rhodes 3.2.0.beta.4 → 3.2.0.beta.5

data/lib/extensions/uri/uri/common.rb

@@ -1,8 +1,8 @@
 # = uri/common.rb
 #
 # Author:: Akira Yamada <akira@ruby-lang.org>
-# Revision:: $Id: common.rb 22760 2009-03-04 09:21:12Z yugui $
-# License:: 
+# Revision:: $Id: common.rb 31799 2011-05-29 22:49:36Z yugui $
+# License::
 #   You can redistribute it and/or modify it under the same term as Ruby.
 #
 
@@ -397,8 +397,8 @@ module URI
       ret = {}
 
       # for URI::split
-      ret[:ABS_URI] = Regexp.new('^' + pattern[:X_ABS_URI] + '$', Regexp::EXTENDED)
-      ret[:REL_URI] = Regexp.new('^' + pattern[:X_REL_URI] + '$', Regexp::EXTENDED)
+      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])
@@ -501,6 +501,7 @@ module URI
     #   # => "@%3F@%21"
     #
     def escape(*arg)
+      warn "#{caller(1)[0]}: warning: URI.escape is obsolete" if $VERBOSE
       DEFAULT_PARSER.escape(*arg)
     end
     alias encode escape
@@ -526,6 +527,7 @@ module URI
     #   # => "http://example.com/?a=\t\r"
     #
     def unescape(*arg)
+      warn "#{caller(1)[0]}: warning: URI.unescape is obsolete" if $VERBOSE
       DEFAULT_PARSER.unescape(*arg)
     end
     alias decode unescape
@@ -714,6 +716,134 @@ module URI
     DEFAULT_PARSER.make_regexp(schemes)
   end
 
+  TBLENCWWWCOMP_ = {} # :nodoc:
+  TBLDECWWWCOMP_ = {} # :nodoc:
+  HTML5ASCIIINCOMPAT = []#[Encoding::UTF_7, Encoding::UTF_16BE, Encoding::UTF_16LE,
+    #Encoding::UTF_32BE, Encoding::UTF_32LE] # :nodoc:
+
+  # Encode given +str+ to URL-encoded form data.
+  #
+  # This doesn't convert *, -, ., 0-9, A-Z, _, a-z,
+  # does convert SP to +, and convert others to %XX.
+  #
+  # This refers http://www.w3.org/TR/html5/forms.html#url-encoded-form-data
+  #
+  # See URI.decode_www_form_component, URI.encode_www_form
+  def self.encode_www_form_component(str)
+    if TBLENCWWWCOMP_.empty?
+      tbl = {}
+      256.times do |i|
+        tbl[i.chr] = '%%%02X' % i
+      end
+      tbl[' '] = '+'
+      begin
+        TBLENCWWWCOMP_.replace(tbl)
+        TBLENCWWWCOMP_.freeze
+      rescue
+      end
+    end
+    str = str.to_s
+    if HTML5ASCIIINCOMPAT.include?(str.encoding)
+      str = str.encode(Encoding::UTF_8)
+    else
+      str = str.dup
+    end
+    str.force_encoding(Encoding::ASCII_8BIT)
+    str.gsub!(/[^*\-.0-9A-Z_a-z]/, TBLENCWWWCOMP_)
+    str.force_encoding(Encoding::US_ASCII)
+  end
+
+  # Decode given +str+ of URL-encoded form data.
+  #
+  # This decods + to SP.
+  #
+  # See URI.encode_www_form_component, URI.decode_www_form
+  def self.decode_www_form_component(str, enc=Encoding::UTF_8)
+    if TBLDECWWWCOMP_.empty?
+      tbl = {}
+      256.times do |i|
+        h, l = i>>4, i&15
+        tbl['%%%X%X' % [h, l]] = i.chr
+        tbl['%%%x%X' % [h, l]] = i.chr
+        tbl['%%%X%x' % [h, l]] = i.chr
+        tbl['%%%x%x' % [h, l]] = i.chr
+      end
+      tbl['+'] = ' '
+      begin
+        TBLDECWWWCOMP_.replace(tbl)
+        TBLDECWWWCOMP_.freeze
+      rescue
+      end
+    end
+    raise ArgumentError, "invalid %-encoding (#{str})" unless /\A(?:%\h\h|[^%]+)*\z/ =~ str
+    str.gsub(/\+|%\h\h/, TBLDECWWWCOMP_).force_encoding(enc)
+  end
+
+  # Generate URL-encoded form data from given +enum+.
+  #
+  # This generates application/x-www-form-urlencoded data defined in HTML5
+  # from given an Enumerable object.
+  #
+  # This internally uses URI.encode_www_form_component(str).
+  #
+  # This doesn't convert encodings of give items, so convert them before call
+  # this method if you want to send data as other than original encoding or
+  # mixed encoding data. (strings which is encoded in HTML5 ASCII incompatible
+  # encoding is converted to UTF-8)
+  #
+  # This doesn't treat files. When you send a file, use multipart/form-data.
+  #
+  # This refers http://www.w3.org/TR/html5/forms.html#url-encoded-form-data
+  #
+  # See URI.encode_www_form_component, URI.decode_www_form
+  def self.encode_www_form(enum)
+    str = nil
+    enum.each do |k,v|
+      if str
+        str << '&'
+      else
+        str = nil.to_s
+      end
+      str << encode_www_form_component(k)
+      str << '='
+      str << encode_www_form_component(v)
+    end
+    str
+  end
+
+  WFKV_ = '(?:%\h\h|[^%#=;&])' # :nodoc:
+
+  # Decode URL-encoded form data from given +str+.
+  #
+  # This decodes application/x-www-form-urlencoded data
+  # and returns array of key-value array.
+  # This internally uses URI.decode_www_form_component.
+  #
+  # _charset_ hack is not supported now because the mapping from given charset
+  # to Ruby's encoding is not clear yet.
+  # see also http://www.w3.org/TR/html5/syntax.html#character-encodings-0
+  #
+  # This refers http://www.w3.org/TR/html5/forms.html#url-encoded-form-data
+  #
+  # ary = URI.decode_www_form("a=1&a=2&b=3")
+  # p ary                  #=> [['a', '1'], ['a', '2'], ['b', '3']]
+  # p ary.assoc('a').last  #=> '1'
+  # p ary.assoc('b').last  #=> '3'
+  # p ary.rassoc('a').last #=> '2'
+  # p Hash[ary]            # => {"a"=>"2", "b"=>"3"}
+  #
+  # See URI.decode_www_form_component, URI.encode_www_form
+  def self.decode_www_form(str, enc=Encoding::UTF_8)
+    return [] if str.empty?
+    unless /\A#{WFKV_}*=#{WFKV_}*(?:[;&]#{WFKV_}*=#{WFKV_}*)*\z/o =~ str
+      raise ArgumentError, "invalid data of application/x-www-form-urlencoded (#{str})"
+    end
+    ary = []
+    $&.scan(/([^=;&]+)=([^;&]*)/) do
+      ary << [decode_www_form_component($1, enc), decode_www_form_component($2, enc)]
+    end
+    ary
+  end
 end
 
 module Kernel

data/lib/extensions/uri/uri/ftp.rb

@@ -3,7 +3,7 @@
 #
 # Author:: Akira Yamada <akira@ruby-lang.org>
 # License:: You can redistribute it and/or modify it under the same term as Ruby.
-# Revision:: $Id: ftp.rb 11708 2007-02-12 23:01:19Z shyouhei $
+# Revision:: $Id: ftp.rb 27350 2010-04-15 13:41:19Z mame $
 #
 
 require 'uri/generic'
@@ -13,6 +13,11 @@ module URI
   #
   # FTP URI syntax is defined by RFC1738 section 3.2.
   #
+  # This class will be redesigned because of difference of implementations;
+  # the structure of its path. draft-hoffman-ftp-uri-04 is a draft but it
+  # is a good summary about the de facto spec.
+  # http://tools.ietf.org/html/draft-hoffman-ftp-uri-04
+  #
   class FTP < Generic
     DEFAULT_PORT = 21
 
@@ -113,13 +118,14 @@ module URI
     # +opaque+, +query+ and +fragment+, in that order.
     #
     def initialize(*arg)
+      arg[5] = arg[5].sub(/^\//,'').sub(/^%2F/,'/')
       super(*arg)
       @typecode = nil
       tmp = @path.index(TYPECODE_PREFIX)
       if tmp
         typecode = @path[tmp + TYPECODE_PREFIX.size..-1]
-        self.set_path(@path[0..tmp - 1])
-        
+        @path = @path[0..tmp - 1]
+
         if arg[-1]
           self.typecode = typecode
         else
@@ -180,6 +186,11 @@ module URI
       return @path.sub(/^\//,'').sub(/^%2F/,'/')
     end
 
+    def set_path(v)
+      super("/" + v.sub(/^\//, "%2F"))
+    end
+    protected :set_path
+
     def to_s
       save_path = nil
       if @typecode

data/lib/extensions/uri/uri/generic.rb

@@ -3,7 +3,7 @@
 #
 # Author:: Akira Yamada <akira@ruby-lang.org>
 # License:: You can redistribute it and/or modify it under the same term as Ruby.
-# Revision:: $Id: generic.rb 20258 2008-11-18 16:46:16Z yugui $
+# Revision:: $Id: generic.rb 31804 2011-05-29 22:50:20Z yugui $
 #
 
 require 'uri/common'
@@ -73,7 +73,7 @@ module URI
         if args.kind_of?(Array)
           return self.build(args.collect{|x| 
             if x
-              @parser.escape(x)
+              parser.escape(x)
             else
               x
             end
@@ -82,7 +82,7 @@ module URI
           tmp = {}
           args.each do |key, value|
             tmp[key] = if value
-                @parser.escape(value)
+                parser.escape(value)
               else
                 value
               end
@@ -121,7 +121,7 @@ module URI
         "expected Array of or Hash of components of #{self.class} (#{self.class.component.join(', ')})"
       end
 
-      tmp << DEFAULT_PARSER
+      tmp << nil
       tmp << true
       return self.new(*tmp)
     end
@@ -172,7 +172,7 @@ module URI
       @opaque = nil
       @registry = nil
       @fragment = nil
-      @parser = parser
+      @parser = parser == DEFAULT_PARSER ? nil : parser
 
       if arg_check
         self.scheme = scheme
@@ -212,7 +212,14 @@ module URI
     attr_reader :query
     attr_reader :opaque
     attr_reader :fragment
-    attr_reader :parser
+
+    def parser
+      if !defined?(@parser) || !@parser
+        DEFAULT_PARSER
+      else
+        @parser || DEFAULT_PARSER
+      end
+    end
 
     # replace self by other URI object
     def replace!(oth)
@@ -231,7 +238,7 @@ module URI
     end
 
     def check_scheme(v)
-      if v && @parser.regexp[:SCHEME] !~ v
+      if v && parser.regexp[:SCHEME] !~ v
         raise InvalidComponentError,
           "bad component(expected scheme component): #{v}"
       end
@@ -270,7 +277,7 @@ module URI
 
       return v unless v
 
-      if @parser.regexp[:USERINFO] !~ v
+      if parser.regexp[:USERINFO] !~ v
         raise InvalidComponentError,
           "bad component(expected userinfo component or user component): #{v}"
       end
@@ -291,7 +298,7 @@ module URI
           "password component depends user component"
       end
 
-      if @parser.regexp[:USERINFO] !~ v
+      if parser.regexp[:USERINFO] !~ v
         raise InvalidComponentError,
           "bad component(expected user component): #{v}"
       end
@@ -356,7 +363,7 @@ module URI
     private :split_userinfo
 
     def escape_userpass(v)
-      v = @parser.escape(v, /[@:\/]/o) # RFC 1738 section 3.1 #/
+      v = parser.escape(v, /[@:\/]/o) # RFC 1738 section 3.1 #/
     end
     private :escape_userpass
 
@@ -384,7 +391,7 @@ module URI
       if @registry || @opaque
         raise InvalidURIError, 
           "can not set host with registry or opaque"
-      elsif @parser.regexp[:HOST] !~ v
+      elsif parser.regexp[:HOST] !~ v
         raise InvalidComponentError,
           "bad component(expected host component): #{v}"
       end
@@ -410,7 +417,7 @@ module URI
       if @registry || @opaque
         raise InvalidURIError, 
           "can not set port with registry or opaque"
-      elsif !v.kind_of?(Fixnum) && @parser.regexp[:PORT] !~ v
+      elsif !v.kind_of?(Fixnum) && parser.regexp[:PORT] !~ v
         raise InvalidComponentError,
           "bad component(expected port component): #{v}"
       end
@@ -446,7 +453,7 @@ module URI
       if @host || @port || @user # userinfo = @user + ':' + @password
         raise InvalidURIError, 
           "can not set registry with host, port, or userinfo"
-      elsif v && @parser.regexp[:REGISTRY] !~ v
+      elsif v && parser.regexp[:REGISTRY] !~ v
         raise InvalidComponentError,
           "bad component(expected registry component): #{v}"
       end
@@ -475,14 +482,16 @@ module URI
           "path conflicts with opaque"
       end
 
-      if @scheme
-        if v && v != '' && @parser.regexp[:ABS_PATH] !~ v
-          raise InvalidComponentError, 
+      # If scheme is ftp, path may be relative.
+      # See RFC 1738 section 3.2.2, and RFC 2396.
+      if @scheme && @scheme != "ftp"
+        if v && v != '' && parser.regexp[:ABS_PATH] !~ v
+          raise InvalidComponentError,
             "bad component(expected absolute path component): #{v}"
         end
       else
-        if v && v != '' && @parser.regexp[:ABS_PATH] !~ v && @parser.regexp[:REL_PATH] !~ v
-          raise InvalidComponentError, 
+        if v && v != '' && parser.regexp[:ABS_PATH] !~ v && parser.regexp[:REL_PATH] !~ v
+          raise InvalidComponentError,
             "bad component(expected relative path component): #{v}"
         end
       end
@@ -513,8 +522,8 @@ module URI
           "query conflicts with opaque"
       end
 
-      if v && v != '' && @parser.regexp[:QUERY] !~ v
-          raise InvalidComponentError, 
+      if v && v != '' && parser.regexp[:QUERY] !~ v
+          raise InvalidComponentError,
             "bad component(expected query component): #{v}"
       end
 
@@ -542,7 +551,7 @@ module URI
       if @host || @port || @user || @path  # userinfo = @user + ':' + @password
         raise InvalidURIError, 
           "can not set opaque with host, port, userinfo or path"
-      elsif v && @parser.regexp[:OPAQUE] !~ v
+      elsif v && parser.regexp[:OPAQUE] !~ v
         raise InvalidComponentError,
           "bad component(expected opaque component): #{v}"
       end
@@ -565,8 +574,8 @@ module URI
     def check_fragment(v)
       return v unless v
 
-      if v && v != '' && @parser.regexp[:FRAGMENT] !~ v
-        raise InvalidComponentError, 
+      if v && v != '' && parser.regexp[:FRAGMENT] !~ v
+        raise InvalidComponentError,
           "bad component(expected fragment component): #{v}"
       end
 
@@ -777,7 +786,7 @@ module URI
       case oth
       when Generic
       when String
-        oth = @parser.parse(oth)
+        oth = parser.parse(oth)
       else
         raise ArgumentError,
           "bad argument(expected URI object or URI string)"
@@ -804,30 +813,26 @@ module URI
     private :merge0
 
     def route_from_path(src, dst)
-      # RFC2396, Section 4.2
-      return '' if src == dst
-
-      src_path = split_path(src)
-      dst_path = split_path(dst)
-
-      # hmm... dst has abnormal absolute path, 
-      # like "/./", "/../", "/x/../", ...
-      if dst_path.include?('..') ||
-          dst_path.include?('.')
+      case dst
+      when src
+        # RFC2396, Section 4.2
+        return ''
+      when %r{(?:\A|/)\.\.?(?:/|\z)}
+        # dst has abnormal absolute path,
+        # like "/./", "/../", "/x/../", ...
         return dst.dup
       end
 
-      src_path.pop
+      src_path = src.scan(%r{(?:\A|[^/]+)/})
+      dst_path = dst.scan(%r{(?:\A|[^/]+)/?})
 
       # discard same parts
-      while dst_path.first == src_path.first
-        break if dst_path.empty?
-
+      while !dst_path.empty? && dst_path.first == src_path.first
         src_path.shift
         dst_path.shift
       end
 
-      tmp = dst_path.join('/')
+      tmp = dst_path.join
 
       # calculate
       if src_path.empty?
@@ -848,7 +853,7 @@ module URI
       case oth
       when Generic
       when String
-        oth = @parser.parse(oth)
+        oth = parser.parse(oth)
       else
         raise ArgumentError,
           "bad argument(expected URI object or URI string)"
@@ -869,7 +874,7 @@ module URI
       rel = URI::Generic.new(nil, # it is relative URI
                              self.userinfo, self.host, self.port, 
                              self.registry, self.path, self.opaque,
-                             self.query, self.fragment, @parser)
+                             self.query, self.fragment, parser)
 
       if rel.userinfo != oth.userinfo ||
           rel.host.to_s.downcase != oth.host.to_s.downcase ||
@@ -960,7 +965,7 @@ module URI
       case oth
       when Generic
       when String
-        oth = @parser.parse(oth)
+        oth = parser.parse(oth)
       else
         raise ArgumentError,
           "bad argument(expected URI object or URI string)"
@@ -985,6 +990,9 @@ module URI
       if path && path == ''
         set_path('/')
       end
+      if scheme && scheme != scheme.downcase
+        set_scheme(self.scheme.downcase)
+      end
       if host && host != host.downcase
         set_host(self.host.downcase)
       end        
@@ -1059,7 +1067,8 @@ module URI
     end
 
     def eql?(oth)
-      @parser == oth.parser &&
+      self.class == oth.class &&
+      parser == oth.parser &&
       self.component_ary.eql?(oth.component_ary)
     end
 
@@ -1117,7 +1126,7 @@ module URI
     def coerce(oth)
       case oth
       when String
-        oth = @parser.parse(oth)
+        oth = parser.parse(oth)
       else
         super
       end