pan3793 commented on PR #48611: URL: https://github.com/apache/spark/pull/48611#issuecomment-2605203551
> If the following two dependencies are in the class path, there will be no corresponding warning logs, but we excluded it in this PR: https://github.com/apache/spark/pull/25481 > - `jakarta.activation:jakarta.activation-api` > - `jakarta.xml.bind:jakarta.xml.bind-api` A little bit of further investigation: Spark already pulls `jakarta.xml.bind-api` into the runtime classpath, in `mllib` module via `jpmml-model` 1.4 https://github.com/apache/spark/blob/98f276730d1dcfb2732a78439270b2578a777a15/dev/deps/spark-deps-hadoop-3-hive-2.3#L122 while this version actually contains `javax.xml.bind.` classes. `jpmml` migrated to `jakarta.xml.bind.` since 1.5, see https://github.com/jpmml/jpmml-model/issues/41. I think Spark eventually needs to follow the Jakarta EE specification to migrate to `jakarta` namespace, I did a quick search on Spark's codebase, seems it has no remaining direct reference for `javax.` that needs to migrate, but 3rd libs like `jpmml` might have the transitive reference, it might be a good time to evaluate upgrading `jpmml` to 1.5 or a newer version before Spark 4.0 ``` find . -type f \( -name "*.scala" -o -name "*.java" \) -exec grep -h "^import javax\." {} + | awk '{ sub(/;$/, ""); print }' | sort | uniq ``` Back to the issue specific to this PR, I think either the current approach or @LuciferYang's suggestion has no real runtime differences because Spark only uses a few set features of Jersey(which integrates with a lot of Jakarta EE APIs), but I lean toward @LuciferYang's suggestion because it simplifies both `pom.xml` and code -we don't need to carefully analysis the runtime code path and explicitly exclude/disable everything that we don't need, as long as it does not have side-effects. -- 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: reviews-unsubscr...@spark.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
