[HARMONY-6640] UTF8 decoder doesn't properly decode supplementary characters

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 5.0M14
• Fix Version/s: 5.0M16
• Component/s: Classlib
• Environment: Windows Vista
• Patch Info: Patch Available
• Assignee: Tim Ellison
• Reporter: Robert Muir
• Created: 02/Sep/10 13:26
• Updated: 15/Sep/10 20:14
• Resolved: 14/Sep/10 14:06

Description

When attempting to build Lucene, I discovered a problem with UTF8 decoding.
(this actually prevents our tests from even compiling without a workaround)

For any codepoint > 0xffff (4-byte utf-8 sequence), the decoder doesn't properly
split the decoded codepoint into surrogate pairs.