[SPARK-21491][GraphX] Enhance GraphX performance: breakOut instead of .toMap

[SereneAnt]

What changes were proposed in this pull request?

Traversable.toMap changed to 'collections.breakOut', that eliminates intermediate tuple collection creation, see Stack Overflow article.

How was this patch tested?

Unit tests run.
No performance tests performed yet.

Please review http://spark.apache.org/contributing.html before opening a pull request.

[srowen]

Ah OK that's what it is. Why is this different? the article you cited isn't comparing to toMap, and toMap doesn't make an intermediate collection.

[SereneAnt]

The principles are the same,
sources.zipWithIndex.map {...}' allocates a collection of tuples, .toMapthen iterates and converts them into the map (hehe). breakOut is the implementation of CanBuildFrom, the implicit parameter passed toTraversable.map` method. Once used, it allows populating newborn map directly, without intermediate collection of tuples.

[SereneAnt]

Optimization nerds have already used it in the spark code:

spark/sql/catalyst/src/main/scala/org/apache/spark/sql/types/StructType.scala

Line 517 in c400848

fields.map(s => (s.name, s))(breakOut)

[SparkQA]

Test build #3850 has finished for PR 18693 at commit 0ae9cc5.

• This patch passes all tests.
• This patch merges cleanly.
• This patch adds no public classes.

[srowen]

Merged to master