Github user mgaido91 commented on a diff in the pull request:
https://github.com/apache/spark/pull/22152#discussion_r211626457
--- Diff: sql/core/src/test/scala/org/apache/spark/sql/DataFrameSuite.scala
---
@@ -2528,4 +2529,27 @@ class DataFrameSuite extends QueryTest with
SharedSQLContext {
checkAnswer(aggPlusFilter1, aggPlusFilter2.collect())
}
}
+
+ test("SPARK-25159: json schema inference should only trigger one job") {
+ withTempPath { path =>
+ // This test is to prove that the `JsonInferSchema` does not use
`RDD#toLocalIterator` which
+ // triggers one Spark job per RDD partition.
+ Seq(1 -> "a", 2 -> "b").toDF("i", "p")
+ // The data set has 2 partitions, so Spark will write at least 2
json files.
+ // Use a non-splittable compression (gzip), to make sure the json
scan RDD has at lease 2
--- End diff --
nit: `at least`
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
