[SPARK-27035][SQL] Get more precise current time

## What changes were proposed in this pull request? In the PR, I propose to replace `System.currentTimeMillis()` by `Instant.now()` in the `CurrentTimestamp` expression. `Instant.now()` uses the best available clock in the system to take current time. See [JDK-8068730](https://bugs.openjdk.java.net/browse/JDK-8068730) for more details. In JDK8, `Instant.now()` provides results with millisecond resolution but starting from JDK9 resolution of results is increased up to microseconds. ## How was this patch tested? The changes were tested by `DateTimeUtilsSuite` and by `DateFunctionsSuite`. Closes apache#23945 from MaxGekk/current-time. Authored-by: Maxim Gekk <[email protected]> Signed-off-by: Sean Owen <[email protected]>
kai-chi · Mar 6, 2019 · 6001258 · 6001258
1 parent 5fa4ba0
commit 6001258
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 3 deletions.
diff --git a/docs/sql-migration-guide-upgrade.md b/docs/sql-migration-guide-upgrade.md
@@ -97,7 +97,9 @@ displayTitle: Spark SQL Upgrading Guide
 
     - the JDBC options `lowerBound` and `upperBound` are converted to TimestampType/DateType values in the same way as casting strings to TimestampType/DateType values. The conversion is based on Proleptic Gregorian calendar, and time zone defined by the SQL config `spark.sql.session.timeZone`. In Spark version 2.4 and earlier, the conversion is based on the hybrid calendar (Julian + Gregorian) and on default system time zone.
 
-  - In Spark version 2.4 and earlier, invalid time zone ids are silently ignored and replaced by GMT time zone, for example, in the from_utc_timestamp function. Since Spark 3.0, such time zone ids are rejected, and Spark throws `java.time.DateTimeException`. 
+  - In Spark version 2.4 and earlier, invalid time zone ids are silently ignored and replaced by GMT time zone, for example, in the from_utc_timestamp function. Since Spark 3.0, such time zone ids are rejected, and Spark throws `java.time.DateTimeException`.
+
+  - In Spark version 2.4 and earlier, the `current_timestamp` function returns a timestamp with millisecond resolution only. Since Spark 3.0, the function can return the result with microsecond resolution if the underlying clock available on the system offers such resolution.
 
 ## Upgrading From Spark SQL 2.3 to 2.4
 

diff --git a/...talyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/datetimeExpressions.scala b/...talyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/datetimeExpressions.scala
@@ -18,7 +18,7 @@
 package org.apache.spark.sql.catalyst.expressions
 
 import java.sql.Timestamp
-import java.time.LocalDate
+import java.time.{Instant, LocalDate}
 import java.time.temporal.IsoFields
 import java.util.{Locale, TimeZone}
 
@@ -96,7 +96,7 @@ case class CurrentTimestamp() extends LeafExpression with CodegenFallback {
   override def dataType: DataType = TimestampType
 
   override def eval(input: InternalRow): Any = {
-    System.currentTimeMillis() * 1000L
+    instantToMicros(Instant.now())
   }
 
   override def prettyName: String = "current_timestamp"