SecondsOfTimeWithFraction¶

Overview¶

SecondsOfTimeWithFraction extracts the seconds component from a time value and returns it as a decimal with fractional precision. This expression is implemented as a RuntimeReplaceable that delegates to DateTimeUtils for the actual computation, preserving subsecond precision based on the input time type's precision.

Syntax¶

-- SQL syntax (function name may vary by implementation)
SECONDS_WITH_FRACTION(time_expression)

// DataFrame API usage
SecondsOfTimeWithFraction(timeColumn)

Arguments¶

Argument	Type	Description
child	Expression	The time expression from which to extract seconds with fractional component

Return Type¶

DecimalType(8, 6) - A decimal with precision 8 and scale 6, allowing values up to 99.999999 seconds.

Supported Data Types¶

TimeType with any precision
Any type conforming to AnyTimeType abstract data type

Algorithm¶

Determines the precision from the input TimeType, defaulting to MIN_PRECISION for non-TimeType inputs
Creates a StaticInvoke expression that calls DateTimeUtils.getSecondsOfTimeWithFraction
Passes the child expression and precision as arguments to the static method
Returns a decimal value representing seconds with up to 6 decimal places of fractional precision
Leverages Catalyst's code generation through StaticInvoke for optimized execution

Partitioning Behavior¶

This expression preserves partitioning characteristics:

Does not require shuffle operations
Maintains existing data partitioning since it's a row-level transformation
Can be safely pushed down in query optimization

Edge Cases¶

Null input values are handled by the underlying DateTimeUtils implementation
Non-TimeType inputs default to minimum precision (TimeType.MIN_PRECISION)
Fractional seconds are preserved up to microsecond precision (6 decimal places)
Invalid time values may result in null or error depending on DateTimeUtils behavior

Code Generation¶

This expression supports Tungsten code generation through the StaticInvoke mechanism, which generates efficient Java code that directly calls the DateTimeUtils static method, avoiding interpreted expression evaluation overhead.

Examples¶

-- Extract seconds with fractional component from time
SELECT SECONDS_WITH_FRACTION(TIME '14:30:25.123456') AS seconds_fraction;
-- Result: 25.123456

// DataFrame API usage
import org.apache.spark.sql.catalyst.expressions.SecondsOfTimeWithFraction

val df = spark.range(1).select(
  SecondsOfTimeWithFraction(col("time_column")).as("seconds_with_fraction")
)