RFR 8223349: [lworld] Reflection support on static <init> factory methods for inline types

[Roger Riggs]

Hi John,

Good point, I should be using Set instead of List.

Roger


On 05/31/2019 03:46 PM, John Rose wrote:
> On May 31, 2019, at 12:13 PM, Roger Riggs <Roger.Riggs at oracle.com> wrote:
>> Hi,
>>
>> A minor update to make the test code checking constructors more generic and used for
>> Point, Line, and NonFlattenedValue.  It only checked the signatures, but does not invoke each constructor.
>>
>> webrev:
>> http://cr.openjdk.java.net/~rriggs/webrev-init-8223349-4/index.html
> Reviewed again, except:  The new test looks fragile with
> respect to declaration order, and (IIRC) method order
> is not preserved by reflection.  This could fail spuriously
> if c.getDeclaredConstructors reports a surprising new order.
> Quick fix:  Add a sorting phase to the stream that produces
> declaredSig.  Make sure your golden list is also sorted.
>
>