gharris1727 opened a new pull request, #15469: URL: https://github.com/apache/kafka/pull/15469
The Values class has checkstyle suppressions for NPathComplexity, MethodLength, CyclomaticComplexity, and JavaNCSS. Rather than suppressing these problems, we should refactor the oversize methods into smaller methods that satisfy the checkstyle heuristics. I added a benchmark to verify that this refactoring does not cause a performance regression. Once I had the benchmark, I found some really obvious optimizations: * parseAsTemporal was catching ParseExceptions, and I eliminated those by checking the ParsePosition. * number parsing was catching multiple ArithmeticExceptions when numbers had decimal parts, or were too large to fit in certain types. The new implementation does some equivalent rounding and cast-checking to avoid the exceptions. With these changes, the parse() profile shows a significant improvement, while all of the other methods are nearly the same or slightly faster than before. I'll include the detailed results from my machine in a follow-up comment. Finally I just did some simplification that makes the generic convertTo method call the specific convertTo(Boolean,Byte, etc) methods, rather than having every one of the specific methods call the single generic method. This eliminates the need for casting Object to the return types, and eliminates flaws that could come from that cast failing. I fixed one such flaw in #15399 that I found while doing this refactor. The SchemaMerger is a refactor of the fix in #15399 because the additional fields and control flow exceeded the checkstyle heuristics. ### Committer Checklist (excluded from commit message) - [ ] Verify design and implementation - [ ] Verify test coverage and CI build status - [ ] Verify documentation (including upgrade notes) -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: jira-unsubscr...@kafka.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
