[BUG] Spark UT framework: "select a single complex field" causes java.lang.IndexOutOfBoundsException

[Feng-Jiang28]

Description:

contacts parquet is defined as following and has saved here: contacts.zip

+---+--------------------+---------------+----+--------------------+----------------------------+-------------------------------+----------------------------+---+
|id |name                |address        |pets|friends             |relatives                   |employer                       |relations                   |p  |
+---+--------------------+---------------+----+--------------------+----------------------------+-------------------------------+----------------------------+---+
|0  |{Jane, X., Doe}     |123 Main Street|1   |[{Susan, Z., Smith}]|{brother -> {John, Y., Doe}}|{0, {abc, 123 Business Street}}|{{John, Y., Doe} -> brother}|1  |
|1  |{John, Y., Doe}     |321 Wall Street|3   |[]                  |{sister -> {Jane, X., Doe}} |{1, null}                      |{{Jane, X., Doe} -> sister} |1  |
|2  |{Janet, null, Jones}|567 Maple Drive|null|null                |null                        |null                           |null                        |2  |
|3  |{Jim, null, Jones}  |6242 Ash Street|null|null                |null                        |null                           |null                        |2  |
+---+--------------------+---------------+----+--------------------+----------------------------+-------------------------------+----------------------------+---+

Code to reproduce:

val dataSourceName = "parquet"
val path = "/home/fejiang/Desktop"

val schema = ("`id` INT,`name` STRUCT<`first`: STRING, `middle`: STRING, `last`: STRING>, " +
  "`address` STRING,`pets` INT,`friends` ARRAY<STRUCT<`first`: STRING, `middle`: STRING, " +
  "`last`: STRING>>,`relatives` MAP<STRING, STRUCT<`first`: STRING, `middle`: STRING, " +
  "`last`: STRING>>,`employer` STRUCT<`id`: INT, `company`: STRUCT<`name`: STRING, " +
  "`address`: STRING>>,`relations` MAP<STRUCT<`first`: STRING, `middle`: STRING, " +
  "`last`: STRING>,STRING>,`p` INT")
spark.read.format(dataSourceName).schema(schema).load(path + "/contacts").createOrReplaceTempView("contacts")
val query = spark.sql("select name.middle from contacts")
query.show()

Spark:

scala> val dataSourceName = "parquet"
dataSourceName: String = parquet

scala> val path = "/home/fejiang/Desktop"
path: String = /home/fejiang/Desktop

scala> 

scala> val schema = ("`id` INT,`name` STRUCT<`first`: STRING, `middle`: STRING, `last`: STRING>, " +
     |   "`address` STRING,`pets` INT,`friends` ARRAY<STRUCT<`first`: STRING, `middle`: STRING, " +
     |   "`last`: STRING>>,`relatives` MAP<STRING, STRUCT<`first`: STRING, `middle`: STRING, " +
     |   "`last`: STRING>>,`employer` STRUCT<`id`: INT, `company`: STRUCT<`name`: STRING, " +
     |   "`address`: STRING>>,`relations` MAP<STRUCT<`first`: STRING, `middle`: STRING, " +
     |   "`last`: STRING>,STRING>,`p` INT")
schema: String = `id` INT,`name` STRUCT<`first`: STRING, `middle`: STRING, `last`: STRING>, `address` STRING,`pets` INT,`friends` ARRAY<STRUCT<`first`: STRING, `middle`: STRING, `last`: STRING>>,`relatives` MAP<STRING, STRUCT<`first`: STRING, `middle`: STRING, `last`: STRING>>,`employer` STRUCT<`id`: INT, `company`: STRUCT<`name`: STRING, `address`: STRING>>,`relations` MAP<STRUCT<`first`: STRING, `middle`: STRING, `last`: STRING>,STRING>,`p` INT

scala> spark.read.format(dataSourceName).schema(schema).load(path + "/contacts").createOrReplaceTempView("contacts")

scala> val query = spark.sql("select name.middle from contacts")
query: org.apache.spark.sql.DataFrame = [middle: string]

scala> query.show()
+------+                                                                        
|middle|
+------+
|    X.|
|    Y.|
|  null|
|  null|
+------+

Rapids:

scala> val dataSourceName = "parquet"
dataSourceName: String = parquet

scala> val path = "/home/fejiang/Desktop"
path: String = /home/fejiang/Desktop

scala> 

scala> val schema = ("`id` INT,`name` STRUCT<`first`: STRING, `middle`: STRING, `last`: STRING>, " +
     |   "`address` STRING,`pets` INT,`friends` ARRAY<STRUCT<`first`: STRING, `middle`: STRING, " +
     |   "`last`: STRING>>,`relatives` MAP<STRING, STRUCT<`first`: STRING, `middle`: STRING, " +
     |   "`last`: STRING>>,`employer` STRUCT<`id`: INT, `company`: STRUCT<`name`: STRING, " +
     |   "`address`: STRING>>,`relations` MAP<STRUCT<`first`: STRING, `middle`: STRING, " +
     |   "`last`: STRING>,STRING>,`p` INT")
schema: String = `id` INT,`name` STRUCT<`first`: STRING, `middle`: STRING, `last`: STRING>, `address` STRING,`pets` INT,`friends` ARRAY<STRUCT<`first`: STRING, `middle`: STRING, `last`: STRING>>,`relatives` MAP<STRING, STRUCT<`first`: STRING, `middle`: STRING, `last`: STRING>>,`employer` STRUCT<`id`: INT, `company`: STRUCT<`name`: STRING, `address`: STRING>>,`relations` MAP<STRUCT<`first`: STRING, `middle`: STRING, `last`: STRING>,STRING>,`p` INT

scala> spark.read.format(dataSourceName).schema(schema).load(path + "/contacts").createOrReplaceTempView("contacts")

scala> val query = spark.sql("select name.middle from contacts")
query: org.apache.spark.sql.DataFrame = [middle: string]

scala> query.show()
24/10/17 13:49:40 WARN GpuOverrides: 
!Exec <CollectLimitExec> cannot run on GPU because the Exec CollectLimitExec has been disabled, and is disabled by default because Collect Limit replacement can be slower on the GPU, if huge number of rows in a batch it could help by limiting the number of rows transferred from GPU to CPU. Set spark.rapids.sql.exec.CollectLimitExec to true if you wish to enable it
  @Partitioning <SinglePartition$> could run on GPU
  *Exec <ProjectExec> will run on GPU
    *Expression <Alias> name#1.middle AS middle#21 will run on GPU
      *Expression <GetStructField> name#1.middle will run on GPU
    *Exec <FileSourceScanExec> will run on GPU

24/10/17 13:49:41 ERROR Executor: Exception in task 1.0 in stage 1.0 (TID 2)    
java.lang.IndexOutOfBoundsException: Index: 0, Size: 0
	at java.util.ArrayList.rangeCheck(ArrayList.java:659)
	at java.util.ArrayList.get(ArrayList.java:435)
	at org.apache.parquet.format.converter.ParquetMetadataConverter.fromParquetMetadata(ParquetMetadataConverter.java:1493)
	at org.apache.parquet.format.converter.ParquetMetadataConverter.readParquetMetadata(ParquetMetadataConverter.java:1450)
	at org.apache.parquet.hadoop.ParquetFileReader.readFooter(ParquetFileReader.java:582)
	at org.apache.parquet.hadoop.ParquetFileReader.readFooter(ParquetFileReader.java:527)
	at org.apache.parquet.hadoop.ParquetFileReader.readFooter(ParquetFileReader.java:521)
	at com.nvidia.spark.rapids.GpuParquetFileFilterHandler.$anonfun$filterBlocks$5(GpuParquetScan.scala:709)
	at com.nvidia.spark.rapids.Arm$.withResource(Arm.scala:30)
	at com.nvidia.spark.rapids.GpuParquetFileFilterHandler.$anonfun$filterBlocks$4(GpuParquetScan.scala:705)
	at com.nvidia.spark.rapids.Arm$.withResource(Arm.scala:30)
	at com.nvidia.spark.rapids.GpuParquetFileFilterHandler.$anonfun$filterBlocks$1(GpuParquetScan.scala:704)