search

Kyligence / ClickHouse

ClickHouse® is a free analytics DBMS for big data
https://clickhouse.com
Apache License 2.0
14 stars 17 forks source link

CHColumnToSparkRow occurs NPE when the size of map field is 0 #422

Closed exmy closed 1 year ago

exmy commented 1 year ago

Describe what's wrong

NPE can happens when converting a CH Block to a Spark UnsafeRow, if the block contains a map field with a size of 0.

The root cause is that when converting the map field, the memory layout does not conform to UnsafeMapData when the map size is 0.

Caused by: java.lang.NullPointerException
    at scala.math.LowPriorityOrderingImplicits$$anon$2.compare(Ordering.scala:150)
    at scala.math.Ordering.equiv(Ordering.scala:103)
    at scala.math.Ordering.equiv$(Ordering.scala:103)
    at scala.math.LowPriorityOrderingImplicits$$anon$2.equiv(Ordering.scala:149)
    at org.apache.spark.sql.catalyst.expressions.GetMapValueUtil.getValueEval(complexTypeExtractors.scala:362)
    at org.apache.spark.sql.catalyst.expressions.GetMapValueUtil.getValueEval$(complexTypeExtractors.scala:342)
    at org.apache.spark.sql.catalyst.expressions.GetMapValue.getValueEval(complexTypeExtractors.scala:442)
    at org.apache.spark.sql.catalyst.expressions.GetMapValue.nullSafeEval(complexTypeExtractors.scala:484)
    at org.apache.spark.sql.catalyst.expressions.BinaryExpression.eval(Expression.scala:574)
    at org.apache.spark.sql.catalyst.expressions.Alias.eval(namedExpressions.scala:168)
    at org.apache.spark.sql.catalyst.expressions.InterpretedUnsafeProjection.apply(InterpretedUnsafeProjection.scala:90)
    at org.apache.spark.sql.catalyst.expressions.InterpretedUnsafeProjection.apply(InterpretedUnsafeProjection.scala:34)