[code-reflection] RFR: Committing a Concatenation transformation to StringBuilder [v2]

[Ian Graves]

On Fri, 14 Jun 2024 00:06:44 GMT, Ian Graves <igraves at openjdk.org> wrote:

>> test/jdk/java/lang/reflect/code/TestStringConcatTransform.java line 125:
>> 
>>> 123:             Assert.assertEquals(transformed_model_interpreted, ssa_interpreted);
>>> 124:             Assert.assertEquals(ssa_interpreted, ssa_transformed_interpreted);
>>> 125:             Assert.assertEquals(ssa_transformed_interpreted, jvm_interpreted);
>> 
>> Won't the asserts errors get wrapped in RuntimeException? suggest you pull 'em out of. e.g.,
>> 
>>   String jvm_interpreted;
>>   try {
>>     jvm_interpreted = method.invoke(null, args);
>>   } catch (...) { ... }
>> 
>>   Assert....
>
> You're right, they should probably come out just so it's clear they're not getting caught. I believe TestNG subclasses from `Error`.

...Ah misunderstood, yes they're getting wrapped and rethrown in the RuntimeException.

-------------

PR Review Comment: https://git.openjdk.org/babylon/pull/135#discussion_r1639065237