diff options
author | Reynold Xin <rxin@databricks.com> | 2016-07-07 18:09:18 +0800 |
---|---|---|
committer | Cheng Lian <lian@databricks.com> | 2016-07-07 18:09:18 +0800 |
commit | 986b2514013ed9ebab526f2cf3dc714cc9e480bf (patch) | |
tree | 2c1fbd3515c18a50702bb67399339163ecd42196 /sql/core/src/test/scala/org | |
parent | ab05db0b48f395543cd7d91e2ad9dd760516868b (diff) | |
download | spark-986b2514013ed9ebab526f2cf3dc714cc9e480bf.tar.gz spark-986b2514013ed9ebab526f2cf3dc714cc9e480bf.tar.bz2 spark-986b2514013ed9ebab526f2cf3dc714cc9e480bf.zip |
[SPARK-16400][SQL] Remove InSet filter pushdown from Parquet
## What changes were proposed in this pull request?
This patch removes InSet filter pushdown from Parquet data source, since row-based pushdown is not beneficial to Spark and brings extra complexity to the code base.
## How was this patch tested?
N/A
Author: Reynold Xin <rxin@databricks.com>
Closes #14076 from rxin/SPARK-16400.
Diffstat (limited to 'sql/core/src/test/scala/org')
-rw-r--r-- | sql/core/src/test/scala/org/apache/spark/sql/execution/datasources/parquet/ParquetFilterSuite.scala | 30 |
1 files changed, 0 insertions, 30 deletions
diff --git a/sql/core/src/test/scala/org/apache/spark/sql/execution/datasources/parquet/ParquetFilterSuite.scala b/sql/core/src/test/scala/org/apache/spark/sql/execution/datasources/parquet/ParquetFilterSuite.scala index 84fdcfea3c..f59d474d00 100644 --- a/sql/core/src/test/scala/org/apache/spark/sql/execution/datasources/parquet/ParquetFilterSuite.scala +++ b/sql/core/src/test/scala/org/apache/spark/sql/execution/datasources/parquet/ParquetFilterSuite.scala @@ -514,36 +514,6 @@ class ParquetFilterSuite extends QueryTest with ParquetTest with SharedSQLContex } } - test("SPARK-11164: test the parquet filter in") { - import testImplicits._ - withSQLConf(SQLConf.PARQUET_FILTER_PUSHDOWN_ENABLED.key -> "true") { - withSQLConf(SQLConf.PARQUET_VECTORIZED_READER_ENABLED.key -> "false") { - withTempPath { dir => - val path = s"${dir.getCanonicalPath}/table1" - (1 to 5).map(i => (i.toFloat, i%3)).toDF("a", "b").write.parquet(path) - - // When a filter is pushed to Parquet, Parquet can apply it to every row. - // So, we can check the number of rows returned from the Parquet - // to make sure our filter pushdown work. - val df = spark.read.parquet(path).where("b in (0,2)") - assert(stripSparkFilter(df).count == 3) - - val df1 = spark.read.parquet(path).where("not (b in (1))") - assert(stripSparkFilter(df1).count == 3) - - val df2 = spark.read.parquet(path).where("not (b in (1,3) or a <= 2)") - assert(stripSparkFilter(df2).count == 2) - - val df3 = spark.read.parquet(path).where("not (b in (1,3) and a <= 2)") - assert(stripSparkFilter(df3).count == 4) - - val df4 = spark.read.parquet(path).where("not (a <= 2)") - assert(stripSparkFilter(df4).count == 3) - } - } - } - } - test("SPARK-16371 Do not push down filters when inner name and outer name are the same") { withParquetDataFrame((1 to 4).map(i => Tuple1(Tuple1(i)))) { implicit df => // Here the schema becomes as below: |