Ruby 1.9.1p0's URI.decode() bug fix

backport to fix Ruby 1.9.1p0 bug on [ruby-dev:38005].

Signed-off-by: Jeremy Kemper <jeremy@bitsweat.net>

2 files changed, 24 insertions, 0 deletions

diff --git a/activesupport/lib/active_support/core_ext/uri.rb b/activesupport/lib/active_support/core_ext/uri.rb
new file mode 100644
index 0000000000..26b050bf2b
--- /dev/null
+++ b/activesupport/lib/active_support/core_ext/uri.rb
@@ -0,0 +1,13 @@
+require 'uri'
+
+module ActiveSupport
+  if RUBY_VERSION == "1.9.1" && defined?(RUBY_PATCHLEVEL) && RUBY_PATCHLEVEL == 0
+    ::URI::Parser.class_eval do
+      def unescape(str, escaped = @regexp[:ESCAPED])
+        enc = (str.encoding == Encoding::US_ASCII) ? Encoding::UTF_8 : str.encoding
+        str.gsub(escaped) { [$&[1, 2].hex].pack('C') }.force_encoding(enc)
+      end
+    end
+  end
+end
+
diff --git a/activesupport/test/core_ext/uri_ext_test.rb b/activesupport/test/core_ext/uri_ext_test.rb
new file mode 100644
index 0000000000..f25bc97aa5
--- /dev/null
+++ b/activesupport/test/core_ext/uri_ext_test.rb
@@ -0,0 +1,11 @@
+require 'abstract_unit'
+
+class URITest < Test::Unit::TestCase
+  def test_uri_decode_handle_multibyte
+    str = "\xE6\x97\xA5\xE6\x9C\xAC\xE8\xAA\x9E" # Ni-ho-nn-go in UTF-8,  means Japanese.
+    str.force_encoding(Encoding::UTF_8) if(defined? Encoding::UTF_8)
+
+    assert_equal str, ::URI.unescape( ::URI.escape(str) )
+    assert_equal str, ::URI.decode( ::URI.escape(str) )
+  end
+end