Zouxxyy opened a new pull request, #10510: URL: https://github.com/apache/paimon/pull/10510
### Purpose A Hive UDTF registered with `CREATE FUNCTION` in a REST catalog works in a `SELECT` list, but `LATERAL VIEW` fails with `ROUTINE_NOT_FOUND`: Spark represents the call as `UnresolvedGenerator`, which Paimon did not rewrite. Resolve these generators through the existing Paimon V1 function registry after their arguments resolve, and retain Spark's generator type validation. Bind `database.function` to the current catalog, including when the database and a configured catalog share a name. Preserve built-in and temporary function precedence, and document Hive UDTF invocation. ### Tests - `PaimonV1FunctionTest`: 17 passed on Spark 3.5.8 (Scala 2.12, JDK 8). - `PaimonV1FunctionTest`: 17 passed on Spark 4.1.2 (Scala 2.13, JDK 17). - Five new cases cover persistent registration and cold lookup, qualified names, Paimon table input, empty/null input and `OUTER`, a fresh session, CTEs, dependent generators, name precedence, and rejection of scalar functions in `LATERAL VIEW`. - `spotless:apply`, `spotless:check`, and `git diff --check` passed. - Spark 4.1 reactor `package -DskipTests` passed without `fast-build`, with Checkstyle, Spotless, and Enforcer enabled. -- 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]
