RFR JDK-8028397: Undo the lenient mimi decoder suport change (JDK-8025003) in jdk8

Alan Bateman Alan.Bateman at oracle.com
Fri Nov 15 21:01:17 UTC 2013

On 15/11/2013 17:12, Xueming Shen wrote:
> I'm happy to put the "IAE with advanced positions" change back, the 
> webrev has been updated
> according.
> http://cr.openjdk.java.net/~sherman/8028397/webrev/

That looks much better.

One case that is still a bit awkward is where there are insufficient 
bits remaining in the source buffer, say when you are consuming from a 
channel and reading a chunk at a time. The decode(ByteBuffer, 
ByteBuffer) method really needs a boolean to indicate to the decoder if 
it capable of providing more bits. I guess it's too late to add that now.

As regards the javadoc updates then a small typo on L660 (=> "is a 
padding character"). On L849, I think I would drop "current".  Do you 
think decode(ByteBuffer) needs a sentence to make it clear that the 
input buffer's position may be advanced even if an exception is thrown?

On the test then it looks like there isn't any coverage to check the 
output buffer's position has been advanced. Otherwise looks fine.


More information about the core-libs-dev mailing list