LuciferYang commented on code in PR #42465:
URL: https://github.com/apache/spark/pull/42465#discussion_r1292329247
##########
connector/connect/server/pom.xml:
##########
@@ -140,6 +140,21 @@
</exclusion>
</exclusions>
</dependency>
+ <dependency>
+ <!--
+ We only need classes from org.apache.spark.sql.connect for testing.
+ spark-connect-client-jvm also contain the client versions of
org.apache.spark.sql classes
+ that are also present in spark-sql,
+ e.g. org.apache.spark.sql.SparkSession, org.apache.spark.sql.Dataset.
+ However, since Maven 2.0.9, Maven uses the order of dependencies in the
POM to build the
+ classpath, so this dependency is below spark-sql dependency is above,
the classes from the
+ server should take precedence.
+ -->
+ <groupId>org.apache.spark</groupId>
+ <artifactId>spark-connect-client-jvm_${scala.binary.version}</artifactId>
+ <version>${project.version}</version>
+ <scope>test</scope>
+ </dependency>
Review Comment:
No, that is not a good practice. While there may be some ways to temporarily
solve the classpath issue, it is not conducive to sustainable maintenance.
--
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]
