xiaonanyang-db commented on code in PR #37933:
URL: https://github.com/apache/spark/pull/37933#discussion_r977144993
##########
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/csv/CSVInferSchema.scala:
##########
@@ -233,7 +238,39 @@ class CSVInferSchema(val options: CSVOptions) extends
Serializable {
* is compatible with both input data types.
*/
private def compatibleType(t1: DataType, t2: DataType): Option[DataType] = {
- TypeCoercion.findTightestCommonType(t1,
t2).orElse(findCompatibleTypeForCSV(t1, t2))
+ (t1, t2) match {
+ case (DateType, TimestampType) | (DateType, TimestampNTZType) |
+ (TimestampNTZType, DateType) | (TimestampType, DateType) =>
+ // For a column containing a mixture of dates and timestamps
+ // infer it as timestamp type if its dates can be inferred as
timestamp type
+ // otherwise infer it as StringType
+ val dateFormat =
options.dateFormatInRead.getOrElse(DateFormatter.defaultPattern)
+ t1 match {
+ case DateType if canParseDateAsTimestamp(dateFormat, t2) =>
+ Some(t2)
+ case TimestampType | TimestampNTZType if
canParseDateAsTimestamp(dateFormat, t1) =>
+ Some(t1)
+ case _ => Some(StringType)
+ }
+ case _ => TypeCoercion.findTightestCommonType(t1,
t2).orElse(findCompatibleTypeForCSV(t1, t2))
+ }
+ }
+
+ /**
+ * Return true if strings of given date format can be parsed as timestamps
+ * 1. If user provides timestamp format, we will parse strings as
timestamps using
+ * Iso8601TimestampFormatter (with strict timestamp parsing). Any date
string can not be parsed
+ * as timestamp type in this case
+ * 2. Otherwise, we will use DefaultTimestampFormatter to parse strings as
timestamps, which
+ * is more lenient and can parse strings of some date formats as timestamps.
+ */
+ private def canParseDateAsTimestamp(dateFormat: String, tsType: DataType):
Boolean = {
+ if ((tsType.isInstanceOf[TimestampType] &&
options.timestampFormatInRead.isEmpty) ||
+ (tsType.isInstanceOf[TimestampNTZType] &&
options.timestampNTZFormatInRead.isEmpty)) {
+ LENIENT_TS_FORMATTER_SUPPORTED_DATE_FORMATS.contains(dateFormat)
Review Comment:
This is a behavior change in terms of Spark 3.3 branch, where a column with
mixed dates and timestamps could be inferred as timestamp type if possible when
no timestamp pattern specified.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
