Fix: Fails to parser when given string includes multibyte chars.

Net::LDAP::Filter::FilterParser fails to parse when a given string
includes multibyte chars such as Japanese characters.
But this should support it according to RFC-2254.

This commit is to fix this failure.

Author: Tatsuya Sato

lib/net/ldap/filter.rb

@@ -755,7 +755,7 @@ def parse_filter_branch(scanner)
         scanner.scan(/\s*/)
         if op = scanner.scan(/<=|>=|!=|:=|=/)
           scanner.scan(/\s*/)
-          if value = scanner.scan(/(?:[-\w*.+@=,#\$%&!'\s\xC3\x80-\xCA\xAF]|\\[a-fA-F\d]{2})+/)
+          if value = scanner.scan(/(?:[-\w*.+@=,#\$%&!'\s\xC3\x80-\xCA\xAF]|[^\x00-\x7F]|\\[a-fA-F\d]{2})+/u)
             # 20100313 AZ: Assumes that "(uid=george*)" is the same as
             # "(uid=george* )". The standard doesn't specify, but I can find
             # no examples that suggest otherwise.