From 08631278ada7a6fd2bafb3ab0f0447b1f6d58790 Mon Sep 17 00:00:00 2001 From: duerst Date: Wed, 5 Mar 2008 08:45:51 +0000 Subject: Web Mar 5 17:43:43 2008 Martin Duerst * transcode.c (transcode_loop): Adjusted detection of invalid (ill-formed) UTF-8 sequences. Fixing potential security issue, see http://www.unicode.org/versions/Unicode5.1.0/#Notable_Changes. * test/ruby/test_transcode.rb: Added two tests for above fix. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@15692 b2dd03c8-39d4-4d8f-98ff-823fe69b080e --- test/ruby/test_transcode.rb | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) (limited to 'test/ruby/test_transcode.rb') diff --git a/test/ruby/test_transcode.rb b/test/ruby/test_transcode.rb index 9edf30882b..5a704fd364 100644 --- a/test/ruby/test_transcode.rb +++ b/test/ruby/test_transcode.rb @@ -242,6 +242,11 @@ class TestTranscode < Test::Unit::TestCase def test_invalid_ignore # arguments only - 'abc'.encode('utf-8', invalid: :ignore) + assert_nothing_raised { 'abc'.encode('utf-8', invalid: :ignore) } + # check handling of UTF-8 ill-formed subsequences + assert_equal("\x00\x41\x00\x3E\x00\x42".force_encoding('UTF-16BE'), + "\x41\xC2\x3E\x42".encode('UTF-16BE', 'UTF-8', invalid: :ignore)) + assert_equal("\x00\x41\x00\xF1\x00\x42".force_encoding('UTF-16BE'), + "\x41\xC2\xC3\xB1\x42".encode('UTF-16BE', 'UTF-8', invalid: :ignore)) end end -- cgit v1.2.3