[GitHub] [spark] Hisoka-X commented on a diff in pull request #41078: [SPARK-39280][SQL] Speed up Timestamp type inference with user-provided format in JSON/CSV data source

["Hisoka-X (via GitHub)" <gi...@apache.org>]

Hisoka-X commented on code in PR #41078:
URL: https://github.com/apache/spark/pull/41078#discussion_r1186872738


##########
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/util/TimestampFormatter.scala:
##########
@@ -163,27 +165,66 @@ class Iso8601TimestampFormatter(
   protected lazy val legacyFormatter = TimestampFormatter.getLegacyFormatter(
     pattern, zoneId, locale, legacyFormat)
 
+  override def parseOptional(s: String): Option[Long] = {
+    try {
+      val parsed = formatter.parseUnresolved(s, new ParsePosition(0))
+      if (parsed != null) {
+        val (epochSeconds, microsOfSecond) = extractSeconds(parsed)
+        Some(Math.addExact(Math.multiplyExact(epochSeconds, MICROS_PER_SECOND), microsOfSecond))
+      } else {
+        None
+      }
+    } catch {
+      case NonFatal(_) => None

Review Comment:
   Yes, we want to avoiding exceptions like `DateTimeParseException`, it cause by field value can't be parsed by format. There only catch JVM error like `VirtualMachineError`, same opeartion with `DateTimeUtils.stringToTimestamp`



-- 
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-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org