Willie Scholtz

@harawata I've removed all unnecessary formatting. I left the headers in tact (2023 -> 2024) please let me know in case you would like me to revert that as well!...

FWIW, I tested this change on a large part of our stack which includes 10's of thousands of mapped statements, both with the `experimentalConstructorCollectionMapping ` enabled, and disabled. However, given...

I wrote a rudimentary performance test for this (_not_ using JMH) The setup is 100 houses (main rows) with 200 rooms subdivided between them, with 50 furniture objects subdivided among...

I managed to improve the performance past the point of the property test baseline, Im doing this by only verifying we can create the result once, (per result set) as...

Hey @harawata, are you referring to the test related to `BoundBlogMapper` (`postConstructorForInit`) in `BindingTest`? I changed it because the original test itself was flawed in the sense that it was...

> Oops, sorry, I was diff-ing wrong commits. > Please disregard my last comment. No worries, I understand this PR is absolutely huge, and I'll do my best to make...

I forgot to actually cache the value of the `experimentalConstructorCollectionMapping` flag while building the `ResultMap` (similar to `hasNestedResultMaps` and `hasNestedQueries`), After doing this, performance has increased even more ;-) Additionally,...

@harawata @hazendaz Is there anything else I can/should do for this PR?

Thanks @harawata , would you mind leaving some comments so I could get it into the desired state, or is that something you prefer we do later?

@GeorgeSalu Indeed it does, I converted the test cases to use records in the attached patch, and it worked out of the box. [101_use_records.patch](https://github.com/mybatis/mybatis-3/files/15430888/101_use_records.patch)