[SPARK-33641] Invalidate new char-like type in public APIs that result incorrect results

• Type: Bug
• Status: Resolved
• Priority: Blocker
• Resolution: Fixed
• Affects Version/s: 3.1.0
• Fix Version/s: 3.1.0
• Component/s: SQL
• Assignee: Kent Yao 2
• Reporter: Kent Yao 2
• Created: 03/Dec/20 07:35
• Updated: 13/Jul/23 09:45
• Resolved: 07/Dec/20 13:40

Description

1. udf

scala> spark.udf.register("abcd", () => "12345", org.apache.spark.sql.types.VarcharType(2))

scala> spark.sql("select abcd()").show
scala.MatchError: CharType(2) (of class org.apache.spark.sql.types.VarcharType)
  at org.apache.spark.sql.catalyst.encoders.RowEncoder$.externalDataTypeFor(RowEncoder.scala:215)
  at org.apache.spark.sql.catalyst.encoders.RowEncoder$.externalDataTypeForInput(RowEncoder.scala:212)
  at org.apache.spark.sql.catalyst.expressions.objects.ValidateExternalType.<init>(objects.scala:1741)
  at org.apache.spark.sql.catalyst.encoders.RowEncoder$.$anonfun$serializerFor$3(RowEncoder.scala:175)
  at scala.collection.TraversableLike.$anonfun$flatMap$1(TraversableLike.scala:245)
  at scala.collection.IndexedSeqOptimized.foreach(IndexedSeqOptimized.scala:36)
  at scala.collection.IndexedSeqOptimized.foreach$(IndexedSeqOptimized.scala:33)
  at scala.collection.mutable.ArrayOps$ofRef.foreach(ArrayOps.scala:198)
  at scala.collection.TraversableLike.flatMap(TraversableLike.scala:245)
  at scala.collection.TraversableLike.flatMap$(TraversableLike.scala:242)
  at scala.collection.mutable.ArrayOps$ofRef.flatMap(ArrayOps.scala:198)
  at org.apache.spark.sql.catalyst.encoders.RowEncoder$.serializerFor(RowEncoder.scala:171)
  at org.apache.spark.sql.catalyst.encoders.RowEncoder$.apply(RowEncoder.scala:66)
  at org.apache.spark.sql.Dataset$.$anonfun$ofRows$2(Dataset.scala:99)
  at org.apache.spark.sql.SparkSession.withActive(SparkSession.scala:768)
  at org.apache.spark.sql.Dataset$.ofRows(Dataset.scala:96)
  at org.apache.spark.sql.SparkSession.$anonfun$sql$1(SparkSession.scala:611)
  at org.apache.spark.sql.SparkSession.withActive(SparkSession.scala:768)
  at org.apache.spark.sql.SparkSession.sql(SparkSession.scala:606)
  ... 47 elided

2. spark.createDataframe

scala> spark.createDataFrame(spark.read.text("README.md").rdd, new org.apache.spark.sql.types.StructType().add("c", "char(1)")).show
+--------------------+
|                   c|
+--------------------+
|      # Apache Spark|
|                    |
|Spark is a unifie...|
|high-level APIs i...|
|supports general ...|
|rich set of highe...|
|MLlib for machine...|
|and Structured St...|
|                    |
|<https://spark.ap...|
|                    |
|[![Jenkins Build]...|
|[![AppVeyor Build...|
|[![PySpark Covera...|
|                    |
|                    |
|## Online Documen...|
|                    |
|You can find the ...|
|guide, on the [pr...|
+--------------------+
only showing top 20 rows

3. reader.schema

```
scala> spark.read.schema("a varchar(2)").text("./README.md").show(100)
--------------------

a

--------------------

Apache Spark

Spark is a unifie...

high-level APIs i...

supports general ...

```
4. etc

Issue Links

• is caused by: SPARK-33480 support char/varchar type
• is related to: SPARK-44409 Handle char/varchar in Dataset.to to keep consistent with others
• links to: [Github] Pull Request #30586 (yaooqinn)
• links to: [Github] Pull Request #30654 (yaooqinn)