uri-idna 0.2.1 → 0.2.2

data/lib/uri/idna/data/virama_combining_classes.rb

@@ -0,0 +1,67 @@
+# frozen_string_literal: true
+
+# This file is automatically generated by bin/generate
+# Unicode version 15.1.0
+
+module URI
+  module IDNA
+    VIRAMA_COMBINING_CLASSES = "["\
+      "\u094D"\
+      "\u09CD"\
+      "\u0A4D"\
+      "\u0ACD"\
+      "\u0B4D"\
+      "\u0BCD"\
+      "\u0C4D"\
+      "\u0CCD"\
+      "\u0D3B\u0D3C"\
+      "\u0D4D"\
+      "\u0DCA"\
+      "\u0E3A"\
+      "\u0EBA"\
+      "\u0F84"\
+      "\u1039\u103A"\
+      "\u1714\u1715"\
+      "\u1734"\
+      "\u17D2"\
+      "\u1A60"\
+      "\u1B44"\
+      "\u1BAA\u1BAB"\
+      "\u1BF2\u1BF3"\
+      "\u2D7F"\
+      "\uA806"\
+      "\uA82C"\
+      "\uA8C4"\
+      "\uA953"\
+      "\uA9C0"\
+      "\uAAF6"\
+      "\uABED"\
+      "\u{10A3F}"\
+      "\u{11046}"\
+      "\u{11070}"\
+      "\u{1107F}"\
+      "\u{110B9}"\
+      "\u{11133}\u{11134}"\
+      "\u{111C0}"\
+      "\u{11235}"\
+      "\u{112EA}"\
+      "\u{1134D}"\
+      "\u{11442}"\
+      "\u{114C2}"\
+      "\u{115BF}"\
+      "\u{1163F}"\
+      "\u{116B6}"\
+      "\u{1172B}"\
+      "\u{11839}"\
+      "\u{1193D}\u{1193E}"\
+      "\u{119E0}"\
+      "\u{11A34}"\
+      "\u{11A47}"\
+      "\u{11A99}"\
+      "\u{11C3F}"\
+      "\u{11D44}\u{11D45}"\
+      "\u{11D97}"\
+      "\u{11F41}\u{11F42}"\
+      "]"
+  end
+end

data/lib/uri/idna/idna2008/processing.rb

@@ -1,17 +1,21 @@
 # frozen_string_literal: true
 
 require_relative "options"
+require_relative "../validation/contextj"
+require_relative "../validation/contexto"
+require_relative "../validation/idna_permitted"
+require_relative "../validation/leading_combining"
 
 module URI
   module IDNA
     module IDNA2008
       class Processing < BaseProcessing
-        private
-
-        def options_class
+        def self.options_class
           Options
         end
 
+        private
+
         def validate(label)
           return if label.empty?
 
@@ -21,35 +25,16 @@ module URI
           else
             Validation::Label.check_ace_prefix(label)
           end
-          Validation::Label.check_leading_combining(label) if options.leading_combining?
-
-          label.each_codepoint.with_index do |cp, pos|
-            begin
-              next if Validation::Codepoint.check_contextj(label, cp, pos)
-            rescue InvalidCodepointContextError => e
-              next unless options.check_joiners?
-
-              raise e
-            end
-
-            begin
-              next if Validation::Codepoint.check_contexto(label, cp, pos)
-            rescue InvalidCodepointContextError => e
-              next unless options.check_others?
-
-              raise e
-            end
-
-            Validation::Codepoint.check_idna_validity(label, cp, pos)
-          end
+          Validation::LeadingCombining.call(label) if options.leading_combining?
+          Validation::ContextJ.call(label) if options.check_joiners?
+          Validation::ContextO.call(label) if options.check_others?
+          Validation::IDNAPermitted.call(label)
           Validation::Bidi.call(label) if check_bidi?
         end
 
         def punycode_decode(label)
           return label unless label.start_with?(ACE_PREFIX)
 
-          raise Error, "A-label must not end with a hyphen" if label[-1] == "-"
-
           super
         end
       end
@@ -95,9 +80,9 @@ module URI
 
             Validation::Label.check_length(a_ulabel) if options.verify_dns_length?
 
-            if alabel && ulabel && (a_ulabel != alabel) && (a_ulabel != alabel)
+            if alabel && ulabel && a_ulabel != alabel
               raise Error,
-                    "Provided alabel #{alabel.inspect} doesn't match de-punycoded ulabel #{u_alabel.inspect}"
+                    "Provided alabel #{alabel.inspect} doesn't match de-punycoded ulabel #{a_ulabel.inspect}"
             end
 
             a_ulabel

data/lib/uri/idna/punycode.rb

@@ -13,6 +13,7 @@ module URI
         DAMP = 700
         INITIAL_BIAS = 72
         INITIAL_N = 0x80
+        ADAPT_THRESHOLD = ((BASE - TMIN) * TMAX) / 2
 
         DELIMITER = 0x2D
         MAXINT = 0x7FFFFFFF
@@ -30,7 +31,9 @@ module URI
         end
 
         def encode_digit(d)
-          d + 22 + 75 * (d < 26 ? 1 : 0)
+          return d + 22 if d >= 26
+
+          d + 97
         end
 
         def adapt(delta, num_points, first_time)
@@ -38,7 +41,7 @@ module URI
           delta += (delta / num_points)
 
           k = 0
-          while delta > (((BASE - TMIN) * TMAX) / 2)
+          while delta > ADAPT_THRESHOLD
             delta /= BASE - TMIN
             k += BASE
           end
@@ -47,18 +50,17 @@ module URI
 
         def encode(input)
           input = input.codepoints
-          output = []
 
           n = INITIAL_N
           delta = 0
           bias = INITIAL_BIAS
 
-          input.each { |cp| output << cp if cp < 0x80 }
+          output = input.select { |cp| cp < 0x80 }
           h = b = output.length
 
           output << DELIMITER if b > 0
-
-          while h < input.length
+          input_length = input.length
+          while h < input_length
             m = MAXINT
             input.each do |cp|
               m = cp if cp >= n && cp < m
@@ -116,15 +118,15 @@ module URI
 
           b = input.rindex(DELIMITER) || 0
 
-          0.upto(b - 1) do |idx|
-            cp = input[idx]
+          input[0, b].each do |cp|
             raise PunycodeError, "Invalid input" unless cp < 0x80
 
             output << cp
           end
 
           inc = b > 0 ? b + 1 : 0
-          while inc < input.length
+          input_length = input.length
+          while inc < input_length
             old_i = i
             w = 1
             k = BASE

data/lib/uri/idna/uts46/mapping.rb

@@ -8,52 +8,54 @@ module URI
       # https://www.unicode.org/reports/tr46/#IDNA_Mapping_Table
       module Mapping
         class << self
+          STATUS_D_REGEX = Regexp.new(REGEX_D_STRING, Regexp::EXTENDED).freeze
+          REGEX_STD3_M_REGEX = Regexp.new(REGEX_STD3_M_STRING, Regexp::EXTENDED).freeze
+
+          MAP_REGEX = Regexp.new("#{REGEX_M_STRING}|#{REGEX_I_STRING}").freeze
+          REGEX_NOT_V = Regexp.new("[^#{REGEX_V_STRING}]").freeze
+          REGEX_NOT_VD = Regexp.new("[^#{REGEX_V_STRING}|#{REGEX_D_STRING}]").freeze
+          REGEX_NOT_V3 = Regexp.new("[^#{REGEX_V_STRING}|#{REGEX_STD3_M_STRING}|#{REGEX_STD3_V_STRING}]").freeze
+          REGEX_NOT_VD3 = Regexp.new(
+            "[^#{REGEX_V_STRING}|#{REGEX_D_STRING}|#{REGEX_STD3_M_STRING}|#{REGEX_STD3_V_STRING}]",
+          ).freeze
+
           def call(domain_name, transitional_processing: false, use_std3_ascii_rules: true)
-            output = []
-            domain_name.each_codepoint do |codepoint|
-              _, status, replacement = status(codepoint)
-              case status
-              when "V", "X" # valid, disallowed
-                output << codepoint
-              when "M" # mapped
-                output += if transitional_processing && codepoint == 7838
-                            [115, 115]
-                          else
-                            replacement
-                          end
-              when "D" # deviation
-                if transitional_processing
-                  output += replacement
-                else
-                  output << codepoint
-                end
-              when "3" # disallowed_STD3_valid, disallowed_STD3_mapped
-                if use_std3_ascii_rules || !replacement
-                  output << codepoint
-                else
-                  output += replacement
-                end
-              when "I" # ignored
-                next
+            return domain_name.downcase if domain_name.ascii_only?
+
+            output = domain_name.gsub(MAP_REGEX) do |match|
+              if transitional_processing && match == "\u1E9E"
+                "ss"
+              else
+                REPLACEMENTS[match]
               end
             end
-            output.pack("U*").unicode_normalize!(:nfc)
+            output.gsub!(STATUS_D_REGEX, REPLACEMENTS) if transitional_processing
+            output.gsub!(REGEX_STD3_M_REGEX, REPLACEMENTS) unless use_std3_ascii_rules
+
+            output.ascii_only? ? output : output.unicode_normalize!(:nfc)
           end
 
-          def validate_status(label, cp, pos, transitional_processing:, use_std3_ascii_rules:)
-            _, status, = status(cp)
-            return if status == "V"
-            return if !transitional_processing && status == "D"
-            return if !use_std3_ascii_rules && status == "3"
+          def validate_label_status(label, transitional_processing:, use_std3_ascii_rules:)
+            regex =
+              if transitional_processing && use_std3_ascii_rules
+                REGEX_NOT_V
+              elsif transitional_processing
+                REGEX_NOT_V3
+              elsif use_std3_ascii_rules
+                REGEX_NOT_VD
+              else
+                REGEX_NOT_VD3
+              end
+
+            return unless (pos = label.index(regex))
 
-            raise InvalidCodepointError, Validation::Codepoint.cp_error_message(label, cp, pos)
+            raise InvalidCodepointError, error_message(label, pos)
           end
 
-          def status(codepoint)
-            return UTS46_DATA[codepoint] if codepoint < 256
+          private
 
-            index = (UTS46_DATA.bsearch_index { |x| x[0] > codepoint } || UTS46_DATA.length) - 1
-            UTS46_DATA[index] || []
+          def error_message(label, pos)
+            format("Codepoint U+%04X at position %d of %p not allowed in UTS46", label[pos].ord, pos + 1, label)
           end
         end
       end

data/lib/uri/idna/uts46/processing.rb

@@ -2,12 +2,18 @@
 
 require_relative "mapping"
 require_relative "options"
+require_relative "../validation/contextj"
+require_relative "../validation/leading_combining"
 
 module URI
   module IDNA
     module UTS46
       # https://www.unicode.org/reports/tr46/#Processing
       class Processing < BaseProcessing
+        def self.options_class
+          Options
+        end
+
         def initialize(domain_name, **options)
           super
           @domain_name = Mapping.call(
@@ -40,10 +46,6 @@ module URI
 
         private
 
-        def options_class
-          Options
-        end
-
         # https://www.unicode.org/reports/tr46/#Validity_Criteria
         def validate(label, transitional_processing: options.transitional_processing?)
           return if label.empty?
@@ -56,16 +58,13 @@ module URI
             Validation::Label.check_ace_prefix(label)
           end
           Validation::Label.check_dot(label)
-          Validation::Label.check_leading_combining(label)
-
-          label.each_codepoint.with_index do |cp, pos|
-            Mapping.validate_status(
-              label, cp, pos,
-              transitional_processing: transitional_processing, use_std3_ascii_rules: options.use_std3_ascii_rules?
-            )
-
-            Validation::Codepoint.check_contextj(label, cp, pos) if options.check_joiners?
-          end
+          Validation::LeadingCombining.call(label)
+          Mapping.validate_label_status(
+            label,
+            transitional_processing: transitional_processing,
+            use_std3_ascii_rules: options.use_std3_ascii_rules?,
+          )
+          Validation::ContextJ.call(label) if options.check_joiners?
           Validation::Bidi.call(label) if check_bidi?
         end
       end
@@ -76,7 +75,7 @@ module URI
 
       # https://www.unicode.org/reports/tr46/#ToASCII
       class ToASCII < Processing
-        def options_class
+        def self.options_class
           ToASCIIOptions
         end
 

data/lib/uri/idna/validation/bidi.rb

@@ -1,5 +1,7 @@
 # frozen_string_literal: true
 
+require_relative "../data/bidi_classes"
+
 module URI
   module IDNA
     module Validation
@@ -8,55 +10,46 @@ module URI
       # https://datatracker.ietf.org/doc/html/rfc5893#section-2
       module Bidi
         class << self
+          BIDI_R1_RTL = Regexp.new(BIDI_CLASSES["RTL"]).freeze
+          BIDI_R1_LTR = Regexp.new(BIDI_CLASSES["L"]).freeze
+          BIDI_R2 = Regexp.new("#{BIDI_CLASSES['L']}|#{BIDI_CLASSES['UNUSED']}").freeze
+          BIDI_R3 = Regexp.new(
+            "(?:#{"#{BIDI_CLASSES['RTL']}|#{BIDI_CLASSES['AN']}|#{BIDI_CLASSES['EN']}"})#{BIDI_CLASSES['NSM']}*\\z",
+          ).freeze
+          BIDI_R4_EN = Regexp.new(BIDI_CLASSES["EN"]).freeze
+          BIDI_R4_AN = Regexp.new(BIDI_CLASSES["AN"]).freeze
+          BIDI_R5 = Regexp.new("#{BIDI_CLASSES['RTL']}|#{BIDI_CLASSES['AN']}").freeze
+          BIDI_R6 = Regexp.new("(?:#{"#{BIDI_CLASSES['L']}|#{BIDI_CLASSES['EN']}"})#{BIDI_CLASSES['NSM']}*\\z").freeze
+
           def call(label)
             # Bidi rule 1
-            if bidi_class(label[0].ord, "RTL")
-              rtl = true
-            elsif bidi_class(label[0].ord, "L")
+            if BIDI_R1_LTR.match?(label[0])
               rtl = false
+            elsif BIDI_R1_RTL.match?(label[0])
+              rtl = true
             else
               raise BidiError, "First codepoint in label #{label} must be directionality L, R or AL"
             end
 
-            valid_ending = false
-            number_type = nil
-            label.each_codepoint.with_index do |cp, pos|
-              if rtl
-                # Bidi rule 2
-                if bidi_class(cp, "L") || bidi_class(cp, "UNUSED")
-                  raise BidiError, "Invalid direction for codepoint at position #{pos + 1} in a right-to-left label"
-                end
-
-                # Bidi rule 3
-                direction = bidi_class(cp, "RTL") || bidi_class(cp, "EN") || bidi_class(cp, "AN")
-                if direction
-                  valid_ending = true
-                elsif !bidi_class(cp, "NSM")
-                  valid_ending = false
-                end
-                # Bidi rule 4
-                if %w[EN AN].include?(direction)
-                  number_type ||= direction
-                  raise BidiError, "Can not mix numeral types in a right-to-left label" if number_type != direction
-                end
-              else
-                # Bidi rule 5
-                if bidi_class(cp, "RTL") || bidi_class(cp, "AN")
-                  raise BidiError, "Invalid direction for codepoint at position #{pos + 1} in a left-to-right label"
-                end
-
-                # Bidi rule 6
-                if bidi_class(cp, "L") || bidi_class(cp, "EN")
-                  valid_ending = true
-                elsif !bidi_class(cp, "NSM")
-                  valid_ending = false
-                end
+            if rtl
+              # Bidi rule 2
+              if (pos = label.index(BIDI_R2))
+                raise BidiError, "Invalid direction for codepoint at position #{pos + 1} in a right-to-left label"
               end
+              # Bidi rule 3
+              raise BidiError, "Label ends with illegal codepoint directionality" unless label.match?(BIDI_R3)
+              # Bidi rule 4
+              if label.match?(BIDI_R4_EN) && label.match?(BIDI_R4_AN)
+                raise BidiError, "Can not mix numeral types in a right-to-left label"
+              end
+            else
+              # Bidi rule 5
+              if (pos = label.index(BIDI_R5))
+                raise BidiError, "Invalid direction for codepoint at position #{pos + 1} in a left-to-right label"
+              end
+              # Bidi rule 6
+              raise BidiError, "Label ends with illegal codepoint directionality" unless label.match?(BIDI_R6)
             end
-
-            raise BidiError, "Label ends with illegal codepoint directionality" unless valid_ending
-
-            true
           end
 
           # https://www.rfc-editor.org/rfc/rfc5891.html#section-4.2.3.4
@@ -71,22 +64,11 @@ module URI
               end
               next if label.ascii_only?
 
-              label.each_codepoint do |cp|
-                next if cp < 256
-                return true if bidi_class(cp, "RTL") || bidi_class(cp, "AN")
-              end
+              return true if label.match?(BIDI_R5)
             end
 
             false
           end
-
-          private
-
-          def bidi_class(codepoint, bidi_class)
-            return bidi_class if Intranges.contain?(codepoint, BIDI_CLASSES[bidi_class])
-
-            false
-          end
         end
       end
     end

data/lib/uri/idna/validation/contextj.rb

@@ -0,0 +1,62 @@
+# frozen_string_literal: true
+
+require_relative "../data/codepoint_classes"
+require_relative "../data/joining_types"
+require_relative "../data/virama_combining_classes"
+
+module URI
+  module IDNA
+    module Validation
+      # https://datatracker.ietf.org/doc/html/rfc5892
+      module ContextJ
+        class << self
+          CONTEXTJ_REGEX = Regexp.new("[#{CODEPOINT_CLASSES['CONTEXTJ']}]").freeze
+
+          def call(label)
+            return if label.ascii_only?
+
+            offset = 0
+            while (pos = label.index(CONTEXTJ_REGEX, offset))
+              raise InvalidCodepointContextError, error_message(label, pos) unless valid_contextj?(label, pos)
+
+              offset = pos + 1
+            end
+          end
+
+          private
+
+          def valid_contextj?(label, pos)
+            case label[pos]
+              # https://datatracker.ietf.org/doc/html/rfc5892#appendix-A.1
+            when "\u200c"
+              return true if pos > 0 && VIRAMA_COMBINING_CLASSES.match?(label[pos - 1])
+
+              ok = false
+              (pos - 1).downto(0) do |i|
+                joining_type = JOINING_TYPES[label[i]]
+                if [0x4c, 0x44].include?(joining_type)
+                  ok = true
+                  break
+                end
+              end
+              return false unless ok
+
+              (pos + 1).upto(label.length - 1) do |i|
+                joining_type = JOINING_TYPES[label[i]]
+                return true if [0x52, 0x44].include?(joining_type)
+              end
+              # https://datatracker.ietf.org/doc/html/rfc5892#appendix-A.2
+            when "\u200d"
+              return VIRAMA_COMBINING_CLASSES.match?(label[pos - 1]) if pos > 0
+            end
+            false
+          end
+
+          def error_message(label, pos)
+            format("ContextJ codepoint U+%04X at position %d of %p not allowed", label[pos].ord, pos + 1, label)
+          end
+        end
+      end
+    end
+  end
+end

data/lib/uri/idna/validation/contexto.rb

@@ -0,0 +1,61 @@
+# frozen_string_literal: true
+
+require_relative "../data/codepoint_classes"
+require_relative "../data/scripts"
+
+module URI
+  module IDNA
+    module Validation
+      # https://datatracker.ietf.org/doc/html/rfc5892
+      module ContextO
+        class << self
+          CONTEXTO_REGEX = Regexp.new("[#{CODEPOINT_CLASSES['CONTEXTO']}]").freeze
+          CONTEXTO_A4_REGEX = Regexp.new(SCRIPTS["Greek"])
+          CONTEXTO_A5_REGEX = Regexp.new(SCRIPTS["Hebrew"])
+          CONTEXTO_A7_REGEX = Regexp.new("#{SCRIPTS['Hiragana']}|#{SCRIPTS['Katakana']}|#{SCRIPTS['Han']}").freeze
+          CONTEXTO_A8_REGEX = /[\u06F0-\u06F9]/.freeze
+          CONTEXTO_A9_REGEX = /[\u0660-\u0669]/.freeze
+
+          def call(label)
+            offset = 0
+            while (pos = label.index(CONTEXTO_REGEX, offset))
+              raise InvalidCodepointContextError, error_message(label, pos) unless valid_contexto?(label, pos)
+
+              offset = pos + 1
+            end
+          end
+
+          private
+
+          def valid_contexto?(label, pos)
+            case label[pos]
+              # https://datatracker.ietf.org/doc/html/rfc5892#appendix-A.3
+            when "\u00b7"
+              pos > 0 && pos < label.length - 1 ? (label[pos - 1] == "\u006c" && label[pos + 1] == "\u006c") : false
+              # https://datatracker.ietf.org/doc/html/rfc5892#appendix-A.4
+            when "\u0375"
+              pos < label.length - 1 ? CONTEXTO_A4_REGEX.match?(label[pos + 1]) : false
+              # https://datatracker.ietf.org/doc/html/rfc5892#appendix-A.5
+              # https://datatracker.ietf.org/doc/html/rfc5892#appendix-A.6
+            when "\u05f3", "\u05f4"
+              pos > 0 ? CONTEXTO_A5_REGEX.match?(label[pos - 1]) : false
+              # https://datatracker.ietf.org/doc/html/rfc5892#appendix-A.7
+            when "\u30fb"
+              CONTEXTO_A7_REGEX.match?(label)
+              # https://datatracker.ietf.org/doc/html/rfc5892#appendix-A.8
+            when "\u0660".."\u0669"
+              !CONTEXTO_A8_REGEX.match?(label)
+              # https://datatracker.ietf.org/doc/html/rfc5892#appendix-A.9
+            when "\u06f0".."\u06f9"
+              !CONTEXTO_A9_REGEX.match?(label)
+            end
+          end
+
+          def error_message(label, pos)
+            format("ContextO codepoint U+%04X at position %d of %p not allowed", label[pos].ord, pos + 1, label)
+          end
+        end
+      end
+    end
+  end
+end

data/lib/uri/idna/validation/idna_permitted.rb

@@ -0,0 +1,30 @@
+# frozen_string_literal: true
+
+require_relative "../data/codepoint_classes"
+
+module URI
+  module IDNA
+    module Validation
+      module IDNAPermitted
+        class << self
+          IDNA_REGEX = Regexp.new(
+            "[^(#{CODEPOINT_CLASSES['PVALID']}|#{CODEPOINT_CLASSES['CONTEXTJ']}|#{CODEPOINT_CLASSES['CONTEXTO']})]",
+          ).freeze
+
+          # https://datatracker.ietf.org/doc/html/rfc5891#section-4.2.2
+          def call(label)
+            return unless (pos = label.index(IDNA_REGEX))
+
+            raise InvalidCodepointError, error_message(label, pos)
+          end
+
+          private
+
+          def error_message(label, pos)
+            format("Codepoint U+%04X at position %d of %p not allowed in IDNA2008", label[pos].ord, pos + 1, label)
+          end
+        end
+      end
+    end
+  end
+end

data/lib/uri/idna/validation/label.rb

@@ -1,8 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "../intranges"
-require_relative "../data/idna"
-
 module URI
   module IDNA
     module Validation
@@ -11,7 +8,7 @@ module URI
           # 4.1. Input to IDNA Registration
           # https://datatracker.ietf.org/doc/html/rfc5891#section-4.1
           def check_nfc(label)
-            return if label.unicode_normalized?(:nfc)
+            return if label.ascii_only? || label.unicode_normalized?(:nfc)
 
             raise Error, "Label must be in Unicode Normalization Form NFC"
           end
@@ -38,16 +35,6 @@ module URI
             raise Error, "Label must not begin with `xn--`"
           end
 
-          # 4.2.3.2. Leading Combining Marks
-          # https://datatracker.ietf.org/doc/html/rfc5891#section-4.2.3.2
-          def check_leading_combining(label)
-            cp = label[0].ord
-            return if cp < 256
-            return unless Intranges.contain?(cp, INITIAL_COMBINERS)
-
-            raise Error, "Label begins with an illegal combining character"
-          end
-
           def check_dot(label)
             raise Error, "Label must not contain a U+002E ( . ) FULL STOP" if label.include?(".")
           end