Python Scope

Advanced Python Scope: Closures & Nonlocal State Mutation

A closure is a nested function that retains access to variables from its enclosing lexical scope even after the outer function has finished executing.

Closure Memory Binding

flowchart LR
    subgraph Enclosing Function Scope
        var["cell reference: rate = 0.15"]
    end
    subgraph Returned Inner Function
        closure["closure_fn(amount)"]
    end
    closure -->|__closure__| var

Requirements for a Closure

1. A nested function exists inside an outer enclosing function.
2. The nested function references a variable defined in the outer function.
3. The outer function returns the nested function reference.

Practical Code Example

from typing import Callable

def create_tax_calculator(tax_rate: float) -> Callable[[float], float]:
    """Factory function generating closure tax calculators."""
    # Enclosing variable 'tax_rate' is captured in the closure
    def calculate_tax(amount: float) -> float:
        return round(amount * tax_rate, 2)

    return calculate_tax

def create_accumulator(initial_value: int = 0) -> Callable[[int], int]:
    """Closure maintaining mutable state via 'nonlocal'."""
    state = initial_value

    def add(value: int) -> int:
        nonlocal state  # Mutate captured enclosing variable
        state += value
        return state

    return add

if __name__ == "__main__":
    vat_calculator = create_tax_calculator(0.20)
    sales_tax_calculator = create_tax_calculator(0.07)

    print(f"VAT on $100: ${vat_calculator(100.0)}")
    print(f"Sales Tax on $100: ${sales_tax_calculator(100.0)}")

    counter = create_accumulator(10)
    print("Accumulator +5:", counter(5))   # 15
    print("Accumulator +20:", counter(20)) # 35

Best Practices & Gotchas

• Late Binding in Closures: Inner functions bind variable names, not values, at lookup time. Beware of creating closures inside loops ([lambda: i for i in range(5)] binds i=4 for all lambdas!).
• Use functools.partial for Simple Currying: For simple function argument binding without mutable state, prefer functools.partial.
• Inspect Closures: You can inspect captured closure variables via fn.__closure__[0].cell_contents.

Self-Check Challenge

Write a function make_multiplier(n: int) returning a closure function that multiplies its input parameter by n.