[SPARK-55710][BUILD] Use Google Mirror of Maven Central for SBT bootstrap

[yaooqinn]

What changes were proposed in this pull request?

Change the default SBT bootstrap repository from Maven Central (repo1.maven.org) to Google's Maven Central mirror (maven-central.storage-download.googleapis.com) for downloading the sbt-launch jar.

Why are the changes needed?

The rest of the SBT build already uses the Google mirror as the primary resolver, as project/SparkBuild.scala. The bootstrap script (build/sbt-launch-lib.bash) was the only place still defaulting to Maven Central. This makes it consistent and avoids flaky Maven Central connections during the initial sbt-launch jar download.

The DEFAULT_ARTIFACT_REPOSITORY environment variable can still be used to override this.

Does this PR introduce any user-facing change?

No.

How was this patch tested?

Manually tested by removing the cached sbt-launch-1.12.4.jar and verifying the bootstrap successfully downloads from the Google mirror with matching checksum.

Was this patch authored or co-authored using generative AI tooling?

Yes, GitHub Copilot CLI was used.

[sarutak]

@yaooqinn This change itself looks good to me. Leave some comments.

The rest of the SBT build already uses the Google mirror as the primary resolver, as configured in project/repositories and project/SparkBuild.scala.

Is project/repositories a typo for something?

I think it's OK to do in a separate PR but should we change the following code as well?
https://github.com/apache/spark/blob/master/common/utils/src/main/scala/org/apache/spark/util/MavenUtils.scala#L183

[yaooqinn]

I revised the PR desc, thank you @sarutak

[yaooqinn]

I think it's OK to do in a separate PR but should we change the following code as well?
https://github.com/apache/spark/blob/master/common/utils/src/main/scala/org/apache/spark/util/MavenUtils.scala#L183

This could be a user-facing change, I'm OK for it w/ a separate PR

[dongjoon-hyun]

+1, LGTM.

[dongjoon-hyun]

Thank you, @yaooqinn , @sarutak , @LuciferYang .
Merged to master for Apache Spark 4.2.0.