LocalTimestamp¶

Overview¶

The LocalTimestamp expression returns the current timestamp without timezone information at the time of query execution. It provides a timestamp in the local timezone as a TimestampNTZ (timestamp without timezone) data type, representing the current date and time when the expression is evaluated.

Syntax¶

LOCALTIMESTAMP()
-- or
SELECT localtimestamp();

// DataFrame API
import org.apache.spark.sql.functions._
df.select(expr("localtimestamp()"))

Arguments¶

This expression takes no arguments. The timeZoneId parameter is used internally for timezone-aware processing but is not exposed to users.

Return Type¶

TimestampNTZType - Timestamp without timezone information, represented as microseconds since epoch.

Supported Data Types¶

This is a leaf expression that generates timestamp values and does not accept input data of any type.

Algorithm¶

Retrieves the current system time using LocalDateTime.now(zoneId)
Converts the LocalDateTime to microseconds since epoch using localDateTimeToMicros
Uses the configured timezone (zoneId) to determine the local time context
Returns the timestamp value as TimestampNTZ type without timezone information
The expression is foldable, meaning it can be constant-folded during optimization

Partitioning Behavior¶

Preserves partitioning as it generates the same value for all rows in a given query execution
Does not require shuffle operations
Can be used safely in partitioned operations without affecting partition boundaries

Edge Cases¶

Never returns null (nullable = false)
Returns the same timestamp value for all rows within a single query execution due to constant folding
Timezone handling depends on the session's configured timezone
The timestamp represents the query execution time, not row processing time

Code Generation¶

This expression uses CodegenFallback, meaning it falls back to interpreted mode rather than generating optimized Java code. The evaluation uses the eval method for all executions.

Examples¶

-- Get current local timestamp
SELECT localtimestamp();
-- Output: 2020-04-25 15:49:11.914

-- Use in SELECT with other columns
SELECT id, name, localtimestamp() as created_at FROM users;

-- Use in WHERE clause for time-based filtering
SELECT * FROM events WHERE event_time > localtimestamp() - INTERVAL 1 HOUR;

// DataFrame API usage
import org.apache.spark.sql.functions._

// Add current timestamp column
df.withColumn("processed_at", expr("localtimestamp()"))

// Filter using current timestamp
df.filter(col("updated_at") > expr("localtimestamp() - INTERVAL 1 DAY"))