[
https://issues.apache.org/jira/browse/SPARK-48568?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17947535#comment-17947535
]
qiwei huang commented on SPARK-48568:
-------------------------------------
I tried using the fix from the Pull Request but couldn't repro the improvement.
For all test cases, the improvement is less than 10%, with some performing even
worse.
> Improve Performance of CodeFormatter with Java StringBuilder
> ------------------------------------------------------------
>
> Key: SPARK-48568
> URL: https://issues.apache.org/jira/browse/SPARK-48568
> Project: Spark
> Issue Type: Improvement
> Components: SQL
> Affects Versions: 4.0.0
> Reporter: Ahir Reddy
> Priority: Trivial
> Labels: pull-request-available
>
> - Update CodeFormatter to use Java's `StringBuilder` directly instead of
> Scala's
> - The Scala `StringBuilder` is a very thin API that wraps Java's.
> - All callsites that need to change just trivially copy what Scala was
> delegating to Java. For example `clear()` simply calls `setLength(0)` under
> the hood.
> - The reason for this change is that it substantially improves the
> performance of the CodeFormatter.
> - From some basic profiling, in a ~100s suite, code formatting took ~2.7
> seconds of CPU time. Post change it takes about 800ms.
> - My hypothesis is that Java StringBuilder is much more JIT friendly. Scala's
> StringBuilder has many layers as it implements a significant portion of the
> Scala mutable collection API. It's also likely the case that the JVM has
> special JIT handling for Java StringBuilder
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
