addressable 2.4.0 → 2.8.0

data/lib/addressable/uri.rb

@@ -1,6 +1,8 @@
+# frozen_string_literal: true
+
 # encoding:utf-8
 #--
-# Copyright (C) 2006-2015 Bob Aman
+# Copyright (C) Bob Aman
 #
 #    Licensed under the Apache License, Version 2.0 (the "License");
 #    you may not use this file except in compliance with the License.
@@ -18,6 +20,7 @@
 
 require "addressable/version"
 require "addressable/idna"
+require "public_suffix"
 
 ##
 # Addressable is a library for processing links and URIs.
@@ -44,13 +47,22 @@ module Addressable
       UNRESERVED = ALPHA + DIGIT + "\\-\\.\\_\\~"
       PCHAR = UNRESERVED + SUB_DELIMS + "\\:\\@"
       SCHEME = ALPHA + DIGIT + "\\-\\+\\."
-      HOST = ALPHA + DIGIT + "\\-\\.\\[\\:\\]"
-      AUTHORITY = PCHAR
+      HOST = UNRESERVED + SUB_DELIMS + "\\[\\:\\]"
+      AUTHORITY = PCHAR + "\\[\\:\\]"
       PATH = PCHAR + "\\/"
       QUERY = PCHAR + "\\/\\?"
       FRAGMENT = PCHAR + "\\/\\?"
     end
 
+    module NormalizeCharacterClasses
+      HOST = /[^#{CharacterClasses::HOST}]/
+      UNRESERVED = /[^#{CharacterClasses::UNRESERVED}]/
+      PCHAR = /[^#{CharacterClasses::PCHAR}]/
+      SCHEME = /[^#{CharacterClasses::SCHEME}]/
+      FRAGMENT = /[^#{CharacterClasses::FRAGMENT}]/
+      QUERY = %r{[^a-zA-Z0-9\-\.\_\~\!\$\'\(\)\*\+\,\=\:\@\/\?%]|%(?!2B|2b)}
+    end
+
     SLASH = '/'
     EMPTY_STR = ''
 
@@ -70,7 +82,7 @@ module Addressable
       "wais" => 210,
       "ldap" => 389,
       "prospero" => 1525
-    }
+    }.freeze
 
     ##
     # Returns a URI object based on the parsed string.
@@ -121,9 +133,9 @@ module Addressable
           user = userinfo.strip[/^([^:]*):?/, 1]
           password = userinfo.strip[/:(.*)$/, 1]
         end
-        host = authority.gsub(
+        host = authority.sub(
           /^([^\[\]]*)@/, EMPTY_STR
-        ).gsub(
+        ).sub(
           /:([^:@\[\]]*?)$/, EMPTY_STR
         )
         port = authority[/:([^:@\[\]]*?)$/, 1]
@@ -176,33 +188,50 @@ module Addressable
         raise TypeError, "Can't convert #{uri.class} into String."
       end
       # Otherwise, convert to a String
-      uri = uri.to_str.dup
+      uri = uri.to_str.dup.strip
       hints = {
         :scheme => "http"
       }.merge(hints)
       case uri
-      when /^http:\/+/
-        uri.gsub!(/^http:\/+/, "http://")
-      when /^https:\/+/
-        uri.gsub!(/^https:\/+/, "https://")
-      when /^feed:\/+http:\/+/
-        uri.gsub!(/^feed:\/+http:\/+/, "feed:http://")
-      when /^feed:\/+/
-        uri.gsub!(/^feed:\/+/, "feed://")
-      when /^file:\/+/
-        uri.gsub!(/^file:\/+/, "file:///")
+      when /^http:\//i
+        uri.sub!(/^http:\/+/i, "http://")
+      when /^https:\//i
+        uri.sub!(/^https:\/+/i, "https://")
+      when /^feed:\/+http:\//i
+        uri.sub!(/^feed:\/+http:\/+/i, "feed:http://")
+      when /^feed:\//i
+        uri.sub!(/^feed:\/+/i, "feed://")
+      when %r[^file:/{4}]i
+        uri.sub!(%r[^file:/+]i, "file:////")
+      when %r[^file://localhost/]i
+        uri.sub!(%r[^file://localhost/+]i, "file:///")
+      when %r[^file:/+]i
+        uri.sub!(%r[^file:/+]i, "file:///")
       when /^\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}/
-        uri.gsub!(/^/, hints[:scheme] + "://")
+        uri.sub!(/^/, hints[:scheme] + "://")
+      when /\A\d+\..*:\d+\z/
+        uri = "#{hints[:scheme]}://#{uri}"
+      end
+      match = uri.match(URIREGEX)
+      fragments = match.captures
+      authority = fragments[3]
+      if authority && authority.length > 0
+        new_authority = authority.tr("\\", "/").gsub(" ", "%20")
+        # NOTE: We want offset 4, not 3!
+        offset = match.offset(4)
+        uri = uri.dup
+        uri[offset[0]...offset[1]] = new_authority
       end
       parsed = self.parse(uri)
       if parsed.scheme =~ /^[^\/?#\.]+\.[^\/?#]+$/
         parsed = self.parse(hints[:scheme] + "://" + uri)
       end
       if parsed.path.include?(".")
-        new_host = parsed.path[/^([^\/]+\.[^\/]*)/, 1]
-        if new_host
+        if parsed.path[/\b@\b/]
+          parsed.scheme = "mailto" unless parsed.scheme
+        elsif new_host = parsed.path[/^([^\/]+\.[^\/]*)/, 1]
           parsed.defer_validation do
-            new_path = parsed.path.gsub(
+            new_path = parsed.path.sub(
               Regexp.new("^" + Regexp.escape(new_host)), EMPTY_STR)
             parsed.host = new_host
             parsed.path = new_path
@@ -253,24 +282,24 @@ module Addressable
       # Otherwise, convert to a String
       path = path.to_str.strip
 
-      path.gsub!(/^file:\/?\/?/, EMPTY_STR) if path =~ /^file:\/?\/?/
+      path.sub!(/^file:\/?\/?/, EMPTY_STR) if path =~ /^file:\/?\/?/
       path = SLASH + path if path =~ /^([a-zA-Z])[\|:]/
       uri = self.parse(path)
 
       if uri.scheme == nil
         # Adjust windows-style uris
-        uri.path.gsub!(/^\/?([a-zA-Z])[\|:][\\\/]/) do
+        uri.path.sub!(/^\/?([a-zA-Z])[\|:][\\\/]/) do
           "/#{$1.downcase}:/"
         end
-        uri.path.gsub!(/\\/, SLASH)
+        uri.path.tr!("\\", SLASH)
         if File.exist?(uri.path) &&
             File.stat(uri.path).directory?
-          uri.path.gsub!(/\/$/, EMPTY_STR)
+          uri.path.chomp!(SLASH)
           uri.path = uri.path + '/'
         end
 
         # If the path is absolute, set the scheme and host.
-        if uri.path =~ /^\//
+        if uri.path.start_with?(SLASH)
           uri.scheme = "file"
           uri.host = EMPTY_STR
         end
@@ -307,6 +336,21 @@ module Addressable
       return result
     end
 
+    ##
+    # Tables used to optimize encoding operations in `self.encode_component`
+    # and `self.normalize_component`
+    SEQUENCE_ENCODING_TABLE = Hash.new do |hash, sequence|
+      hash[sequence] = sequence.unpack("C*").map do |c|
+        format("%02x", c)
+      end.join
+    end
+
+    SEQUENCE_UPCASED_PERCENT_ENCODING_TABLE = Hash.new do |hash, sequence|
+      hash[sequence] = sequence.unpack("C*").map do |c|
+        format("%%%02X", c)
+      end.join
+    end
+
     ##
     # Percent encodes a URI component.
     #
@@ -367,26 +411,26 @@ module Addressable
       if character_class.kind_of?(String)
         character_class = /[^#{character_class}]/
       end
-      if component.respond_to?(:force_encoding)
-        # We can't perform regexps on invalid UTF sequences, but
-        # here we need to, so switch to ASCII.
-        component = component.dup
-        component.force_encoding(Encoding::ASCII_8BIT)
-      end
+      # We can't perform regexps on invalid UTF sequences, but
+      # here we need to, so switch to ASCII.
+      component = component.dup
+      component.force_encoding(Encoding::ASCII_8BIT)
       # Avoiding gsub! because there are edge cases with frozen strings
       component = component.gsub(character_class) do |sequence|
-        (sequence.unpack('C*').map { |c| "%" + ("%02x" % c).upcase }).join
+        SEQUENCE_UPCASED_PERCENT_ENCODING_TABLE[sequence]
       end
       if upcase_encoded.length > 0
-        component = component.gsub(/%(#{upcase_encoded.chars.map do |char|
-          char.unpack('C*').map { |c| '%02x' % c }.join
-        end.join('|')})/i) { |s| s.upcase }
+        upcase_encoded_chars = upcase_encoded.chars.map do |char|
+          SEQUENCE_ENCODING_TABLE[char]
+        end
+        component = component.gsub(/%(#{upcase_encoded_chars.join('|')})/,
+                                   &:upcase)
       end
       return component
     end
 
     class << self
-      alias_method :encode_component, :encode_component
+      alias_method :escape_component, :encode_component
     end
 
     ##
@@ -427,14 +471,18 @@ module Addressable
       end
       uri = uri.dup
       # Seriously, only use UTF-8. I'm really not kidding!
-      uri.force_encoding("utf-8") if uri.respond_to?(:force_encoding)
-      leave_encoded.force_encoding("utf-8") if leave_encoded.respond_to?(:force_encoding)
+      uri.force_encoding("utf-8")
+
+      unless leave_encoded.empty?
+        leave_encoded = leave_encoded.dup.force_encoding("utf-8")
+      end
+
       result = uri.gsub(/%[0-9a-f]{2}/iu) do |sequence|
         c = sequence[1..3].to_i(16).chr
-        c.force_encoding("utf-8") if c.respond_to?(:force_encoding)
+        c.force_encoding("utf-8")
         leave_encoded.include?(c) ? sequence : c
       end
-      result.force_encoding("utf-8") if result.respond_to?(:force_encoding)
+      result.force_encoding("utf-8")
       if return_type == String
         return result
       elsif return_type == ::Addressable::URI
@@ -514,19 +562,21 @@ module Addressable
           character_class = "#{character_class}%" unless character_class.include?('%')
 
           "|%(?!#{leave_encoded.chars.map do |char|
-            seq = char.unpack('C*').map { |c| '%02x' % c }.join
+            seq = SEQUENCE_ENCODING_TABLE[char]
             [seq.upcase, seq.downcase]
           end.flatten.join('|')})"
         end
 
-        character_class = /[^#{character_class}]#{leave_re}/
-      end
-      if component.respond_to?(:force_encoding)
-        # We can't perform regexps on invalid UTF sequences, but
-        # here we need to, so switch to ASCII.
-        component = component.dup
-        component.force_encoding(Encoding::ASCII_8BIT)
+        character_class = if leave_re
+                            /[^#{character_class}]#{leave_re}/
+                          else
+                            /[^#{character_class}]/
+                          end
       end
+      # We can't perform regexps on invalid UTF sequences, but
+      # here we need to, so switch to ASCII.
+      component = component.dup
+      component.force_encoding(Encoding::ASCII_8BIT)
       unencoded = self.unencode_component(component, String, leave_encoded)
       begin
         encoded = self.encode_component(
@@ -537,9 +587,7 @@ module Addressable
       rescue ArgumentError
         encoded = self.encode_component(unencoded)
       end
-      if encoded.respond_to?(:force_encoding)
-        encoded.force_encoding(Encoding::UTF_8)
-      end
+      encoded.force_encoding(Encoding::UTF_8)
       return encoded
     end
 
@@ -843,14 +891,17 @@ module Addressable
       return nil unless self.scheme
       @normalized_scheme ||= begin
         if self.scheme =~ /^\s*ssh\+svn\s*$/i
-          "svn+ssh"
+          "svn+ssh".dup
         else
           Addressable::URI.normalize_component(
             self.scheme.strip.downcase,
-            Addressable::URI::CharacterClasses::SCHEME
+            Addressable::URI::NormalizeCharacterClasses::SCHEME
           )
         end
       end
+      # All normalized values should be UTF-8
+      @normalized_scheme.force_encoding(Encoding::UTF_8) if @normalized_scheme
+      @normalized_scheme
     end
 
     ##
@@ -864,7 +915,7 @@ module Addressable
         new_scheme = new_scheme.to_str
       end
       if new_scheme && new_scheme !~ /\A[a-z][a-z0-9\.\+\-]*\z/i
-        raise InvalidURIError, "Invalid scheme format: #{new_scheme}"
+        raise InvalidURIError, "Invalid scheme format: '#{new_scheme}'"
       end
       @scheme = new_scheme
       @scheme = nil if @scheme.to_s.strip.empty?
@@ -899,10 +950,13 @@ module Addressable
         else
           Addressable::URI.normalize_component(
             self.user.strip,
-            Addressable::URI::CharacterClasses::UNRESERVED
+            Addressable::URI::NormalizeCharacterClasses::UNRESERVED
           )
         end
       end
+      # All normalized values should be UTF-8
+      @normalized_user.force_encoding(Encoding::UTF_8) if @normalized_user
+      @normalized_user
     end
 
     ##
@@ -953,10 +1007,15 @@ module Addressable
         else
           Addressable::URI.normalize_component(
             self.password.strip,
-            Addressable::URI::CharacterClasses::UNRESERVED
+            Addressable::URI::NormalizeCharacterClasses::UNRESERVED
           )
         end
       end
+      # All normalized values should be UTF-8
+      if @normalized_password
+        @normalized_password.force_encoding(Encoding::UTF_8)
+      end
+      @normalized_password
     end
 
     ##
@@ -1017,11 +1076,16 @@ module Addressable
         if !current_user && !current_password
           nil
         elsif current_user && current_password
-          "#{current_user}:#{current_password}"
+          "#{current_user}:#{current_password}".dup
         elsif current_user && !current_password
-          "#{current_user}"
+          "#{current_user}".dup
         end
       end
+      # All normalized values should be UTF-8
+      if @normalized_userinfo
+        @normalized_userinfo.force_encoding(Encoding::UTF_8)
+      end
+      @normalized_userinfo
     end
 
     ##
@@ -1067,6 +1131,7 @@ module Addressable
     # @return [String] The host component, normalized.
     def normalized_host
       return nil unless self.host
+
       @normalized_host ||= begin
         if !self.host.strip.empty?
           result = ::Addressable::IDNA.to_ascii(
@@ -1078,12 +1143,18 @@ module Addressable
           end
           result = Addressable::URI.normalize_component(
             result,
-            CharacterClasses::HOST)
+            NormalizeCharacterClasses::HOST
+          )
           result
         else
-          EMPTY_STR
+          EMPTY_STR.dup
         end
       end
+      # All normalized values should be UTF-8
+      if @normalized_host && !@normalized_host.empty?
+        @normalized_host.force_encoding(Encoding::UTF_8)
+      end
+      @normalized_host
     end
 
     ##
@@ -1096,14 +1167,6 @@ module Addressable
       end
       @host = new_host ? new_host.to_str : nil
 
-      unreserved = CharacterClasses::UNRESERVED
-      sub_delims = CharacterClasses::SUB_DELIMS
-      if !@host.nil? && (@host =~ /[<>{}\/\?\#\@"[[:space:]]]/ ||
-          (@host[/^\[(.*)\]$/, 1] != nil && @host[/^\[(.*)\]$/, 1] !~
-          Regexp.new("^[#{unreserved}#{sub_delims}:]*$")))
-        raise InvalidURIError, "Invalid character in host: '#{@host.to_s}'"
-      end
-
       # Reset dependent values
       remove_instance_variable(:@authority) if defined?(@authority)
       remove_instance_variable(:@normalized_host) if defined?(@normalized_host)
@@ -1144,6 +1207,33 @@ module Addressable
       self.host = v
     end
 
+    ##
+    # Returns the top-level domain for this host.
+    #
+    # @example
+    #   Addressable::URI.parse("http://www.example.co.uk").tld # => "co.uk"
+    def tld
+      PublicSuffix.parse(self.host, ignore_private: true).tld
+    end
+
+    ##
+    # Sets the top-level domain for this URI.
+    #
+    # @param [String, #to_str] new_tld The new top-level domain.
+    def tld=(new_tld)
+      replaced_tld = host.sub(/#{tld}\z/, new_tld)
+      self.host = PublicSuffix::Domain.new(replaced_tld).to_s
+    end
+
+    ##
+    # Returns the public suffix domain for this host.
+    #
+    # @example
+    #   Addressable::URI.parse("http://www.example.co.uk").domain # => "example.co.uk"
+    def domain
+      PublicSuffix.domain(self.host, ignore_private: true)
+    end
+
     ##
     # The authority component for this URI.
     # Combines the user, password, host, and port components.
@@ -1151,7 +1241,7 @@ module Addressable
     # @return [String] The authority component.
     def authority
       self.host && @authority ||= begin
-        authority = ""
+        authority = String.new
         if self.userinfo != nil
           authority << "#{self.userinfo}@"
         end
@@ -1170,7 +1260,7 @@ module Addressable
     def normalized_authority
       return nil unless self.authority
       @normalized_authority ||= begin
-        authority = ""
+        authority = String.new
         if self.normalized_userinfo != nil
           authority << "#{self.normalized_userinfo}@"
         end
@@ -1180,6 +1270,11 @@ module Addressable
         end
         authority
       end
+      # All normalized values should be UTF-8
+      if @normalized_authority
+        @normalized_authority.force_encoding(Encoding::UTF_8)
+      end
+      @normalized_authority
     end
 
     ##
@@ -1197,9 +1292,9 @@ module Addressable
           new_user = new_userinfo.strip[/^([^:]*):?/, 1]
           new_password = new_userinfo.strip[/:(.*)$/, 1]
         end
-        new_host = new_authority.gsub(
+        new_host = new_authority.sub(
           /^([^\[\]]*)@/, EMPTY_STR
-        ).gsub(
+        ).sub(
           /:([^:@\[\]]*?)$/, EMPTY_STR
         )
         new_port =
@@ -1383,7 +1478,7 @@ module Addressable
     # @return [String] The components that identify a site.
     def site
       (self.scheme || self.authority) && @site ||= begin
-        site_string = ""
+        site_string = "".dup
         site_string << "#{self.scheme}:" if self.scheme != nil
         site_string << "//#{self.authority}" if self.authority != nil
         site_string
@@ -1402,7 +1497,7 @@ module Addressable
     def normalized_site
       return nil unless self.site
       @normalized_site ||= begin
-        site_string = ""
+        site_string = "".dup
         if self.normalized_scheme != nil
           site_string << "#{self.normalized_scheme}:"
         end
@@ -1411,6 +1506,9 @@ module Addressable
         end
         site_string
       end
+      # All normalized values should be UTF-8
+      @normalized_site.force_encoding(Encoding::UTF_8) if @normalized_site
+      @normalized_site
     end
 
     ##
@@ -1460,17 +1558,20 @@ module Addressable
         result = path.strip.split(SLASH, -1).map do |segment|
           Addressable::URI.normalize_component(
             segment,
-            Addressable::URI::CharacterClasses::PCHAR
+            Addressable::URI::NormalizeCharacterClasses::PCHAR
           )
         end.join(SLASH)
 
         result = URI.normalize_path(result)
         if result.empty? &&
             ["http", "https", "ftp", "tftp"].include?(self.normalized_scheme)
-          result = SLASH
+          result = SLASH.dup
         end
         result
       end
+      # All normalized values should be UTF-8
+      @normalized_path.force_encoding(Encoding::UTF_8) if @normalized_path
+      @normalized_path
     end
 
     ##
@@ -1489,6 +1590,9 @@ module Addressable
       # Reset dependent values
       remove_instance_variable(:@normalized_path) if defined?(@normalized_path)
       remove_composite_values
+
+      # Ensure we haven't created an invalid URI
+      validate()
     end
 
     ##
@@ -1497,7 +1601,7 @@ module Addressable
     # @return [String] The path's basename.
     def basename
       # Path cannot be nil
-      return File.basename(self.path).gsub(/;[^\/]*$/, EMPTY_STR)
+      return File.basename(self.path).sub(/;[^\/]*$/, EMPTY_STR)
     end
 
     ##
@@ -1529,13 +1633,21 @@ module Addressable
         modified_query_class = Addressable::URI::CharacterClasses::QUERY.dup
         # Make sure possible key-value pair delimiters are escaped.
         modified_query_class.sub!("\\&", "").sub!("\\;", "")
-        pairs = (self.query || "").split("&", -1)
+        pairs = (query || "").split("&", -1)
+        pairs.delete_if(&:empty?).uniq! if flags.include?(:compacted)
         pairs.sort! if flags.include?(:sorted)
         component = pairs.map do |pair|
-          Addressable::URI.normalize_component(pair, modified_query_class, "+")
+          Addressable::URI.normalize_component(
+            pair,
+            Addressable::URI::NormalizeCharacterClasses::QUERY,
+            "+"
+          )
         end.join("&")
         component == "" ? nil : component
       end
+      # All normalized values should be UTF-8
+      @normalized_query.force_encoding(Encoding::UTF_8) if @normalized_query
+      @normalized_query
     end
 
     ##
@@ -1588,11 +1700,13 @@ module Addressable
         # so it's best to make all changes in-place.
         pair[0] = URI.unencode_component(pair[0])
         if pair[1].respond_to?(:to_str)
+          value = pair[1].to_str
           # I loathe the fact that I have to do this. Stupid HTML 4.01.
           # Treating '+' as a space was just an unbelievably bad idea.
           # There was nothing wrong with '%20'!
           # If it ain't broke, don't fix it!
-          pair[1] = URI.unencode_component(pair[1].to_str.gsub(/\+/, " "))
+          value = value.tr("+", " ") if ["http", "https", nil].include?(scheme)
+          pair[1] = URI.unencode_component(value)
         end
         if return_type == Hash
           accu[pair[0]] = pair[1]
@@ -1644,7 +1758,7 @@ module Addressable
       end
 
       # new_query_values have form [['key1', 'value1'], ['key2', 'value2']]
-      buffer = ""
+      buffer = "".dup
       new_query_values.each do |key, value|
         encoded_key = URI.encode_component(
           key, CharacterClasses::UNRESERVED
@@ -1674,7 +1788,7 @@ module Addressable
     #
     # @return [String] The request URI required for an HTTP request.
     def request_uri
-      return nil if self.absolute? && self.scheme !~ /^https?$/
+      return nil if self.absolute? && self.scheme !~ /^https?$/i
       return (
         (!self.path.empty? ? self.path : SLASH) +
         (self.query ? "?#{self.query}" : EMPTY_STR)
@@ -1689,12 +1803,12 @@ module Addressable
       if !new_request_uri.respond_to?(:to_str)
         raise TypeError, "Can't convert #{new_request_uri.class} into String."
       end
-      if self.absolute? && self.scheme !~ /^https?$/
+      if self.absolute? && self.scheme !~ /^https?$/i
         raise InvalidURIError,
           "Cannot set an HTTP request URI for a non-HTTP URI."
       end
       new_request_uri = new_request_uri.to_str
-      path_component = new_request_uri[/^([^\?]*)\?(?:.*)$/, 1]
+      path_component = new_request_uri[/^([^\?]*)\??(?:.*)$/, 1]
       query_component = new_request_uri[/^(?:[^\?]*)\?(.*)$/, 1]
       path_component = path_component.to_s
       path_component = (!path_component.empty? ? path_component : SLASH)
@@ -1723,10 +1837,15 @@ module Addressable
       @normalized_fragment ||= begin
         component = Addressable::URI.normalize_component(
           self.fragment,
-          Addressable::URI::CharacterClasses::FRAGMENT
+          Addressable::URI::NormalizeCharacterClasses::FRAGMENT
         )
         component == "" ? nil : component
       end
+      # All normalized values should be UTF-8
+      if @normalized_fragment
+        @normalized_fragment.force_encoding(Encoding::UTF_8)
+      end
+      @normalized_fragment
     end
 
     ##
@@ -1844,8 +1963,8 @@ module Addressable
               # Section 5.2.3 of RFC 3986
               #
               # Removes the right-most path segment from the base path.
-              if base_path =~ /\//
-                base_path.gsub!(/\/[^\/]+$/, SLASH)
+              if base_path.include?(SLASH)
+                base_path.sub!(/\/[^\/]+$/, SLASH)
               else
                 base_path = EMPTY_STR
               end
@@ -2246,15 +2365,13 @@ module Addressable
           "Cannot assemble URI string with ambiguous path: '#{self.path}'"
       end
       @uri_string ||= begin
-        uri_string = ""
+        uri_string = String.new
         uri_string << "#{self.scheme}:" if self.scheme != nil
         uri_string << "//#{self.authority}" if self.authority != nil
         uri_string << self.path.to_s
         uri_string << "?#{self.query}" if self.query != nil
         uri_string << "##{self.fragment}" if self.fragment != nil
-        if uri_string.respond_to?(:force_encoding)
-          uri_string.force_encoding(Encoding::UTF_8)
-        end
+        uri_string.force_encoding(Encoding::UTF_8)
         uri_string
       end
     end
@@ -2296,10 +2413,10 @@ module Addressable
     #
     # @param [Proc] block
     #   A set of operations to perform on a given URI.
-    def defer_validation(&block)
-      raise LocalJumpError, "No block given." unless block
+    def defer_validation
+      raise LocalJumpError, "No block given." unless block_given?
       @validation_deferred = true
-      block.call()
+      yield
       @validation_deferred = false
       validate
       return nil
@@ -2371,6 +2488,19 @@ module Addressable
         raise InvalidURIError,
           "Cannot have a relative path with an authority set: '#{self.to_s}'"
       end
+      if self.path != nil && !self.path.empty? &&
+          self.path[0..1] == SLASH + SLASH && self.authority == nil
+        raise InvalidURIError,
+          "Cannot have a path with two leading slashes " +
+          "without an authority set: '#{self.to_s}'"
+      end
+      unreserved = CharacterClasses::UNRESERVED
+      sub_delims = CharacterClasses::SUB_DELIMS
+      if !self.host.nil? && (self.host =~ /[<>{}\/\\\?\#\@"[[:space:]]]/ ||
+          (self.host[/^\[(.*)\]$/, 1] != nil && self.host[/^\[(.*)\]$/, 1] !~
+          Regexp.new("^[#{unreserved}#{sub_delims}:]*$")))
+        raise InvalidURIError, "Invalid character in host: '#{self.host.to_s}'"
+      end
       return nil
     end
 
@@ -2384,7 +2514,9 @@ module Addressable
     def replace_self(uri)
       # Reset dependent values
       instance_variables.each do |var|
-        remove_instance_variable(var) if instance_variable_defined?(var)
+        if instance_variable_defined?(var) && var != :@validation_deferred
+          remove_instance_variable(var)
+        end
       end
 
       @scheme = uri.scheme