Declaration annotations in type-annotation-only positions

[Jan Lahoda]

Werner,

Thanks for the bug report, was filled as:
https://bugs.openjdk.java.net/browse/JDK-8028699

Also thanks for the fix. I've prepared a webrev with the changes for 
convenience (the only intentional change compared to the changesets 
below was changing the bug number in the test):
http://cr.openjdk.java.net/~jlahoda/8028699/webrev.00/

I would like to ask about the "tree.def.mods != null" check. Is this 
check needed? Looking at other places, it seems that if tree.def != null 
then tree.def.mods cannot be null.

Thanks,
     Jan

On 11/21/2013 12:55 AM, Werner Dietl wrote:
> Dear type annotation developers,
>
> I have some changes in the type-annotations repository that should be
> propagated to TL.
> It's three changesets that should be merged into one:
>
> http://hg.openjdk.java.net/type-annotations/type-annotations/langtools/rev/098327afad8d
> http://hg.openjdk.java.net/type-annotations/type-annotations/langtools/rev/9cb6d18db886
> http://hg.openjdk.java.net/type-annotations/type-annotations/langtools/rev/2da59ead687f
>
> These changes prevent a NullPointerException in a simple test case and
> then refactor how the checks are implemented.
>
> All comments welcome!
> cu, WMD.
>