emoji   < UnicodeProperty::Base; end
-    class Script  < UnicodeProperty::Base; end
-    class Block   < UnicodeProperty::Base; end
+    class Age        < UnicodeProperty::Base; end
+    class Block      < UnicodeProperty::Base; end
+    class Derived    < UnicodeProperty::Base; end
+    class Emoji      < UnicodeProperty::Base; end
+    class Enumerated < UnicodeProperty::Base; end
+    class Script     < UnicodeProperty::Base; end
   end
 
+  # alias for symmetry between token symbol and Expression class name
+  Property    = UnicodeProperty
+  Nonproperty = UnicodeProperty
 end # module Regexp::Expression

data/lib/regexp_parser/expression/methods/construct.rb

@@ -0,0 +1,41 @@
+module Regexp::Expression
+  module Shared
+    module ClassMethods
+      # Convenience method to init a valid Expression without a Regexp::Token
+      def construct(params = {})
+        attrs = construct_defaults.merge(params)
+        options = attrs.delete(:options)
+        token_args = Regexp::TOKEN_KEYS.map { |k| attrs.delete(k) }
+        token = Regexp::Token.new(*token_args)
+        raise ArgumentError, "unsupported attribute(s): #{attrs}" if attrs.any?
+
+        new(token, options)
+      end
+
+      def construct_defaults
+        if self == Root
+          { type: :expression, token: :root, ts: 0 }
+        elsif self < Sequence
+          { type: :expression, token: :sequence }
+        else
+          { type: token_class::Type }
+        end.merge(level: 0, set_level: 0, conditional_level: 0, text: '')
+      end
+
+      def token_class
+        if self == Root || self < Sequence
+          nil # no token class because these objects are Parser-generated
+        # TODO: synch exp class, token class & type names for this in v3.0.0
+        elsif self == CharacterType::Any
+          Regexp::Syntax::Token::Meta
+        else
+          Regexp::Syntax::Token.const_get(name.split('::')[2])
+        end
+      end
+    end
+
+    def token_class
+      self.class.token_class
+    end
+  end
+end