json 2.7.3 → 2.10.1

data/lib/json/ext.rb

@@ -6,20 +6,37 @@ module JSON
   # This module holds all the modules/classes that implement JSON's
   # functionality as C extensions.
   module Ext
+    class Parser
+      class << self
+        def parse(...)
+          new(...).parse
+        end
+        alias_method :parse, :parse # Allow redefinition by extensions
+      end
+
+      def initialize(source, opts = nil)
+        @source = source
+        @config = Config.new(opts)
+      end
+
+      def source
+        @source.dup
+      end
+
+      def parse
+        @config.parse(@source)
+      end
+    end
+
+    require 'json/ext/parser'
+    Ext::Parser::Config = Ext::ParserConfig
+    JSON.parser = Ext::Parser
+
     if RUBY_ENGINE == 'truffleruby'
-      require 'json/ext/parser'
-      require 'json/pure'
-      $DEBUG and warn "Using Ext extension for JSON parser and Pure library for JSON generator."
-      JSON.parser = Parser
-      JSON.generator = JSON::Pure::Generator
+      require 'json/truffle_ruby/generator'
+      JSON.generator = ::JSON::TruffleRuby::Generator
     else
-      require 'json/ext/parser'
       require 'json/ext/generator'
-      unless RUBY_ENGINE == 'jruby'
-        require 'json/ext/generator/state'
-      end
-      $DEBUG and warn "Using Ext extension for JSON."
-      JSON.parser = Parser
       JSON.generator = Generator
     end
   end

data/lib/json/{pure → truffle_ruby}/generator.rb

@@ -1,116 +1,123 @@
 # frozen_string_literal: true
 module JSON
-  MAP = {
-    "\x0" => '\u0000',
-    "\x1" => '\u0001',
-    "\x2" => '\u0002',
-    "\x3" => '\u0003',
-    "\x4" => '\u0004',
-    "\x5" => '\u0005',
-    "\x6" => '\u0006',
-    "\x7" => '\u0007',
-    "\b"  =>  '\b',
-    "\t"  =>  '\t',
-    "\n"  =>  '\n',
-    "\xb" => '\u000b',
-    "\f"  =>  '\f',
-    "\r"  =>  '\r',
-    "\xe" => '\u000e',
-    "\xf" => '\u000f',
-    "\x10" => '\u0010',
-    "\x11" => '\u0011',
-    "\x12" => '\u0012',
-    "\x13" => '\u0013',
-    "\x14" => '\u0014',
-    "\x15" => '\u0015',
-    "\x16" => '\u0016',
-    "\x17" => '\u0017',
-    "\x18" => '\u0018',
-    "\x19" => '\u0019',
-    "\x1a" => '\u001a',
-    "\x1b" => '\u001b',
-    "\x1c" => '\u001c',
-    "\x1d" => '\u001d',
-    "\x1e" => '\u001e',
-    "\x1f" => '\u001f',
-    '"'   =>  '\"',
-    '\\'  =>  '\\\\',
-  }.freeze # :nodoc:
-
-  ESCAPE_PATTERN = /[\/"\\\x0-\x1f]/n # :nodoc:
-
-  SCRIPT_SAFE_MAP = MAP.merge(
-    '/'  =>  '\\/',
-    "\u2028".b => '\u2028',
-    "\u2029".b => '\u2029',
-  ).freeze
-
-  SCRIPT_SAFE_ESCAPE_PATTERN = Regexp.union(ESCAPE_PATTERN, "\u2028".b, "\u2029".b)
-
-  # Convert a UTF8 encoded Ruby string _string_ to a JSON string, encoded with
-  # UTF16 big endian characters as \u????, and return it.
-  def utf8_to_json(string, script_safe = false) # :nodoc:
-    string = string.b
-    if script_safe
-      string.gsub!(SCRIPT_SAFE_ESCAPE_PATTERN) { SCRIPT_SAFE_MAP[$&] || $& }
-    else
-      string.gsub!(ESCAPE_PATTERN) { MAP[$&] || $& }
-    end
-    string.force_encoding(::Encoding::UTF_8)
-    string
-  end
+  module TruffleRuby
+    module Generator
+      MAP = {
+        "\x0" => '\u0000',
+        "\x1" => '\u0001',
+        "\x2" => '\u0002',
+        "\x3" => '\u0003',
+        "\x4" => '\u0004',
+        "\x5" => '\u0005',
+        "\x6" => '\u0006',
+        "\x7" => '\u0007',
+        "\b"  =>  '\b',
+        "\t"  =>  '\t',
+        "\n"  =>  '\n',
+        "\xb" => '\u000b',
+        "\f"  =>  '\f',
+        "\r"  =>  '\r',
+        "\xe" => '\u000e',
+        "\xf" => '\u000f',
+        "\x10" => '\u0010',
+        "\x11" => '\u0011',
+        "\x12" => '\u0012',
+        "\x13" => '\u0013',
+        "\x14" => '\u0014',
+        "\x15" => '\u0015',
+        "\x16" => '\u0016',
+        "\x17" => '\u0017',
+        "\x18" => '\u0018',
+        "\x19" => '\u0019',
+        "\x1a" => '\u001a',
+        "\x1b" => '\u001b',
+        "\x1c" => '\u001c',
+        "\x1d" => '\u001d',
+        "\x1e" => '\u001e',
+        "\x1f" => '\u001f',
+        '"'   =>  '\"',
+        '\\'  =>  '\\\\',
+      }.freeze # :nodoc:
+
+      SCRIPT_SAFE_MAP = MAP.merge(
+        '/'  =>  '\\/',
+        "\u2028" => '\u2028',
+        "\u2029" => '\u2029',
+      ).freeze
+
+      SCRIPT_SAFE_ESCAPE_PATTERN = /[\/"\\\x0-\x1f\u2028-\u2029]/
+
+      # Convert a UTF8 encoded Ruby string _string_ to a JSON string, encoded with
+      # UTF16 big endian characters as \u????, and return it.
+      def self.utf8_to_json(string, script_safe = false) # :nodoc:
+        if script_safe
+          if SCRIPT_SAFE_ESCAPE_PATTERN.match?(string)
+            string.gsub(SCRIPT_SAFE_ESCAPE_PATTERN, SCRIPT_SAFE_MAP)
+          else
+            string
+          end
+        else
+          if /["\\\x0-\x1f]/.match?(string)
+            string.gsub(/["\\\x0-\x1f]/, MAP)
+          else
+            string
+          end
+        end
+      end
 
-  def utf8_to_json_ascii(string, script_safe = false) # :nodoc:
-    string = string.b
-    map = script_safe ? SCRIPT_SAFE_MAP : MAP
-    string.gsub!(/[\/"\\\x0-\x1f]/n) { map[$&] || $& }
-    string.gsub!(/(
-      (?:
-       [\xc2-\xdf][\x80-\xbf]    |
-       [\xe0-\xef][\x80-\xbf]{2} |
-       [\xf0-\xf4][\x80-\xbf]{3}
-      )+ |
-      [\x80-\xc1\xf5-\xff]       # invalid
-    )/nx) { |c|
-      c.size == 1 and raise GeneratorError, "invalid utf8 byte: '#{c}'"
-      s = c.encode(::Encoding::UTF_16BE, ::Encoding::UTF_8).unpack('H*')[0]
-      s.force_encoding(::Encoding::ASCII_8BIT)
-      s.gsub!(/.{4}/n, '\\\\u\&')
-      s.force_encoding(::Encoding::UTF_8)
-    }
-    string.force_encoding(::Encoding::UTF_8)
-    string
-  rescue => e
-    raise GeneratorError.wrap(e)
-  end
+      def self.utf8_to_json_ascii(original_string, script_safe = false) # :nodoc:
+        string = original_string.b
+        map = script_safe ? SCRIPT_SAFE_MAP : MAP
+        string.gsub!(/[\/"\\\x0-\x1f]/n) { map[$&] || $& }
+        string.gsub!(/(
+          (?:
+           [\xc2-\xdf][\x80-\xbf]    |
+           [\xe0-\xef][\x80-\xbf]{2} |
+           [\xf0-\xf4][\x80-\xbf]{3}
+          )+ |
+          [\x80-\xc1\xf5-\xff]       # invalid
+        )/nx) { |c|
+          c.size == 1 and raise GeneratorError.new("invalid utf8 byte: '#{c}'", original_string)
+          s = c.encode(::Encoding::UTF_16BE, ::Encoding::UTF_8).unpack('H*')[0]
+          s.force_encoding(::Encoding::BINARY)
+          s.gsub!(/.{4}/n, '\\\\u\&')
+          s.force_encoding(::Encoding::UTF_8)
+        }
+        string.force_encoding(::Encoding::UTF_8)
+        string
+      rescue => e
+        raise GeneratorError.new(e.message, original_string)
+      end
 
-  def valid_utf8?(string)
-    encoding = string.encoding
-    (encoding == Encoding::UTF_8 || encoding == Encoding::ASCII) &&
-      string.valid_encoding?
-  end
-  module_function :utf8_to_json, :utf8_to_json_ascii, :valid_utf8?
+      def self.valid_utf8?(string)
+        encoding = string.encoding
+        (encoding == Encoding::UTF_8 || encoding == Encoding::ASCII) &&
+          string.valid_encoding?
+      end
 
-  module Pure
-    module Generator
       # This class is used to create State instances, that are use to hold data
       # while generating a JSON text from a Ruby data structure.
       class State
+        def self.generate(obj, opts = nil, io = nil)
+          new(opts).generate(obj, io)
+        end
+
         # Creates a State object from _opts_, which ought to be Hash to create
         # a new State instance configured by _opts_, something else to create
         # an unconfigured instance. If _opts_ is a State object, it is just
         # returned.
         def self.from_state(opts)
-          case
-          when self === opts
-            opts
-          when opts.respond_to?(:to_hash)
-            new(opts.to_hash)
-          when opts.respond_to?(:to_h)
-            new(opts.to_h)
-          else
-            SAFE_STATE_PROTOTYPE.dup
+          if opts
+            case
+            when self === opts
+              return opts
+            when opts.respond_to?(:to_hash)
+              return new(opts.to_hash)
+            when opts.respond_to?(:to_h)
+              return new(opts.to_h)
+            end
           end
+          SAFE_STATE_PROTOTYPE.dup
         end
 
         # Instantiates a new State object, configured by _opts_.
@@ -130,7 +137,7 @@ module JSON
         # * *allow_nan*: true if NaN, Infinity, and -Infinity should be
         #   generated, otherwise an exception is thrown, if these values are
         #   encountered. This options defaults to false.
-        def initialize(opts = {})
+        def initialize(opts = nil)
           @indent                = ''
           @space                 = ''
           @space_before          = ''
@@ -138,10 +145,13 @@ module JSON
           @array_nl              = ''
           @allow_nan             = false
           @ascii_only            = false
-          @script_safe          = false
-          @strict                = false
+          @as_json               = false
+          @depth                 = 0
           @buffer_initial_length = 1024
-          configure opts
+          @script_safe           = false
+          @strict                = false
+          @max_nesting           = 100
+          configure(opts) if opts
         end
 
         # This string is used to indent levels in the JSON text.
@@ -161,6 +171,9 @@ module JSON
         # This string is put at the end of a line that holds a JSON array.
         attr_accessor :array_nl
 
+        # This proc converts unsupported types into native JSON types.
+        attr_accessor :as_json
+
         # This integer returns the maximum level of data structure nesting in
         # the generated JSON, max_nesting = 0 if no maximum is checked.
         attr_accessor :max_nesting
@@ -239,13 +252,14 @@ module JSON
           end
 
           # NOTE: If adding new instance variables here, check whether #generate should check them for #generate_json
-          @indent                = opts[:indent] if opts.key?(:indent)
-          @space                 = opts[:space] if opts.key?(:space)
-          @space_before          = opts[:space_before] if opts.key?(:space_before)
-          @object_nl             = opts[:object_nl] if opts.key?(:object_nl)
-          @array_nl              = opts[:array_nl] if opts.key?(:array_nl)
-          @allow_nan             = !!opts[:allow_nan] if opts.key?(:allow_nan)
-          @ascii_only            = opts[:ascii_only] if opts.key?(:ascii_only)
+          @indent                = opts[:indent]        || '' if opts.key?(:indent)
+          @space                 = opts[:space]         || '' if opts.key?(:space)
+          @space_before          = opts[:space_before]  || '' if opts.key?(:space_before)
+          @object_nl             = opts[:object_nl]     || '' if opts.key?(:object_nl)
+          @array_nl              = opts[:array_nl]      || '' if opts.key?(:array_nl)
+          @allow_nan             = !!opts[:allow_nan]         if opts.key?(:allow_nan)
+          @as_json               = opts[:as_json].to_proc     if opts[:as_json]
+          @ascii_only            = opts[:ascii_only]          if opts.key?(:ascii_only)
           @depth                 = opts[:depth] || 0
           @buffer_initial_length ||= opts[:buffer_initial_length]
 
@@ -287,16 +301,27 @@ module JSON
         # returns the result. If no valid JSON document can be
         # created this method raises a
         # GeneratorError exception.
-        def generate(obj)
+        def generate(obj, anIO = nil)
           if @indent.empty? and @space.empty? and @space_before.empty? and @object_nl.empty? and @array_nl.empty? and
-              !@ascii_only and !@script_safe and @max_nesting == 0 and !@strict
+              !@ascii_only and !@script_safe and @max_nesting == 0 and (!@strict || Symbol === obj)
             result = generate_json(obj, ''.dup)
           else
             result = obj.to_json(self)
           end
-          JSON.valid_utf8?(result) or raise GeneratorError,
-            "source sequence #{result.inspect} is illegal/malformed utf-8"
-          result
+          JSON::TruffleRuby::Generator.valid_utf8?(result) or raise GeneratorError.new(
+            "source sequence #{result.inspect} is illegal/malformed utf-8",
+            obj
+          )
+          if anIO
+            anIO.write(result)
+            anIO
+          else
+            result
+          end
+        end
+
+        def generate_new(obj, anIO = nil) # :nodoc:
+          dup.generate(obj, anIO)
         end
 
         # Handles @allow_nan, @buffer_initial_length, other ivars must be the default value (see above)
@@ -307,7 +332,16 @@ module JSON
             first = true
             obj.each_pair do |k,v|
               buf << ',' unless first
-              fast_serialize_string(k.to_s, buf)
+
+              key_str = k.to_s
+              if key_str.class == String
+                fast_serialize_string(key_str, buf)
+              elsif key_str.is_a?(String)
+                generate_json(key_str, buf)
+              else
+                raise TypeError, "#{k.class}#to_s returns an instance of #{key_str.class}, expected a String"
+              end
+
               buf << ':'
               generate_json(v, buf)
               first = false
@@ -323,9 +357,19 @@ module JSON
             end
             buf << ']'
           when String
-            fast_serialize_string(obj, buf)
+            if obj.class == String
+              fast_serialize_string(obj, buf)
+            else
+              buf << obj.to_json(self)
+            end
           when Integer
             buf << obj.to_s
+          when Symbol
+            if @strict
+              fast_serialize_string(obj.name, buf)
+            else
+              buf << obj.to_json(self)
+            end
           else
             # Note: Float is handled this way since Float#to_s is slow anyway
             buf << obj.to_json(self)
@@ -333,24 +377,23 @@ module JSON
         end
 
         # Assumes !@ascii_only, !@script_safe
-        if Regexp.method_defined?(:match?)
-          private def fast_serialize_string(string, buf) # :nodoc:
-            buf << '"'
-            string = string.encode(::Encoding::UTF_8) unless string.encoding == ::Encoding::UTF_8
-            raise GeneratorError, "source sequence is illegal/malformed utf-8" unless string.valid_encoding?
-
-            if /["\\\x0-\x1f]/n.match?(string)
-              buf << string.gsub(/["\\\x0-\x1f]/n, MAP)
-            else
-              buf << string
+        private def fast_serialize_string(string, buf) # :nodoc:
+          buf << '"'
+          unless string.encoding == ::Encoding::UTF_8
+            begin
+              string = string.encode(::Encoding::UTF_8)
+            rescue Encoding::UndefinedConversionError => error
+              raise GeneratorError.new(error.message, string)
             end
-            buf << '"'
           end
-        else
-          # Ruby 2.3 compatibility
-          private def fast_serialize_string(string, buf) # :nodoc:
-            buf << string.to_json(self)
+          raise GeneratorError.new("source sequence is illegal/malformed utf-8", string) unless string.valid_encoding?
+
+          if /["\\\x0-\x1f]/.match?(string)
+            buf << string.gsub(/["\\\x0-\x1f]/, MAP)
+          else
+            buf << string
           end
+          buf << '"'
         end
 
         # Return the value returned by method +name+.
@@ -378,8 +421,20 @@ module JSON
           # it to a JSON string, and returns the result. This is a fallback, if no
           # special method #to_json was defined for some object.
           def to_json(state = nil, *)
-            if state && State.from_state(state).strict?
-              raise GeneratorError, "#{self.class} not allowed in JSON"
+            state = State.from_state(state) if state
+            if state&.strict?
+              value = self
+              if state.strict? && !(false == value || true == value || nil == value || String === value || Array === value || Hash === value || Integer === value || Float === value || Fragment === value)
+                if state.as_json
+                  value = state.as_json.call(value)
+                  unless false == value || true == value || nil == value || String === value || Array === value || Hash === value || Integer === value || Float === value || Fragment === value
+                    raise GeneratorError.new("#{value.class} returned by #{state.as_json} not allowed in JSON", value)
+                  end
+                  value.to_json(state)
+                else
+                  raise GeneratorError.new("#{value.class} not allowed in JSON", value)
+                end
+              end
             else
               to_s.to_json
             end
@@ -406,17 +461,39 @@ module JSON
           end
 
           def json_transform(state)
+            depth = state.depth += 1
+
+            if empty?
+              state.depth -= 1
+              return '{}'
+            end
+
             delim = ",#{state.object_nl}"
             result = +"{#{state.object_nl}"
-            depth = state.depth += 1
             first = true
             indent = !state.object_nl.empty?
             each { |key, value|
               result << delim unless first
               result << state.indent * depth if indent
-              result = +"#{result}#{key.to_s.to_json(state)}#{state.space_before}:#{state.space}"
-              if state.strict? && !(false == value || true == value || nil == value || String === value || Array === value || Hash === value || Integer === value || Float === value)
-                raise GeneratorError, "#{value.class} not allowed in JSON"
+
+              key_str = key.to_s
+              if key_str.is_a?(String)
+                key_json = key_str.to_json(state)
+              else
+                raise TypeError, "#{key.class}#to_s returns an instance of #{key_str.class}, expected a String"
+              end
+
+              result = +"#{result}#{key_json}#{state.space_before}:#{state.space}"
+              if state.strict? && !(false == value || true == value || nil == value || String === value || Array === value || Hash === value || Integer === value || Float === value || Fragment === value)
+                if state.as_json
+                  value = state.as_json.call(value)
+                  unless false == value || true == value || nil == value || String === value || Array === value || Hash === value || Integer === value || Float === value || Fragment === value
+                    raise GeneratorError.new("#{value.class} returned by #{state.as_json} not allowed in JSON", value)
+                  end
+                  result << value.to_json(state)
+                else
+                  raise GeneratorError.new("#{value.class} not allowed in JSON", value)
+                end
               elsif value.respond_to?(:to_json)
                 result << value.to_json(state)
               else
@@ -448,6 +525,13 @@ module JSON
           private
 
           def json_transform(state)
+            depth = state.depth += 1
+
+            if empty?
+              state.depth -= 1
+              return '[]'
+            end
+
             result = '['.dup
             if state.array_nl.empty?
               delim = ","
@@ -455,14 +539,22 @@ module JSON
               result << state.array_nl
               delim = ",#{state.array_nl}"
             end
-            depth = state.depth += 1
+
             first = true
             indent = !state.array_nl.empty?
             each { |value|
               result << delim unless first
               result << state.indent * depth if indent
-              if state.strict? && !(false == value || true == value || nil == value || String === value || Array === value || Hash === value || Integer === value || Float === value)
-                raise GeneratorError, "#{value.class} not allowed in JSON"
+              if state.strict? && !(false == value || true == value || nil == value || String === value || Array === value || Hash === value || Integer === value || Float === value || Fragment === value || Symbol == value)
+                if state.as_json
+                  value = state.as_json.call(value)
+                  unless false == value || true == value || nil == value || String === value || Array === value || Hash === value || Integer === value || Float === value || Fragment === value || Symbol === value
+                    raise GeneratorError.new("#{value.class} returned by #{state.as_json} not allowed in JSON", value)
+                  end
+                  result << value.to_json(state)
+                else
+                  raise GeneratorError.new("#{value.class} not allowed in JSON", value)
+                end
               elsif value.respond_to?(:to_json)
                 result << value.to_json(state)
               else
@@ -484,20 +576,25 @@ module JSON
 
         module Float
           # Returns a JSON string representation for this Float number.
-          def to_json(state = nil, *)
+          def to_json(state = nil, *args)
             state = State.from_state(state)
-            case
-            when infinite?
+            if infinite? || nan?
               if state.allow_nan?
                 to_s
+              elsif state.strict? && state.as_json
+                casted_value = state.as_json.call(self)
+
+                if casted_value.equal?(self)
+                  raise GeneratorError.new("#{self} not allowed in JSON", self)
+                end
+
+                state.check_max_nesting
+                state.depth += 1
+                result = casted_value.to_json(state, *args)
+                state.depth -= 1
+                result
               else
-                raise GeneratorError, "#{self} not allowed in JSON"
-              end
-            when nan?
-              if state.allow_nan?
-                to_s
-              else
-                raise GeneratorError, "#{self} not allowed in JSON"
+                raise GeneratorError.new("#{self} not allowed in JSON", self)
               end
             else
               to_s
@@ -505,6 +602,17 @@ module JSON
           end
         end
 
+        module Symbol
+          def to_json(state = nil, *args)
+            state = State.from_state(state)
+            if state.strict?
+              name.to_json(state, *args)
+            else
+              super
+            end
+          end
+        end
+
         module String
           # This string should be encoded with UTF-8 A call to this method
           # returns a JSON string encoded with UTF16 big endian characters as
@@ -513,17 +621,19 @@ module JSON
             state = State.from_state(state)
             if encoding == ::Encoding::UTF_8
               unless valid_encoding?
-                raise GeneratorError, "source sequence is illegal/malformed utf-8"
+                raise GeneratorError.new("source sequence is illegal/malformed utf-8", self)
               end
               string = self
             else
               string = encode(::Encoding::UTF_8)
             end
             if state.ascii_only?
-              %("#{JSON.utf8_to_json_ascii(string, state.script_safe)}")
+              %("#{JSON::TruffleRuby::Generator.utf8_to_json_ascii(string, state.script_safe)}")
             else
-              %("#{JSON.utf8_to_json(string, state.script_safe)}")
+              %("#{JSON::TruffleRuby::Generator.utf8_to_json(string, state.script_safe)}")
             end
+          rescue Encoding::UndefinedConversionError => error
+            raise ::JSON::GeneratorError.new(error.message, self)
           end
 
           # Module that holds the extending methods if, the String module is

data/lib/json/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module JSON
-  VERSION = '2.7.3'
+  VERSION = '2.10.1'
 end