[SPARK-18877][SQL] `CSVInferSchema.inferField` on DecimalType should find a common type with `typeSoFar`

## What changes were proposed in this pull request?

CSV type inferencing causes `IllegalArgumentException` on decimal numbers with heterogeneous precisions and scales because the current logic uses the last decimal type in a **partition**. Specifically, `inferRowType`, the **seqOp** of **aggregate**, returns the last decimal type. This PR fixes it to use `findTightestCommonType`.

**decimal.csv**
```
9.03E+12
1.19E+11
```

**BEFORE**
```scala
scala> spark.read.format("csv").option("inferSchema", true).load("decimal.csv").printSchema
root
 |-- _c0: decimal(3,-9) (nullable = true)

scala> spark.read.format("csv").option("inferSchema", true).load("decimal.csv").show
16/12/16 14:32:49 ERROR Executor: Exception in task 0.0 in stage 4.0 (TID 4)
java.lang.IllegalArgumentException: requirement failed: Decimal precision 4 exceeds max precision 3
```

**AFTER**
```scala
scala> spark.read.format("csv").option("inferSchema", true).load("decimal.csv").printSchema
root
 |-- _c0: decimal(4,-9) (nullable = true)

scala> spark.read.format("csv").option("inferSchema", true).load("decimal.csv").show
+---------+
|      _c0|
+---------+
|9.030E+12|
| 1.19E+11|
+---------+
```

## How was this patch tested?

Pass the newly add test case.

Author: Dongjoon Hyun <dongjoon@apache.org>

Closes #16320 from dongjoon-hyun/SPARK-18877.

2 files changed, 20 insertions, 1 deletions

diff --git a/sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/csv/CSVInferSchema.scala b/sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/csv/CSVInferSchema.scala
index 88c608add1..adc92fe5a3 100644
--- a/sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/csv/CSVInferSchema.scala
+++ b/sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/csv/CSVInferSchema.scala
@@ -85,7 +85,9 @@ private[csv] object CSVInferSchema {
         case NullType => tryParseInteger(field, options)
         case IntegerType => tryParseInteger(field, options)
         case LongType => tryParseLong(field, options)
-        case _: DecimalType => tryParseDecimal(field, options)
+        case _: DecimalType =>
+          // DecimalTypes have different precisions and scales, so we try to find the common type.
+          findTightestCommonType(typeSoFar, tryParseDecimal(field, options)).getOrElse(StringType)
         case DoubleType => tryParseDouble(field, options)
         case TimestampType => tryParseTimestamp(field, options)
         case BooleanType => tryParseBoolean(field, options)
diff --git a/sql/core/src/test/scala/org/apache/spark/sql/execution/datasources/csv/CSVInferSchemaSuite.scala b/sql/core/src/test/scala/org/apache/spark/sql/execution/datasources/csv/CSVInferSchemaSuite.scala
index 93f752d107..8620bb9f65 100644
--- a/sql/core/src/test/scala/org/apache/spark/sql/execution/datasources/csv/CSVInferSchemaSuite.scala
+++ b/sql/core/src/test/scala/org/apache/spark/sql/execution/datasources/csv/CSVInferSchemaSuite.scala
@@ -114,4 +114,21 @@ class CSVInferSchemaSuite extends SparkFunSuite {
     val options = new CSVOptions(Map("TiMeStampFormat" -> "yyyy-mm"))
     assert(CSVInferSchema.inferField(TimestampType, "2015-08", options) == TimestampType)
   }
+
+  test("SPARK-18877: `inferField` on DecimalType should find a common type with `typeSoFar`") {
+    val options = new CSVOptions(Map.empty[String, String])
+
+    // 9.03E+12 is Decimal(3, -10) and 1.19E+11 is Decimal(3, -9).
+    assert(CSVInferSchema.inferField(DecimalType(3, -10), "1.19E+11", options) ==
+      DecimalType(4, -9))
+
+    // BigDecimal("12345678901234567890.01234567890123456789") is precision 40 and scale 20.
+    val value = "12345678901234567890.01234567890123456789"
+    assert(CSVInferSchema.inferField(DecimalType(3, -10), value, options) == DoubleType)
+
+    // Seq(s"${Long.MaxValue}1", "2015-12-01 00:00:00") should be StringType
+    assert(CSVInferSchema.inferField(NullType, s"${Long.MaxValue}1", options) == DecimalType(20, 0))
+    assert(CSVInferSchema.inferField(DecimalType(20, 0), "2015-12-01 00:00:00", options)
+      == StringType)
+  }
 }