LDAP uses UTF-8 strings

Author: danabr

lib/net/ber.rb

@@ -295,6 +295,8 @@ class Net::BER::BerIdentifiedString < String
   attr_accessor :ber_identifier
   def initialize args
     super args
+    # LDAP uses UTF-8 encoded strings
+    force_encoding('UTF-8') if respond_to?(:encoding)
   end
 end
 

lib/net/ber/core_ext/string.rb

@@ -12,9 +12,21 @@ module Net::BER::Extensions::String
   # User code should call either #to_ber_application_string or
   # #to_ber_contextspecific.
   def to_ber(code = 0x04)
-    [code].pack('C') + length.to_ber_length_encoding + self
+    raw_string = raw_utf8_encoded
+    [code].pack('C') + raw_string.length.to_ber_length_encoding + raw_string
   end
 
+  def raw_utf8_encoded
+    if self.respond_to?(:encode)
+      # Strings should be UTF-8 encoded according to LDAP.
+      # However, the BER code is not necessarily valid UTF-8
+      self.encode('UTF-8').force_encoding('ASCII-8BIT')
+    else
+      self
+    end
+  end
+  private :raw_utf8_encoded
+
   ##
   # Creates an application-specific BER string encoded value with the
   # provided syntax code value.

spec/unit/ber/ber_spec.rb

@@ -75,6 +75,21 @@
       end 
     end
   end
+  if "Ruby 1.9".respond_to?(:encoding)
+    context "strings" do
+      it "should properly encode UTF-8 strings" do
+        "\u00e5".force_encoding("UTF-8").to_ber.should ==
+          "\x04\x02\xC3\xA5"
+      end
+      it "should properly encode strings encodable as UTF-8" do
+        "teststring".encode("US-ASCII").to_ber.should == "\x04\nteststring"
+      end
+      it "should fail on strings that can not be converted to UTF-8" do
+        error = Encoding::UndefinedConversionError
+        lambda {"\x81".to_ber }.should raise_exception(error)
+      end
+    end
+  end
 end
 
 describe "BER decoding of" do
@@ -91,4 +106,4 @@
           [1, [3, "Administrator", "ad_is_bogus"]]
     end 
   end
-end
+end