\ue863 is in the Unicode private use area. Therefore, conversion to GBK 
should not succeed. I get Encoding::UndefinedConversionError just for
    ruby -e '"\ue863".encode("gbk")'

I don't understand why the title of the bug is "when convert ... from 
GBK to UTF-8". The last method call of the test, .encode('utf-8'), is 
quite irrelevant. Justin, can you explain what you want?

Regards,   Martin.

On 2012/07/02 10:03, naruse (Yui NARUSE) wrote:
>
> Issue #6675 has been updated by naruse (Yui NARUSE).
>
> Status changed from Open to Feedback
>
> The test passes my environment:
> * ruby 1.9.3p194 (2012-04-20 revision 35410) [x86_64-darwin11.4.0]
> * ruby 1.9.3p203 (2012-05-04 revision 35536) [x86_64-freebsd9.0]
> * ruby 1.9.3p243 (2012-06-28 revision 36244) [x86_64-freebsd9.0]
> * ruby 2.0.0dev (2012-07-02 trunk 36265) [x86_64-freebsd9.0]
>
> Anyone can reproduce it?
> ----------------------------------------
> Bug #6675: Raise exception when convert encoding of a character from GBK to UTF
> https://bugs.ruby-lang.org/issues/6675#change-27700
>
> Author: mghomn (Justin Peal)
> Status: Feedback
> Priority: Normal
> Assignee:
> Category:
> Target version:
> ruby -v: ruby 1.9.3p194 (2012-04-20) [i386-mingw32]
>
>
> #!/usr/bin/env ruby -w
> # encoding: utf-8
>
> require 'test/unit'
>
> class Tests<  Test::Unit::TestCase
>    def test_gbk
>      assert_raise Encoding::UndefinedConversionError do
>        "\ue863".encode('gbk').encode('utf-8')
>      end
>    end
> end
>
>
>