Issue #235 has been updated by Martin Dürst.


I agree with Yui for the input encoding. The default for output should be UTF-8, however.
An exception should be raised if the actual encoding and the declared encoding on input
don't agree. An exception should not be raised if data from two or more input documents
in different encodings is combined.
----------------------------------------
http://redmine.ruby-lang.org/issues/show/235

----------------------------------------
http://redmine.ruby-lang.org