Managing State

Testability

• In order to test code, it helps for it to be written with tests in mind
• Isolation of functionality is important

Readability

• Code that is easier to test is often easy to read
• Breaking code at distinct test boundaries can simplify the architecture

Pure Functions

• No side effects
• No reliance on external state
• Always return the same output given the same inputs

Pure Function

def absolute_value(x):
    if x < 0:
        return -x
    else:
        return x

Impure Function

n = 2

def square():
  return n*n

n = 4
print(square())

Global State

• Global state exists for the life a program
• Local state exists in a section of code
• Code that references global state can be more difficult to reason about

Local Variables

• Are only defined within a function
• Torn down when function terminates
• Useful for storing temporary data

Local Variable

def is_freezing(temp, unit):
    if unit == 'C':
        temp_f = temp * 9/5 + 32
    else:
        temp_f = temp

    if temp_f < 32:
        return True
    else:
        return False

Global Variable

• Defined over the entire scope
• Torn down only on program termination
• Useful for storing global data

Global Variable

def get_tip(price):
    return int(percentage) / 100 * float(price)

percentage = input("Tip percentage: ")
price = input("Meal price: ")

print("Tip:", get_tip(price))

How would we test get_tip?

State

• Managing state is critical for creating readable code
• Testability is enhance when we limit external state