support RFC3986 [Feature #2542]

* lib/uri/common.rb (URI::REGEXP): move to lib/uri/rfc2396_parser.rb.

* lib/uri/common.rb (URI::Parser): ditto.

* lib/uri/common.rb (URI.split): use RFC3986_Parser.

* lib/uri/common.rb (URI.parse): ditto.

* lib/uri/common.rb (URI.join): ditto.

* lib/uri/common.rb (URI.extract): deprecated.

* lib/uri/common.rb (URI.regexp): ditto.

* lib/uri/rfc2396_parser.rb: added.

* lib/uri/rfc3986_parser.rb: added.

git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@46491 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

3 files changed, 655 insertions, 533 deletions

diff --git a/lib/uri/common.rb b/lib/uri/common.rb
index 385eca2494..b310c4dc6c 100644
--- a/lib/uri/common.rb
+++ b/lib/uri/common.rb
@@ -9,537 +9,13 @@
 # See URI for general documentation
 #
 
-module URI
-  #
-  # Includes URI::REGEXP::PATTERN
-  #
-  module REGEXP
-    #
-    # Patterns used to parse URI's
-    #
-    module PATTERN
-      # :stopdoc:
-
-      # RFC 2396 (URI Generic Syntax)
-      # RFC 2732 (IPv6 Literal Addresses in URL's)
-      # RFC 2373 (IPv6 Addressing Architecture)
-
-      # alpha         = lowalpha | upalpha
-      ALPHA = "a-zA-Z"
-      # alphanum      = alpha | digit
-      ALNUM = "#{ALPHA}\\d"
-
-      # hex           = digit | "A" | "B" | "C" | "D" | "E" | "F" |
-      #                         "a" | "b" | "c" | "d" | "e" | "f"
-      HEX     = "a-fA-F\\d"
-      # escaped       = "%" hex hex
-      ESCAPED = "%[#{HEX}]{2}"
-      # mark          = "-" | "_" | "." | "!" | "~" | "*" | "'" |
-      #                 "(" | ")"
-      # unreserved    = alphanum | mark
-      UNRESERVED = "\\-_.!~*'()#{ALNUM}"
-      # reserved      = ";" | "/" | "?" | ":" | "@" | "&" | "=" | "+" |
-      #                 "$" | ","
-      # reserved      = ";" | "/" | "?" | ":" | "@" | "&" | "=" | "+" |
-      #                 "$" | "," | "[" | "]" (RFC 2732)
-      RESERVED = ";/?:@&=+$,\\[\\]"
-
-      # domainlabel   = alphanum | alphanum *( alphanum | "-" ) alphanum
-      DOMLABEL = "(?:[#{ALNUM}](?:[-#{ALNUM}]*[#{ALNUM}])?)"
-      # toplabel      = alpha | alpha *( alphanum | "-" ) alphanum
-      TOPLABEL = "(?:[#{ALPHA}](?:[-#{ALNUM}]*[#{ALNUM}])?)"
-      # hostname      = *( domainlabel "." ) toplabel [ "." ]
-      HOSTNAME = "(?:#{DOMLABEL}\\.)*#{TOPLABEL}\\.?"
-
-      # :startdoc:
-    end # PATTERN
-
-    # :startdoc:
-  end # REGEXP
+require "uri/rfc2396_parser"
+require "uri/rfc3986_parser"
 
-  # class that Parses String's into URI's
-  #
-  # It contains a Hash set of patterns and Regexp's that match and validate.
-  #
-  class Parser
-    include REGEXP
-
-    #
-    # == Synopsis
-    #
-    #   URI::Parser.new([opts])
-    #
-    # == Args
-    #
-    # The constructor accepts a hash as options for parser.
-    # Keys of options are pattern names of URI components
-    # and values of options are pattern strings.
-    # The constructor generates set of regexps for parsing URIs.
-    #
-    # You can use the following keys:
-    #
-    #   * :ESCAPED (URI::PATTERN::ESCAPED in default)
-    #   * :UNRESERVED (URI::PATTERN::UNRESERVED in default)
-    #   * :DOMLABEL (URI::PATTERN::DOMLABEL in default)
-    #   * :TOPLABEL (URI::PATTERN::TOPLABEL in default)
-    #   * :HOSTNAME (URI::PATTERN::HOSTNAME in default)
-    #
-    # == Examples
-    #
-    #   p = URI::Parser.new(:ESCAPED => "(?:%[a-fA-F0-9]{2}|%u[a-fA-F0-9]{4})")
-    #   u = p.parse("http://example.jp/%uABCD") #=> #<URI::HTTP:0xb78cf4f8 URL:http://example.jp/%uABCD>
-    #   URI.parse(u.to_s) #=> raises URI::InvalidURIError
-    #
-    #   s = "http://example.com/ABCD"
-    #   u1 = p.parse(s) #=> #<URI::HTTP:0xb78c3220 URL:http://example.com/ABCD>
-    #   u2 = URI.parse(s) #=> #<URI::HTTP:0xb78b6d54 URL:http://example.com/ABCD>
-    #   u1 == u2 #=> true
-    #   u1.eql?(u2) #=> false
-    #
-    def initialize(opts = {})
-      @pattern = initialize_pattern(opts)
-      @pattern.each_value {|v| v.freeze}
-      @pattern.freeze
-
-      @regexp = initialize_regexp(@pattern)
-      @regexp.each_value {|v| v.freeze}
-      @regexp.freeze
-    end
-
-    # The Hash of patterns.
-    #
-    # see also URI::Parser.initialize_pattern
-    attr_reader :pattern
-
-    # The Hash of Regexp
-    #
-    # see also URI::Parser.initialize_regexp
-    attr_reader :regexp
-
-    # Returns a split URI against regexp[:ABS_URI]
-    def split(uri)
-      case uri
-      when ''
-        # null uri
-
-      when @regexp[:ABS_URI]
-        scheme, opaque, userinfo, host, port,
-          registry, path, query, fragment = $~[1..-1]
-
-        # URI-reference = [ absoluteURI | relativeURI ] [ "#" fragment ]
-
-        # absoluteURI   = scheme ":" ( hier_part | opaque_part )
-        # hier_part     = ( net_path | abs_path ) [ "?" query ]
-        # opaque_part   = uric_no_slash *uric
-
-        # abs_path      = "/"  path_segments
-        # net_path      = "//" authority [ abs_path ]
-
-        # authority     = server | reg_name
-        # server        = [ [ userinfo "@" ] hostport ]
-
-        if !scheme
-          raise InvalidURIError,
-            "bad URI(absolute but no scheme): #{uri}"
-        end
-        if !opaque && (!path && (!host && !registry))
-          raise InvalidURIError,
-            "bad URI(absolute but no path): #{uri}"
-        end
-
-      when @regexp[:REL_URI]
-        scheme = nil
-        opaque = nil
-
-        userinfo, host, port, registry,
-          rel_segment, abs_path, query, fragment = $~[1..-1]
-        if rel_segment && abs_path
-          path = rel_segment + abs_path
-        elsif rel_segment
-          path = rel_segment
-        elsif abs_path
-          path = abs_path
-        end
-
-        # URI-reference = [ absoluteURI | relativeURI ] [ "#" fragment ]
-
-        # relativeURI   = ( net_path | abs_path | rel_path ) [ "?" query ]
-
-        # net_path      = "//" authority [ abs_path ]
-        # abs_path      = "/"  path_segments
-        # rel_path      = rel_segment [ abs_path ]
-
-        # authority     = server | reg_name
-        # server        = [ [ userinfo "@" ] hostport ]
-
-      else
-        raise InvalidURIError, "bad URI(is not URI?): #{uri}"
-      end
-
-      path = '' if !path && !opaque # (see RFC2396 Section 5.2)
-      ret = [
-        scheme,
-        userinfo, host, port,         # X
-        registry,                     # X
-        path,                         # Y
-        opaque,                       # Y
-        query,
-        fragment
-      ]
-      return ret
-    end
-
-    #
-    # == Args
-    #
-    # +uri+::
-    #    String
-    #
-    # == Description
-    #
-    # parses +uri+ and constructs either matching URI scheme object
-    # (FTP, HTTP, HTTPS, LDAP, LDAPS, or MailTo) or URI::Generic
-    #
-    # == Usage
-    #
-    #   p = URI::Parser.new
-    #   p.parse("ldap://ldap.example.com/dc=example?user=john")
-    #   #=> #<URI::LDAP:0x00000000b9e7e8 URL:ldap://ldap.example.com/dc=example?user=john>
-    #
-    def parse(uri)
-      scheme, userinfo, host, port,
-        registry, path, opaque, query, fragment = self.split(uri)
-
-      if scheme && URI.scheme_list.include?(scheme.upcase)
-        URI.scheme_list[scheme.upcase].new(scheme, userinfo, host, port,
-                                           registry, path, opaque, query,
-                                           fragment, self)
-      else
-        Generic.new(scheme, userinfo, host, port,
-                    registry, path, opaque, query,
-                    fragment, self)
-      end
-    end
-
-
-    #
-    # == Args
-    #
-    # +uris+::
-    #    an Array of Strings
-    #
-    # == Description
-    #
-    # Attempts to parse and merge a set of URIs
-    #
-    def join(*uris)
-      uris[0] = convert_to_uri(uris[0])
-      uris.inject :merge
-    end
-
-    #
-    # :call-seq:
-    #   extract( str )
-    #   extract( str, schemes )
-    #   extract( str, schemes ) {|item| block }
-    #
-    # == Args
-    #
-    # +str+::
-    #    String to search
-    # +schemes+::
-    #    Patterns to apply to +str+
-    #
-    # == Description
-    #
-    # Attempts to parse and merge a set of URIs
-    # If no +block+ given , then returns the result,
-    # else it calls +block+ for each element in result.
-    #
-    # see also URI::Parser.make_regexp
-    #
-    def extract(str, schemes = nil)
-      if block_given?
-        str.scan(make_regexp(schemes)) { yield $& }
-        nil
-      else
-        result = []
-        str.scan(make_regexp(schemes)) { result.push $& }
-        result
-      end
-    end
-
-    # returns Regexp that is default self.regexp[:ABS_URI_REF],
-    # unless +schemes+ is provided. Then it is a Regexp.union with self.pattern[:X_ABS_URI]
-    def make_regexp(schemes = nil)
-      unless schemes
-        @regexp[:ABS_URI_REF]
-      else
-        /(?=#{Regexp.union(*schemes)}:)#{@pattern[:X_ABS_URI]}/x
-      end
-    end
-
-    #
-    # :call-seq:
-    #   escape( str )
-    #   escape( str, unsafe )
-    #
-    # == Args
-    #
-    # +str+::
-    #    String to make safe
-    # +unsafe+::
-    #    Regexp to apply. Defaults to self.regexp[:UNSAFE]
-    #
-    # == Description
-    #
-    # constructs a safe String from +str+, removing unsafe characters,
-    # replacing them with codes.
-    #
-    def escape(str, unsafe = @regexp[:UNSAFE])
-      unless unsafe.kind_of?(Regexp)
-        # perhaps unsafe is String object
-        unsafe = Regexp.new("[#{Regexp.quote(unsafe)}]", false)
-      end
-      str.gsub(unsafe) do
-        us = $&
-        tmp = ''
-        us.each_byte do |uc|
-          tmp << sprintf('%%%02X', uc)
-        end
-        tmp
-      end.force_encoding(Encoding::US_ASCII)
-    end
-
-    #
-    # :call-seq:
-    #   unescape( str )
-    #   unescape( str, unsafe )
-    #
-    # == Args
-    #
-    # +str+::
-    #    String to remove escapes from
-    # +unsafe+::
-    #    Regexp to apply. Defaults to self.regexp[:ESCAPED]
-    #
-    # == Description
-    #
-    # Removes escapes from +str+
-    #
-    def unescape(str, escaped = @regexp[:ESCAPED])
-      str.gsub(escaped) { [$&[1, 2].hex].pack('C') }.force_encoding(str.encoding)
-    end
-
-    @@to_s = Kernel.instance_method(:to_s)
-    def inspect
-      @@to_s.bind(self).call
-    end
-
-    private
-
-    # Constructs the default Hash of patterns
-    def initialize_pattern(opts = {})
-      ret = {}
-      ret[:ESCAPED] = escaped = (opts.delete(:ESCAPED) || PATTERN::ESCAPED)
-      ret[:UNRESERVED] = unreserved = opts.delete(:UNRESERVED) || PATTERN::UNRESERVED
-      ret[:RESERVED] = reserved = opts.delete(:RESERVED) || PATTERN::RESERVED
-      ret[:DOMLABEL] = opts.delete(:DOMLABEL) || PATTERN::DOMLABEL
-      ret[:TOPLABEL] = opts.delete(:TOPLABEL) || PATTERN::TOPLABEL
-      ret[:HOSTNAME] = hostname = opts.delete(:HOSTNAME)
-
-      # RFC 2396 (URI Generic Syntax)
-      # RFC 2732 (IPv6 Literal Addresses in URL's)
-      # RFC 2373 (IPv6 Addressing Architecture)
-
-      # uric          = reserved | unreserved | escaped
-      ret[:URIC] = uric = "(?:[#{unreserved}#{reserved}]|#{escaped})"
-      # uric_no_slash = unreserved | escaped | ";" | "?" | ":" | "@" |
-      #                 "&" | "=" | "+" | "$" | ","
-      ret[:URIC_NO_SLASH] = uric_no_slash = "(?:[#{unreserved};?:@&=+$,]|#{escaped})"
-      # query         = *uric
-      ret[:QUERY] = query = "#{uric}*"
-      # fragment      = *uric
-      ret[:FRAGMENT] = fragment = "#{uric}*"
-
-      # hostname      = *( domainlabel "." ) toplabel [ "." ]
-      # reg-name      = *( unreserved / pct-encoded / sub-delims ) # RFC3986
-      unless hostname
-        ret[:HOSTNAME] = hostname = "(?:[a-zA-Z0-9\\-.]|%\\h\\h)+"
-      end
-
-      # RFC 2373, APPENDIX B:
-      # IPv6address = hexpart [ ":" IPv4address ]
-      # IPv4address   = 1*3DIGIT "." 1*3DIGIT "." 1*3DIGIT "." 1*3DIGIT
-      # hexpart = hexseq | hexseq "::" [ hexseq ] | "::" [ hexseq ]
-      # hexseq  = hex4 *( ":" hex4)
-      # hex4    = 1*4HEXDIG
-      #
-      # XXX: This definition has a flaw. "::" + IPv4address must be
-      # allowed too.  Here is a replacement.
-      #
-      # IPv4address = 1*3DIGIT "." 1*3DIGIT "." 1*3DIGIT "." 1*3DIGIT
-      ret[:IPV4ADDR] = ipv4addr = "\\d{1,3}\\.\\d{1,3}\\.\\d{1,3}\\.\\d{1,3}"
-      # hex4     = 1*4HEXDIG
-      hex4 = "[#{PATTERN::HEX}]{1,4}"
-      # lastpart = hex4 | IPv4address
-      lastpart = "(?:#{hex4}|#{ipv4addr})"
-      # hexseq1  = *( hex4 ":" ) hex4
-      hexseq1 = "(?:#{hex4}:)*#{hex4}"
-      # hexseq2  = *( hex4 ":" ) lastpart
-      hexseq2 = "(?:#{hex4}:)*#{lastpart}"
-      # IPv6address = hexseq2 | [ hexseq1 ] "::" [ hexseq2 ]
-      ret[:IPV6ADDR] = ipv6addr = "(?:#{hexseq2}|(?:#{hexseq1})?::(?:#{hexseq2})?)"
-
-      # IPv6prefix  = ( hexseq1 | [ hexseq1 ] "::" [ hexseq1 ] ) "/" 1*2DIGIT
-      # unused
-
-      # ipv6reference = "[" IPv6address "]" (RFC 2732)
-      ret[:IPV6REF] = ipv6ref = "\\[#{ipv6addr}\\]"
-
-      # host          = hostname | IPv4address
-      # host          = hostname | IPv4address | IPv6reference (RFC 2732)
-      ret[:HOST] = host = "(?:#{hostname}|#{ipv4addr}|#{ipv6ref})"
-      # port          = *digit
-      port = '\d*'
-      # hostport      = host [ ":" port ]
-      ret[:HOSTPORT] = hostport = "#{host}(?::#{port})?"
-
-      # userinfo      = *( unreserved | escaped |
-      #                    ";" | ":" | "&" | "=" | "+" | "$" | "," )
-      ret[:USERINFO] = userinfo = "(?:[#{unreserved};:&=+$,]|#{escaped})*"
-
-      # pchar         = unreserved | escaped |
-      #                 ":" | "@" | "&" | "=" | "+" | "$" | ","
-      pchar = "(?:[#{unreserved}:@&=+$,]|#{escaped})"
-      # param         = *pchar
-      param = "#{pchar}*"
-      # segment       = *pchar *( ";" param )
-      segment = "#{pchar}*(?:;#{param})*"
-      # path_segments = segment *( "/" segment )
-      ret[:PATH_SEGMENTS] = path_segments = "#{segment}(?:/#{segment})*"
-
-      # server        = [ [ userinfo "@" ] hostport ]
-      server = "(?:#{userinfo}@)?#{hostport}"
-      # reg_name      = 1*( unreserved | escaped | "$" | "," |
-      #                     ";" | ":" | "@" | "&" | "=" | "+" )
-      ret[:REG_NAME] = reg_name = "(?:[#{unreserved}$,;:@&=+]|#{escaped})+"
-      # authority     = server | reg_name
-      authority = "(?:#{server}|#{reg_name})"
-
-      # rel_segment   = 1*( unreserved | escaped |
-      #                     ";" | "@" | "&" | "=" | "+" | "$" | "," )
-      ret[:REL_SEGMENT] = rel_segment = "(?:[#{unreserved};@&=+$,]|#{escaped})+"
-
-      # scheme        = alpha *( alpha | digit | "+" | "-" | "." )
-      ret[:SCHEME] = scheme = "[#{PATTERN::ALPHA}][\\-+.#{PATTERN::ALPHA}\\d]*"
-
-      # abs_path      = "/"  path_segments
-      ret[:ABS_PATH] = abs_path = "/#{path_segments}"
-      # rel_path      = rel_segment [ abs_path ]
-      ret[:REL_PATH] = rel_path = "#{rel_segment}(?:#{abs_path})?"
-      # net_path      = "//" authority [ abs_path ]
-      ret[:NET_PATH] = net_path = "//#{authority}(?:#{abs_path})?"
-
-      # hier_part     = ( net_path | abs_path ) [ "?" query ]
-      ret[:HIER_PART] = hier_part = "(?:#{net_path}|#{abs_path})(?:\\?(?:#{query}))?"
-      # opaque_part   = uric_no_slash *uric
-      ret[:OPAQUE_PART] = opaque_part = "#{uric_no_slash}#{uric}*"
-
-      # absoluteURI   = scheme ":" ( hier_part | opaque_part )
-      ret[:ABS_URI] = abs_uri = "#{scheme}:(?:#{hier_part}|#{opaque_part})"
-      # relativeURI   = ( net_path | abs_path | rel_path ) [ "?" query ]
-      ret[:REL_URI] = rel_uri = "(?:#{net_path}|#{abs_path}|#{rel_path})(?:\\?#{query})?"
-
-      # URI-reference = [ absoluteURI | relativeURI ] [ "#" fragment ]
-      ret[:URI_REF] = "(?:#{abs_uri}|#{rel_uri})?(?:##{fragment})?"
-
-      ret[:X_ABS_URI] = "
-        (#{scheme}):                           (?# 1: scheme)
-        (?:
-           (#{opaque_part})                    (?# 2: opaque)
-        |
-           (?:(?:
-             //(?:
-                 (?:(?:(#{userinfo})@)?        (?# 3: userinfo)
-                   (?:(#{host})(?::(\\d*))?))? (?# 4: host, 5: port)
-               |
-                 (#{reg_name})                 (?# 6: registry)
-               )
-             |
-             (?!//))                           (?# XXX: '//' is the mark for hostport)
-             (#{abs_path})?                    (?# 7: path)
-           )(?:\\?(#{query}))?                 (?# 8: query)
-        )
-        (?:\\#(#{fragment}))?                  (?# 9: fragment)
-      "
-
-      ret[:X_REL_URI] = "
-        (?:
-          (?:
-            //
-            (?:
-              (?:(#{userinfo})@)?       (?# 1: userinfo)
-                (#{host})?(?::(\\d*))?  (?# 2: host, 3: port)
-            |
-              (#{reg_name})             (?# 4: registry)
-            )
-          )
-        |
-          (#{rel_segment})              (?# 5: rel_segment)
-        )?
-        (#{abs_path})?                  (?# 6: abs_path)
-        (?:\\?(#{query}))?              (?# 7: query)
-        (?:\\#(#{fragment}))?           (?# 8: fragment)
-      "
-
-      ret
-    end
-
-    # Constructs the default Hash of Regexp's
-    def initialize_regexp(pattern)
-      ret = {}
-
-      # for URI::split
-      ret[:ABS_URI] = Regexp.new('\A\s*' + pattern[:X_ABS_URI] + '\s*\z', Regexp::EXTENDED)
-      ret[:REL_URI] = Regexp.new('\A\s*' + pattern[:X_REL_URI] + '\s*\z', Regexp::EXTENDED)
-
-      # for URI::extract
-      ret[:URI_REF]     = Regexp.new(pattern[:URI_REF])
-      ret[:ABS_URI_REF] = Regexp.new(pattern[:X_ABS_URI], Regexp::EXTENDED)
-      ret[:REL_URI_REF] = Regexp.new(pattern[:X_REL_URI], Regexp::EXTENDED)
-
-      # for URI::escape/unescape
-      ret[:ESCAPED] = Regexp.new(pattern[:ESCAPED])
-      ret[:UNSAFE]  = Regexp.new("[^#{pattern[:UNRESERVED]}#{pattern[:RESERVED]}]")
-
-      # for Generic#initialize
-      ret[:SCHEME]   = Regexp.new("\\A#{pattern[:SCHEME]}\\z")
-      ret[:USERINFO] = Regexp.new("\\A#{pattern[:USERINFO]}\\z")
-      ret[:HOST]     = Regexp.new("\\A#{pattern[:HOST]}\\z")
-      ret[:PORT]     = Regexp.new("\\A#{pattern[:PORT]}\\z")
-      ret[:OPAQUE]   = Regexp.new("\\A#{pattern[:OPAQUE_PART]}\\z")
-      ret[:REGISTRY] = Regexp.new("\\A#{pattern[:REG_NAME]}\\z")
-      ret[:ABS_PATH] = Regexp.new("\\A#{pattern[:ABS_PATH]}\\z")
-      ret[:REL_PATH] = Regexp.new("\\A#{pattern[:REL_PATH]}\\z")
-      ret[:QUERY]    = Regexp.new("\\A#{pattern[:QUERY]}\\z")
-      ret[:FRAGMENT] = Regexp.new("\\A#{pattern[:FRAGMENT]}\\z")
-
-      ret
-    end
-
-    def convert_to_uri(uri)
-      if uri.is_a?(URI::Generic)
-        uri
-      elsif uri = String.try_convert(uri)
-        parse(uri)
-      else
-        raise ArgumentError,
-          "bad argument (expected URI object or URI string)"
-      end
-    end
-
-  end # class Parser
+module URI
+  REGEXP = RFC2396_REGEXP
+  Parser = RFC2396_Parser
+  RFC3986_PARSER = RFC3986_Parser.new
 
   # URI::Parser.new
   DEFAULT_PARSER = Parser.new
@@ -709,7 +185,7 @@ module URI
   #   # => ["http", nil, "www.ruby-lang.org", nil, nil, "/", nil, nil, nil]
   #
   def self.split(uri)
-    DEFAULT_PARSER.split(uri)
+    RFC3986_PARSER.split(uri)
   end
 
   #
@@ -744,7 +220,7 @@ module URI
   #   # => "www.ruby-lang.org"
   #
   def self.parse(uri)
-    DEFAULT_PARSER.parse(uri)
+    RFC3986_PARSER.parse(uri)
   end
 
   #
@@ -782,7 +258,7 @@ module URI
   #
   #
   def self.join(*str)
-    DEFAULT_PARSER.join(*str)
+    RFC3986_PARSER.join(*str)
   end
 
   #
@@ -810,6 +286,7 @@ module URI
   #   # => ["http://foo.example.com/bla", "mailto:test@example.com"]
   #
   def self.extract(str, schemes = nil, &block)
+    warn "#{caller(1)[0]}: warning: URI.extract is obsolete" if $VERBOSE
     DEFAULT_PARSER.extract(str, schemes, &block)
   end
 
@@ -845,6 +322,7 @@ module URI
   #   end
   #
   def self.regexp(schemes = nil)
+    warn "#{caller(1)[0]}: warning: URI.regexp is obsolete" if $VERBOSE
     DEFAULT_PARSER.make_regexp(schemes)
   end
 
diff --git a/lib/uri/rfc2396_parser.rb b/lib/uri/rfc2396_parser.rb
new file mode 100644
index 0000000000..50e3ae6973
--- /dev/null
+++ b/lib/uri/rfc2396_parser.rb
@@ -0,0 +1,543 @@
+#--
+# = uri/common.rb
+#
+# Author:: Akira Yamada <akira@ruby-lang.org>
+# Revision:: $Id$
+# License::
+#   You can redistribute it and/or modify it under the same term as Ruby.
+#
+# See URI for general documentation
+#
+
+module URI
+  #
+  # Includes URI::REGEXP::PATTERN
+  #
+  module RFC2396_REGEXP
+    #
+    # Patterns used to parse URI's
+    #
+    module PATTERN
+      # :stopdoc:
+
+      # RFC 2396 (URI Generic Syntax)
+      # RFC 2732 (IPv6 Literal Addresses in URL's)
+      # RFC 2373 (IPv6 Addressing Architecture)
+
+      # alpha         = lowalpha | upalpha
+      ALPHA = "a-zA-Z"
+      # alphanum      = alpha | digit
+      ALNUM = "#{ALPHA}\\d"
+
+      # hex           = digit | "A" | "B" | "C" | "D" | "E" | "F" |
+      #                         "a" | "b" | "c" | "d" | "e" | "f"
+      HEX     = "a-fA-F\\d"
+      # escaped       = "%" hex hex
+      ESCAPED = "%[#{HEX}]{2}"
+      # mark          = "-" | "_" | "." | "!" | "~" | "*" | "'" |
+      #                 "(" | ")"
+      # unreserved    = alphanum | mark
+      UNRESERVED = "\\-_.!~*'()#{ALNUM}"
+      # reserved      = ";" | "/" | "?" | ":" | "@" | "&" | "=" | "+" |
+      #                 "$" | ","
+      # reserved      = ";" | "/" | "?" | ":" | "@" | "&" | "=" | "+" |
+      #                 "$" | "," | "[" | "]" (RFC 2732)
+      RESERVED = ";/?:@&=+$,\\[\\]"
+
+      # domainlabel   = alphanum | alphanum *( alphanum | "-" ) alphanum
+      DOMLABEL = "(?:[#{ALNUM}](?:[-#{ALNUM}]*[#{ALNUM}])?)"
+      # toplabel      = alpha | alpha *( alphanum | "-" ) alphanum
+      TOPLABEL = "(?:[#{ALPHA}](?:[-#{ALNUM}]*[#{ALNUM}])?)"
+      # hostname      = *( domainlabel "." ) toplabel [ "." ]
+      HOSTNAME = "(?:#{DOMLABEL}\\.)*#{TOPLABEL}\\.?"
+
+      # :startdoc:
+    end # PATTERN
+
+    # :startdoc:
+  end # REGEXP
+
+  # class that Parses String's into URI's
+  #
+  # It contains a Hash set of patterns and Regexp's that match and validate.
+  #
+  class RFC2396_Parser
+    include RFC2396_REGEXP
+
+    #
+    # == Synopsis
+    #
+    #   URI::Parser.new([opts])
+    #
+    # == Args
+    #
+    # The constructor accepts a hash as options for parser.
+    # Keys of options are pattern names of URI components
+    # and values of options are pattern strings.
+    # The constructor generates set of regexps for parsing URIs.
+    #
+    # You can use the following keys:
+    #
+    #   * :ESCAPED (URI::PATTERN::ESCAPED in default)
+    #   * :UNRESERVED (URI::PATTERN::UNRESERVED in default)
+    #   * :DOMLABEL (URI::PATTERN::DOMLABEL in default)
+    #   * :TOPLABEL (URI::PATTERN::TOPLABEL in default)
+    #   * :HOSTNAME (URI::PATTERN::HOSTNAME in default)
+    #
+    # == Examples
+    #
+    #   p = URI::Parser.new(:ESCAPED => "(?:%[a-fA-F0-9]{2}|%u[a-fA-F0-9]{4})")
+    #   u = p.parse("http://example.jp/%uABCD") #=> #<URI::HTTP:0xb78cf4f8 URL:http://example.jp/%uABCD>
+    #   URI.parse(u.to_s) #=> raises URI::InvalidURIError
+    #
+    #   s = "http://example.com/ABCD"
+    #   u1 = p.parse(s) #=> #<URI::HTTP:0xb78c3220 URL:http://example.com/ABCD>
+    #   u2 = URI.parse(s) #=> #<URI::HTTP:0xb78b6d54 URL:http://example.com/ABCD>
+    #   u1 == u2 #=> true
+    #   u1.eql?(u2) #=> false
+    #
+    def initialize(opts = {})
+      @pattern = initialize_pattern(opts)
+      @pattern.each_value {|v| v.freeze}
+      @pattern.freeze
+
+      @regexp = initialize_regexp(@pattern)
+      @regexp.each_value {|v| v.freeze}
+      @regexp.freeze
+    end
+
+    # The Hash of patterns.
+    #
+    # see also URI::Parser.initialize_pattern
+    attr_reader :pattern
+
+    # The Hash of Regexp
+    #
+    # see also URI::Parser.initialize_regexp
+    attr_reader :regexp
+
+    # Returns a split URI against regexp[:ABS_URI]
+    def split(uri)
+      case uri
+      when ''
+        # null uri
+
+      when @regexp[:ABS_URI]
+        scheme, opaque, userinfo, host, port,
+          registry, path, query, fragment = $~[1..-1]
+
+        # URI-reference = [ absoluteURI | relativeURI ] [ "#" fragment ]
+
+        # absoluteURI   = scheme ":" ( hier_part | opaque_part )
+        # hier_part     = ( net_path | abs_path ) [ "?" query ]
+        # opaque_part   = uric_no_slash *uric
+
+        # abs_path      = "/"  path_segments
+        # net_path      = "//" authority [ abs_path ]
+
+        # authority     = server | reg_name
+        # server        = [ [ userinfo "@" ] hostport ]
+
+        if !scheme
+          raise InvalidURIError,
+            "bad URI(absolute but no scheme): #{uri}"
+        end
+        if !opaque && (!path && (!host && !registry))
+          raise InvalidURIError,
+            "bad URI(absolute but no path): #{uri}"
+        end
+
+      when @regexp[:REL_URI]
+        scheme = nil
+        opaque = nil
+
+        userinfo, host, port, registry,
+          rel_segment, abs_path, query, fragment = $~[1..-1]
+        if rel_segment && abs_path
+          path = rel_segment + abs_path
+        elsif rel_segment
+          path = rel_segment
+        elsif abs_path
+          path = abs_path
+        end
+
+        # URI-reference = [ absoluteURI | relativeURI ] [ "#" fragment ]
+
+        # relativeURI   = ( net_path | abs_path | rel_path ) [ "?" query ]
+
+        # net_path      = "//" authority [ abs_path ]
+        # abs_path      = "/"  path_segments
+        # rel_path      = rel_segment [ abs_path ]
+
+        # authority     = server | reg_name
+        # server        = [ [ userinfo "@" ] hostport ]
+
+      else
+        raise InvalidURIError, "bad URI(is not URI?): #{uri}"
+      end
+
+      path = '' if !path && !opaque # (see RFC2396 Section 5.2)
+      ret = [
+        scheme,
+        userinfo, host, port,         # X
+        registry,                     # X
+        path,                         # Y
+        opaque,                       # Y
+        query,
+        fragment
+      ]
+      return ret
+    end
+
+    #
+    # == Args
+    #
+    # +uri+::
+    #    String
+    #
+    # == Description
+    #
+    # parses +uri+ and constructs either matching URI scheme object
+    # (FTP, HTTP, HTTPS, LDAP, LDAPS, or MailTo) or URI::Generic
+    #
+    # == Usage
+    #
+    #   p = URI::Parser.new
+    #   p.parse("ldap://ldap.example.com/dc=example?user=john")
+    #   #=> #<URI::LDAP:0x00000000b9e7e8 URL:ldap://ldap.example.com/dc=example?user=john>
+    #
+    def parse(uri)
+      scheme, userinfo, host, port,
+        registry, path, opaque, query, fragment = self.split(uri)
+
+      if scheme && URI.scheme_list.include?(scheme.upcase)
+        URI.scheme_list[scheme.upcase].new(scheme, userinfo, host, port,
+                                           registry, path, opaque, query,
+                                           fragment, self)
+      else
+        Generic.new(scheme, userinfo, host, port,
+                    registry, path, opaque, query,
+                    fragment, self)
+      end
+    end
+
+
+    #
+    # == Args
+    #
+    # +uris+::
+    #    an Array of Strings
+    #
+    # == Description
+    #
+    # Attempts to parse and merge a set of URIs
+    #
+    def join(*uris)
+      uris[0] = convert_to_uri(uris[0])
+      uris.inject :merge
+    end
+
+    #
+    # :call-seq:
+    #   extract( str )
+    #   extract( str, schemes )
+    #   extract( str, schemes ) {|item| block }
+    #
+    # == Args
+    #
+    # +str+::
+    #    String to search
+    # +schemes+::
+    #    Patterns to apply to +str+
+    #
+    # == Description
+    #
+    # Attempts to parse and merge a set of URIs
+    # If no +block+ given , then returns the result,
+    # else it calls +block+ for each element in result.
+    #
+    # see also URI::Parser.make_regexp
+    #
+    def extract(str, schemes = nil)
+      if block_given?
+        str.scan(make_regexp(schemes)) { yield $& }
+        nil
+      else
+        result = []
+        str.scan(make_regexp(schemes)) { result.push $& }
+        result
+      end
+    end
+
+    # returns Regexp that is default self.regexp[:ABS_URI_REF],
+    # unless +schemes+ is provided. Then it is a Regexp.union with self.pattern[:X_ABS_URI]
+    def make_regexp(schemes = nil)
+      unless schemes
+        @regexp[:ABS_URI_REF]
+      else
+        /(?=#{Regexp.union(*schemes)}:)#{@pattern[:X_ABS_URI]}/x
+      end
+    end
+
+    #
+    # :call-seq:
+    #   escape( str )
+    #   escape( str, unsafe )
+    #
+    # == Args
+    #
+    # +str+::
+    #    String to make safe
+    # +unsafe+::
+    #    Regexp to apply. Defaults to self.regexp[:UNSAFE]
+    #
+    # == Description
+    #
+    # constructs a safe String from +str+, removing unsafe characters,
+    # replacing them with codes.
+    #
+    def escape(str, unsafe = @regexp[:UNSAFE])
+      unless unsafe.kind_of?(Regexp)
+        # perhaps unsafe is String object
+        unsafe = Regexp.new("[#{Regexp.quote(unsafe)}]", false)
+      end
+      str.gsub(unsafe) do
+        us = $&
+        tmp = ''
+        us.each_byte do |uc|
+          tmp << sprintf('%%%02X', uc)
+        end
+        tmp
+      end.force_encoding(Encoding::US_ASCII)
+    end
+
+    #
+    # :call-seq:
+    #   unescape( str )
+    #   unescape( str, unsafe )
+    #
+    # == Args
+    #
+    # +str+::
+    #    String to remove escapes from
+    # +unsafe+::
+    #    Regexp to apply. Defaults to self.regexp[:ESCAPED]
+    #
+    # == Description
+    #
+    # Removes escapes from +str+
+    #
+    def unescape(str, escaped = @regexp[:ESCAPED])
+      str.gsub(escaped) { [$&[1, 2].hex].pack('C') }.force_encoding(str.encoding)
+    end
+
+    @@to_s = Kernel.instance_method(:to_s)
+    def inspect
+      @@to_s.bind(self).call
+    end
+
+    private
+
+    # Constructs the default Hash of patterns
+    def initialize_pattern(opts = {})
+      ret = {}
+      ret[:ESCAPED] = escaped = (opts.delete(:ESCAPED) || PATTERN::ESCAPED)
+      ret[:UNRESERVED] = unreserved = opts.delete(:UNRESERVED) || PATTERN::UNRESERVED
+      ret[:RESERVED] = reserved = opts.delete(:RESERVED) || PATTERN::RESERVED
+      ret[:DOMLABEL] = opts.delete(:DOMLABEL) || PATTERN::DOMLABEL
+      ret[:TOPLABEL] = opts.delete(:TOPLABEL) || PATTERN::TOPLABEL
+      ret[:HOSTNAME] = hostname = opts.delete(:HOSTNAME)
+
+      # RFC 2396 (URI Generic Syntax)
+      # RFC 2732 (IPv6 Literal Addresses in URL's)
+      # RFC 2373 (IPv6 Addressing Architecture)
+
+      # uric          = reserved | unreserved | escaped
+      ret[:URIC] = uric = "(?:[#{unreserved}#{reserved}]|#{escaped})"
+      # uric_no_slash = unreserved | escaped | ";" | "?" | ":" | "@" |
+      #                 "&" | "=" | "+" | "$" | ","
+      ret[:URIC_NO_SLASH] = uric_no_slash = "(?:[#{unreserved};?:@&=+$,]|#{escaped})"
+      # query         = *uric
+      ret[:QUERY] = query = "#{uric}*"
+      # fragment      = *uric
+      ret[:FRAGMENT] = fragment = "#{uric}*"
+
+      # hostname      = *( domainlabel "." ) toplabel [ "." ]
+      # reg-name      = *( unreserved / pct-encoded / sub-delims ) # RFC3986
+      unless hostname
+        ret[:HOSTNAME] = hostname = "(?:[a-zA-Z0-9\\-.]|%\\h\\h)+"
+      end
+
+      # RFC 2373, APPENDIX B:
+      # IPv6address = hexpart [ ":" IPv4address ]
+      # IPv4address   = 1*3DIGIT "." 1*3DIGIT "." 1*3DIGIT "." 1*3DIGIT
+      # hexpart = hexseq | hexseq "::" [ hexseq ] | "::" [ hexseq ]
+      # hexseq  = hex4 *( ":" hex4)
+      # hex4    = 1*4HEXDIG
+      #
+      # XXX: This definition has a flaw. "::" + IPv4address must be
+      # allowed too.  Here is a replacement.
+      #
+      # IPv4address = 1*3DIGIT "." 1*3DIGIT "." 1*3DIGIT "." 1*3DIGIT
+      ret[:IPV4ADDR] = ipv4addr = "\\d{1,3}\\.\\d{1,3}\\.\\d{1,3}\\.\\d{1,3}"
+      # hex4     = 1*4HEXDIG
+      hex4 = "[#{PATTERN::HEX}]{1,4}"
+      # lastpart = hex4 | IPv4address
+      lastpart = "(?:#{hex4}|#{ipv4addr})"
+      # hexseq1  = *( hex4 ":" ) hex4
+      hexseq1 = "(?:#{hex4}:)*#{hex4}"
+      # hexseq2  = *( hex4 ":" ) lastpart
+      hexseq2 = "(?:#{hex4}:)*#{lastpart}"
+      # IPv6address = hexseq2 | [ hexseq1 ] "::" [ hexseq2 ]
+      ret[:IPV6ADDR] = ipv6addr = "(?:#{hexseq2}|(?:#{hexseq1})?::(?:#{hexseq2})?)"
+
+      # IPv6prefix  = ( hexseq1 | [ hexseq1 ] "::" [ hexseq1 ] ) "/" 1*2DIGIT
+      # unused
+
+      # ipv6reference = "[" IPv6address "]" (RFC 2732)
+      ret[:IPV6REF] = ipv6ref = "\\[#{ipv6addr}\\]"
+
+      # host          = hostname | IPv4address
+      # host          = hostname | IPv4address | IPv6reference (RFC 2732)
+      ret[:HOST] = host = "(?:#{hostname}|#{ipv4addr}|#{ipv6ref})"
+      # port          = *digit
+      port = '\d*'
+      # hostport      = host [ ":" port ]
+      ret[:HOSTPORT] = hostport = "#{host}(?::#{port})?"
+
+      # userinfo      = *( unreserved | escaped |
+      #                    ";" | ":" | "&" | "=" | "+" | "$" | "," )
+      ret[:USERINFO] = userinfo = "(?:[#{unreserved};:&=+$,]|#{escaped})*"
+
+      # pchar         = unreserved | escaped |
+      #                 ":" | "@" | "&" | "=" | "+" | "$" | ","
+      pchar = "(?:[#{unreserved}:@&=+$,]|#{escaped})"
+      # param         = *pchar
+      param = "#{pchar}*"
+      # segment       = *pchar *( ";" param )
+      segment = "#{pchar}*(?:;#{param})*"
+      # path_segments = segment *( "/" segment )
+      ret[:PATH_SEGMENTS] = path_segments = "#{segment}(?:/#{segment})*"
+
+      # server        = [ [ userinfo "@" ] hostport ]
+      server = "(?:#{userinfo}@)?#{hostport}"
+      # reg_name      = 1*( unreserved | escaped | "$" | "," |
+      #                     ";" | ":" | "@" | "&" | "=" | "+" )
+      ret[:REG_NAME] = reg_name = "(?:[#{unreserved}$,;:@&=+]|#{escaped})+"
+      # authority     = server | reg_name
+      authority = "(?:#{server}|#{reg_name})"
+
+      # rel_segment   = 1*( unreserved | escaped |
+      #                     ";" | "@" | "&" | "=" | "+" | "$" | "," )
+      ret[:REL_SEGMENT] = rel_segment = "(?:[#{unreserved};@&=+$,]|#{escaped})+"
+
+      # scheme        = alpha *( alpha | digit | "+" | "-" | "." )
+      ret[:SCHEME] = scheme = "[#{PATTERN::ALPHA}][\\-+.#{PATTERN::ALPHA}\\d]*"
+
+      # abs_path      = "/"  path_segments
+      ret[:ABS_PATH] = abs_path = "/#{path_segments}"
+      # rel_path      = rel_segment [ abs_path ]
+      ret[:REL_PATH] = rel_path = "#{rel_segment}(?:#{abs_path})?"
+      # net_path      = "//" authority [ abs_path ]
+      ret[:NET_PATH] = net_path = "//#{authority}(?:#{abs_path})?"
+
+      # hier_part     = ( net_path | abs_path ) [ "?" query ]
+      ret[:HIER_PART] = hier_part = "(?:#{net_path}|#{abs_path})(?:\\?(?:#{query}))?"
+      # opaque_part   = uric_no_slash *uric
+      ret[:OPAQUE_PART] = opaque_part = "#{uric_no_slash}#{uric}*"
+
+      # absoluteURI   = scheme ":" ( hier_part | opaque_part )
+      ret[:ABS_URI] = abs_uri = "#{scheme}:(?:#{hier_part}|#{opaque_part})"
+      # relativeURI   = ( net_path | abs_path | rel_path ) [ "?" query ]
+      ret[:REL_URI] = rel_uri = "(?:#{net_path}|#{abs_path}|#{rel_path})(?:\\?#{query})?"
+
+      # URI-reference = [ absoluteURI | relativeURI ] [ "#" fragment ]
+      ret[:URI_REF] = "(?:#{abs_uri}|#{rel_uri})?(?:##{fragment})?"
+
+      ret[:X_ABS_URI] = "
+        (#{scheme}):                           (?# 1: scheme)
+        (?:
+           (#{opaque_part})                    (?# 2: opaque)
+        |
+           (?:(?:
+             //(?:
+                 (?:(?:(#{userinfo})@)?        (?# 3: userinfo)
+                   (?:(#{host})(?::(\\d*))?))? (?# 4: host, 5: port)
+               |
+                 (#{reg_name})                 (?# 6: registry)
+               )
+             |
+             (?!//))                           (?# XXX: '//' is the mark for hostport)
+             (#{abs_path})?                    (?# 7: path)
+           )(?:\\?(#{query}))?                 (?# 8: query)
+        )
+        (?:\\#(#{fragment}))?                  (?# 9: fragment)
+      "
+
+      ret[:X_REL_URI] = "
+        (?:
+          (?:
+            //
+            (?:
+              (?:(#{userinfo})@)?       (?# 1: userinfo)
+                (#{host})?(?::(\\d*))?  (?# 2: host, 3: port)
+            |
+              (#{reg_name})             (?# 4: registry)
+            )
+          )
+        |
+          (#{rel_segment})              (?# 5: rel_segment)
+        )?
+        (#{abs_path})?                  (?# 6: abs_path)
+        (?:\\?(#{query}))?              (?# 7: query)
+        (?:\\#(#{fragment}))?           (?# 8: fragment)
+      "
+
+      ret
+    end
+
+    # Constructs the default Hash of Regexp's
+    def initialize_regexp(pattern)
+      ret = {}
+
+      # for URI::split
+      ret[:ABS_URI] = Regexp.new('\A\s*' + pattern[:X_ABS_URI] + '\s*\z', Regexp::EXTENDED)
+      ret[:REL_URI] = Regexp.new('\A\s*' + pattern[:X_REL_URI] + '\s*\z', Regexp::EXTENDED)
+
+      # for URI::extract
+      ret[:URI_REF]     = Regexp.new(pattern[:URI_REF])
+      ret[:ABS_URI_REF] = Regexp.new(pattern[:X_ABS_URI], Regexp::EXTENDED)
+      ret[:REL_URI_REF] = Regexp.new(pattern[:X_REL_URI], Regexp::EXTENDED)
+
+      # for URI::escape/unescape
+      ret[:ESCAPED] = Regexp.new(pattern[:ESCAPED])
+      ret[:UNSAFE]  = Regexp.new("[^#{pattern[:UNRESERVED]}#{pattern[:RESERVED]}]")
+
+      # for Generic#initialize
+      ret[:SCHEME]   = Regexp.new("\\A#{pattern[:SCHEME]}\\z")
+      ret[:USERINFO] = Regexp.new("\\A#{pattern[:USERINFO]}\\z")
+      ret[:HOST]     = Regexp.new("\\A#{pattern[:HOST]}\\z")
+      ret[:PORT]     = Regexp.new("\\A#{pattern[:PORT]}\\z")
+      ret[:OPAQUE]   = Regexp.new("\\A#{pattern[:OPAQUE_PART]}\\z")
+      ret[:REGISTRY] = Regexp.new("\\A#{pattern[:REG_NAME]}\\z")
+      ret[:ABS_PATH] = Regexp.new("\\A#{pattern[:ABS_PATH]}\\z")
+      ret[:REL_PATH] = Regexp.new("\\A#{pattern[:REL_PATH]}\\z")
+      ret[:QUERY]    = Regexp.new("\\A#{pattern[:QUERY]}\\z")
+      ret[:FRAGMENT] = Regexp.new("\\A#{pattern[:FRAGMENT]}\\z")
+
+      ret
+    end
+
+    def convert_to_uri(uri)
+      if uri.is_a?(URI::Generic)
+        uri
+      elsif uri = String.try_convert(uri)
+        parse(uri)
+      else
+        raise ArgumentError,
+          "bad argument (expected URI object or URI string)"
+      end
+    end
+
+  end # class Parser
+end # module URI
diff --git a/lib/uri/rfc3986_parser.rb b/lib/uri/rfc3986_parser.rb
new file mode 100644
index 0000000000..779693c80c
--- /dev/null
+++ b/lib/uri/rfc3986_parser.rb
@@ -0,0 +1,101 @@
+module URI
+  class RFC3986_Parser # :nodoc:
+    # URI defined in RFC3986
+    # this regexp is modified not to host is not empty string
+    RFC3986_URI = /\A(?<URI>(?<scheme>[A-Za-z][+\-.0-9A-Za-z]*):(?<hier-part>\/\/(?<authority>(?:(?<userinfo>(?:%\h\h|[!$&-.0-;=A-Z_a-z~])*)@)?(?<host>(?<IP-literal>\[(?:(?<IPv6address>(?:\h{1,4}:){6}(?<ls32>\h{1,4}:\h{1,4}|(?<IPv4address>(?<dec-octet>[1-9]\d|1\d{2}|2[0-4]\d|25[0-5]|\d)\.\g<dec-octet>\.\g<dec-octet>\.\g<dec-octet>))|::(?:\h{1,4}:){5}\g<ls32>|\h{1,4}?::(?:\h{1,4}:){4}\g<ls32>|(?:(?:\h{1,4}:)?\h{1,4})?::(?:\h{1,4}:){3}\g<ls32>|(?:(?:\h{1,4}:){,2}\h{1,4})?::(?:\h{1,4}:){2}\g<ls32>|(?:(?:\h{1,4}:){,3}\h{1,4})?::\h{1,4}:\g<ls32>|(?:(?:\h{1,4}:){,4}\h{1,4})?::\g<ls32>|(?:(?:\h{1,4}:){,5}\h{1,4})?::\h{1,4}|(?:(?:\h{1,4}:){,6}\h{1,4})?::)|(?<IPvFuture>v\h+\.[!$&-.0-;=A-Z_a-z~]+))\])|\g<IPv4address>|(?<reg-name>(?:%\h\h|[!$&-.0-9;=A-Z_a-z~])+))?(?::(?<port>\d*))?)(?<path-abempty>(?:\/(?<segment>(?:%\h\h|[!$&-.0-;=@-Z_a-z~])*))*)|(?<path-absolute>\/(?:(?<segment-nz>(?:%\h\h|[!$&-.0-;=@-Z_a-z~])+)(?:\/\g<segment>)*)?)|(?<path-rootless>\g<segment-nz>(?:\/\g<segment>)*)|(?<path-empty>))(?:\?(?<query>(?:%\h\h|[!$&-.0-;=@-Z_a-z~]|[\/?])*))?(?:\#(?<fragment>(?:%\h\h|[!$&-.0-;=@-Z_a-z~]|[\/?])*))?)\z/
+    RFC3986_relative_ref = /\A(?<relative-ref>(?<relative-part>\/\/(?<authority>(?:(?<userinfo>(?:%\h\h|[!$&-.0-;=A-Z_a-z~])*)@)?(?<host>(?<IP-literal>\[(?<IPv6address>(?:\h{1,4}:){6}(?<ls32>\h{1,4}:\h{1,4}|(?<IPv4address>(?<dec-octet>[1-9]\d|1\d{2}|2[0-4]\d|25[0-5]|\d)\.\g<dec-octet>\.\g<dec-octet>\.\g<dec-octet>))|::(?:\h{1,4}:){5}\g<ls32>|\h{1,4}?::(?:\h{1,4}:){4}\g<ls32>|(?:(?:\h{1,4}:){,1}\h{1,4})?::(?:\h{1,4}:){3}\g<ls32>|(?:(?:\h{1,4}:){,2}\h{1,4})?::(?:\h{1,4}:){2}\g<ls32>|(?:(?:\h{1,4}:){,3}\h{1,4})?::\h{1,4}:\g<ls32>|(?:(?:\h{1,4}:){,4}\h{1,4})?::\g<ls32>|(?:(?:\h{1,4}:){,5}\h{1,4})?::\h{1,4}|(?:(?:\h{1,4}:){,6}\h{1,4})?::)|(?<IPvFuture>v\h+\.[!$&-.0-;=A-Z_a-z~]+)\])|\g<IPv4address>|(?<reg-name>(?:%\h\h|[!$&-.0-9;=A-Z_a-z~])+))?(?::(?<port>\d*))?)(?<path-abempty>(?:\/(?<segment>(?:%\h\h|[!$&-.0-;=@-Z_a-z~])*))*)|(?<path-absolute>\/(?:(?<segment-nz>(?:%\h\h|[!$&-.0-;=@-Z_a-z~])+)(?:\/\g<segment>)*)?)|(?<path-noscheme>(?<segment-nz-nc>(?:%\h\h|[!$&-.0-9;=@-Z_a-z~])+)(?:\/\g<segment>)*)|(?<path-empty>(?:%\h\h|[!$&-.0-;=@-Z_a-z~]){0}))(?:\?(?<query>(?:%\h\h|[!$&-.0-;=@-Z_a-z~]|[\/?])*))?(?:\#(?<fragment>(?:%\h\h|[!$&-.0-;=@-Z_a-z~]|[\/?])*))?)\z/
+
+    def split(uri) #:nodoc:
+      if m = RFC3986_URI.match(uri)
+        ary = []
+        ary << m["scheme"]
+        if m["path-rootless"] # opaque
+          ary << nil # userinfo
+          ary << nil # host
+          ary << nil # port
+          ary << nil # registry
+          ary << nil # path
+          ary << m["path-rootless"]
+          ary[-1] << '?' << m["query"] if m["query"]
+          ary << nil # query
+          ary << m["fragment"]
+        else # normal
+          ary << m["userinfo"]
+          ary << m["host"]
+          ary << m["port"]
+          ary << nil # registry
+          ary << (m["path-abempty"] || m["path-absolute"] || m["path-empty"])
+          ary << nil # opaque
+          ary << m["query"]
+          ary << m["fragment"]
+        end
+      elsif m = RFC3986_relative_ref.match(uri)
+        ary = [nil]
+        ary << m["userinfo"]
+        ary << m["host"]
+        ary << m["port"]
+        ary << nil # registry
+        ary << (m["path-abempty"] || m["path-absolute"] || m["path-noscheme"] || m["path-empty"])
+        ary << nil # opaque
+        ary << m["query"]
+        ary << m["fragment"]
+      else
+        raise InvalidURIError, "bad URI(is not URI?): #{uri}"
+      end
+    end
+
+    def parse(uri) # :nodoc:
+      scheme, userinfo, host, port,
+        registry, path, opaque, query, fragment = self.split(uri)
+
+      if scheme && URI.scheme_list.include?(scheme.upcase)
+        URI.scheme_list[scheme.upcase].new(scheme, userinfo, host, port,
+                                           registry, path, opaque, query,
+                                           fragment, self)
+      else
+        Generic.new(scheme, userinfo, host, port,
+                    registry, path, opaque, query,
+                    fragment, self)
+      end
+    end
+
+
+    def join(*uris) # :nodoc:
+      uris[0] = convert_to_uri(uris[0])
+      uris.inject :merge
+    end
+
+    @@to_s = Kernel.instance_method(:to_s)
+    def inspect
+      @@to_s.bind(self).call
+    end
+
+    def regexp
+      {
+        SCHEME: /\A[A-Za-z][A-Za-z0-9+\-.]*\z/,
+        USERINFO: /\A(?:%\h\h|[!$&-.0-;=A-Z_a-z~])*\z/,
+        HOST: /\A(?:(?<IP-literal>\[(?:(?<IPv6address>(?:\h{1,4}:){6}(?<ls32>\h{1,4}:\h{1,4}|(?<IPv4address>(?<dec-octet>[1-9]\d|1\d{2}|2[0-4]\d|25[0-5]|\d)\.\g<dec-octet>\.\g<dec-octet>\.\g<dec-octet>))|::(?:\h{1,4}:){5}\g<ls32>|\h{,4}::(?:\h{1,4}:){4}\g<ls32>|(?:(?:\h{1,4}:)?\h{1,4})?::(?:\h{1,4}:){3}\g<ls32>|(?:(?:\h{1,4}:){,2}\h{1,4})?::(?:\h{1,4}:){2}\g<ls32>|(?:(?:\h{1,4}:){,3}\h{1,4})?::\h{1,4}:\g<ls32>|(?:(?:\h{1,4}:){,4}\h{1,4})?::\g<ls32>|(?:(?:\h{1,4}:){,5}\h{1,4})?::\h{1,4}|(?:(?:\h{1,4}:){,6}\h{1,4})?::)|(?<IPvFuture>v\h+\.[!$&-.0-;=A-Z_a-z~]+))\])|\g<IPv4address>|(?<reg-name>(?:%\h\h|[!$&-.0-9;=A-Z_a-z~])*))\z/,
+        ABS_PATH: /\A\/(?:%\h\h|[!$&-.0-;=@-Z_a-z~])*(?:\/(?:%\h\h|[!$&-.0-;=@-Z_a-z~])*)*\z/,
+        REL_PATH: /\A(?:%\h\h|[!$&-.0-;=@-Z_a-z~])+(?:\/(?:%\h\h|[!$&-.0-;=@-Z_a-z~])*)*\z/,
+        QUERY: /\A(?:%\h\h|[!$&-.0-;=@-Z_a-z~]|[\/?])*\z/,
+        FRAGMENT: /\A(?:%\h\h|[!$&-.0-;=@-Z_a-z~]|[\/?])*\z/,
+        OPAQUE: nil,
+        PORT: nil,
+      }
+    end
+
+    private
+
+    def convert_to_uri(uri)
+      if uri.is_a?(URI::Generic)
+        uri
+      elsif uri = String.try_convert(uri)
+        parse(uri)
+      else
+        raise ArgumentError,
+          "bad argument (expected URI object or URI string)"
+      end
+    end
+
+  end # class Parser
+end # module URI