xiangfu0 commented on PR #18756: URL: https://github.com/apache/pinot/pull/18756#issuecomment-5975586596
@Jackie-Jiang No additional benefit for ingestion JSON extraction from this cache: [ExpressionTransformer runs before the schema DataTypeTransformer](https://github.com/apache/pinot/blob/cf6b73f11c9eb5149c4d137c5d3afc789e6f8364/pinot-segment-local/src/main/java/org/apache/pinot/segment/local/recordtransformer/RecordTransformerUtils.java#L88-L99). Existing `JSONPATH*` functions can already consume parsed Map/List inputs (for example, an intermediate [jsonExtractObject(...)](https://github.com/apache/pinot/blob/cf6b73f11c9eb5149c4d137c5d3afc789e6f8364/pinot-common/src/main/java/org/apache/pinot/common/function/scalar/JsonFunctions.java#L587-L609)), independently of this PR. This PR avoids the subsequent parse for JSON-typed columns feeding [the built-in mutable JSON index](https://github.com/apache/pinot/blob/cf6b73f11c9eb5149c4d137c5d3afc789e6f8364/pinot-segment-local/src/main/java/org/apache/pinot/segment/local/indexsegment/mutable/MutableSegmentImpl.java#L956-L967). Sharing the cache across extraction expressions would need earlier population and evaluator support with source-value invalidation; I propose keeping that in a separate follow-up. At `cf6b73f`, [JDK 25 unit CI](https://github.com/apache/pinot/actions/runs/33958583035) passed `JsonUtilsTest`, `GenericRowTest`, `DataTypeTransformerTest`, `JsonIndexTest`, and `IndexingFailureTest` (117 tests, including sanitization invalidation and `testJsonIndexUsesParsedCache`). Could you re-review this scoped approach? -- 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: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
