UnixDate¶

Overview¶

The UnixDate expression converts a date value to its Unix date representation, which is the number of days since the Unix epoch (1970-01-01). This expression provides a way to get the integer representation of a date for numerical operations or storage optimization.

Syntax¶

unix_date(date_expr)

// DataFrame API
col("date_column").expr("unix_date(date_column)")

Arguments¶

Argument	Type	Description
date_expr	DATE	The date value to convert to Unix date representation

Return Type¶

INTEGER - Returns the number of days since 1970-01-01 as an integer value.

Supported Data Types¶

DateType - Only date type inputs are supported

Algorithm¶

Takes the internal date representation (already stored as days since epoch)
Performs a direct cast to integer without any transformation
Returns the integer value representing days since Unix epoch
Uses null-safe evaluation to handle null inputs appropriately
Implements code generation for optimized execution

Partitioning Behavior¶

This expression preserves partitioning characteristics:

Does not require shuffle operations
Maintains deterministic output for the same input
Can be used safely in partitioning schemes
Preserves data locality during processing

Edge Cases¶

Null handling: Returns null when input date is null (null-safe evaluation)
Negative dates: Dates before 1970-01-01 will return negative integer values
Type enforcement: Only accepts DATE type inputs, other types will cause compilation errors
Integer overflow: Extremely large date values could theoretically cause integer overflow

Code Generation¶

This expression fully supports Tungsten code generation:

Implements doGenCode method for compiled code paths
Uses direct value passing without complex transformations
Generates efficient native code for high-performance execution
Falls back gracefully to interpreted mode if code generation fails

Examples¶

-- Convert specific date to Unix date
SELECT unix_date(DATE('1970-01-02'));
-- Returns: 1

-- Convert date column
SELECT unix_date(birth_date) FROM users;

-- Use in calculations
SELECT unix_date(end_date) - unix_date(start_date) AS duration_days FROM events;

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

df.select(expr("unix_date(date_column)"))

// In transformations
df.withColumn("unix_date", expr("unix_date(original_date)"))

// Calculate date differences
df.select(
  expr("unix_date(end_date) - unix_date(start_date)").alias("duration")
)