UnresolvedNamedLambdaVariable¶

Overview¶

UnresolvedNamedLambdaVariable is a placeholder expression used during the analysis phase to represent lambda variables in higher-order functions before they are resolved. It prevents unexpected resolution of lambda functions during the initial parsing and analysis stages of query planning.

Syntax¶

This is an internal Catalyst expression that is not directly accessible through SQL or DataFrame API. It is created automatically during the parsing of lambda expressions in higher-order functions like transform, filter, exists, etc.

Arguments¶

Argument	Type	Description
nameParts	Seq[String]	Sequence of string parts that make up the lambda variable name

Return Type¶

This expression is Unevaluable and throws UnresolvedException for all type-related methods. It does not have a concrete return type as it must be resolved before evaluation.

Supported Data Types¶

Not applicable - this expression is a placeholder that must be resolved before type checking occurs.

Algorithm¶

Stores the lambda variable name as a sequence of string parts
Marks itself as unresolved (resolved = false)
Throws UnresolvedException for any attempt to access type information or evaluation methods
Provides string representation for debugging and SQL generation
Uses LAMBDA_VARIABLE tree pattern for pattern matching during analysis

Partitioning Behavior¶

Not applicable - this expression is resolved before physical planning and does not affect partitioning behavior.

Edge Cases¶

All type-related methods (dataType, nullable, qualifier) throw UnresolvedException
All evaluation-related methods throw UnresolvedException since it extends Unevaluable
Expression ID and attribute conversion methods throw UnresolvedException
Name parts containing dots are wrapped in backticks for proper SQL representation

Code Generation¶

This expression does not support code generation as it is Unevaluable and must be resolved to concrete expressions before the code generation phase.

Examples¶

-- Lambda variables are created implicitly in higher-order functions
-- The 'x' below would initially be represented as UnresolvedNamedLambdaVariable
SELECT transform(array(1, 2, 3), x -> x * 2);

// DataFrame API usage where lambda variables are created
// The lambda parameter would be UnresolvedNamedLambdaVariable initially
import org.apache.spark.sql.functions._
df.select(transform(col("numbers"), x => x * 2))