Or¶

Overview¶

The Or expression implements the logical OR operation between two boolean expressions. It follows three-valued logic (3VL) rules, where the result can be true, false, or null depending on the input values and their nullability.

Syntax¶

expression1 OR expression2
-- Alternative symbol syntax
expression1 || expression2

// DataFrame API
col("column1") || col("column2")
// or using the or method
col("column1").or(col("column2"))

Arguments¶

Argument	Type	Description
left	Expression	Left operand boolean expression
right	Expression	Right operand boolean expression

Return Type¶

BooleanType (can be true, false, or null)

Supported Data Types¶

BooleanType for both left and right operands

Algorithm¶

If the left operand evaluates to true, return true immediately (short-circuit evaluation)
If the left operand is not true, evaluate the right operand
If the right operand evaluates to true, return true
If both operands are non-null and neither is true, return false
If either operand is null and neither evaluates to true, return null

Partitioning Behavior¶

This expression preserves partitioning as it operates on individual rows
Does not require shuffle operations
Can be pushed down to data sources for predicate pushdown optimization

Edge Cases¶

Null handling: Follows SQL three-valued logic where true OR null = true, false OR null = null, and null OR null = null
Short-circuit evaluation: If the left operand is true, the right operand is not evaluated
Type safety: Both operands must be boolean expressions; type mismatches are caught at analysis time

Code Generation¶

This expression supports Tungsten code generation with optimized paths for both nullable and non-nullable operands. The generated code includes short-circuit evaluation and efficient null handling without falling back to interpreted mode.

Examples¶

-- Basic OR operation
SELECT true OR false;  -- Result: true

-- OR with NULL values
SELECT false OR NULL;  -- Result: NULL
SELECT true OR NULL;   -- Result: true

-- Using in WHERE clauses
SELECT * FROM table WHERE status = 'active' OR priority = 'high';

-- Complex expressions
SELECT * FROM table WHERE (age > 65 OR income < 30000) OR status = 'special';

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

// Basic OR operation
df.filter(col("status") === "active" || col("priority") === "high")

// Using or method
df.filter(col("status").equalTo("active").or(col("priority").equalTo("high")))

// Complex boolean logic
df.filter((col("age") > 65 || col("income") < 30000) || col("status") === "special")

Or¶