Short circuit evaluation

Python has a useful feature called short circuit evaluation that acts on logical operators or and and.

Short circuiting with the or operator

Consider this code:

x = 5
y = 10
if x > 6 or y < 20:
    print('OK')

When Python reaches the if statement, it starts evaluating the terms from left to right:

• x > 6 is false
• y < 20 is true
• therefore the result of the or expression is true
• the print statement executes

Now compare this code:

x = 7
y = 10
if x > 6 or y < 20:
    print('OK')

• x > 6 is true
• therefore the result of the or expression is true
• the print statement executes

In this case, as soon as Python discovers that x is greater than 6, it knows that the whole or statement is true, so it doesn't bother checking y. It doesn't matter whether y is less than 20 or not, it won't affect the outcome because an or statement is always true if either of the terms is true.

Python guarantees not to evaluate the second term if the first term is true.

Short circuiting with the and operator

Short circuiting with the and operator works in a similar way.

x = 5
y = 10
if x > 6 and y < 20:
    print('OK')

In this case:

• x > 6 is false
• therefore the result of the and expression is false
• the print statement doesn't execute

In this case, Python never evaluates y < 20 because it already knows the final result will be false.

In summary:

• An or statement stops as soon as if encounters the first True case
• An and statement stops as soon as if encounters the first False case

Short circuiting to avoid errors

We can often use short circuit evaluation to avoid errors. A common case is checking for a value being 0 before performing a calculation:

n = 3
m = 2
if m < 1/n:
    print('m < 1/n')

This code looks alright at first glance. But what happens if n is zero? Dividing by zero is an error, so our code would thrown an exception and stop working.

A quick way to avoid this is to use short circuiting, with the and operator:

n = 0
m = 2
if n and m < 1/n:
    print('m < 1/n')

With this new code, we have an and expression. The first term in n. If n is 0, which of course counts as false, Python knows that the whole an expression is false, so it never evaluates 1/n and there is no error.

If n is not zero, Python evaluates the second expression as normal.

You need to take care with this technique. Although we have successfully avoided throwing a divide by zero error, we have also ignored the fact that n is zero. This could lead to further errors in other parts of the code.

The value of an or/and expression

You might expect the value of a logical expression to always be True or False, but in fact this isn't the case. Python does something more useful.

a = 0
b = 2
c = 3
x = c or a:
print(x)

You may be surprised to see it print 3.

What actually happens is that Python prints the last value it evaluates. In this case, Python evaluates c, which has a value 3. Since this counts as true, it doesn't need to evaluate a, so it returns the value 3.

Here are some more examples, using the same values of a, b and c:

x = a or b:
print(x)  #prints 2
x = b or c:
print(x)  #prints 2
x = 0 or a:
print(x)  #prints 0 (the result is false because 0 and a are both false)
x = a or b+c or b+a:
print(x)  #prints 5 (b+c, the first true value)

Here is a real life example:

username=''
username = username or input('Enter username')

In this case, the or statement first evaluates username, which is an empty string, so it counts as false. It then evaluates the input statement - that is, it calls the input function, waits for the user to type in their name, and returns that value.

If we were to execute this statement again, later, when username has a value, the or expression would simply return the current value, and would not call input() at all.

Getting a true or false value

It doesn't usually matter that a or b gives a value of 2 rather than True, because it works fine in an if statement.

If you really need to, you can always use the bool() function. bool() converts any value to True or False.

See also

• List comprehensions
• Objects and variables
• Objects and identity
• Immutable objects
• Global variables
• Data types
• Lists vs tuples
• Sequences
• Named tuples
• Operators
• Walrus Operator
• For loops
• For loop using range vs iterables
• Changing the loop order
• Using enumerate in a for loop
• Using zip in a for loop
• Looping over multiple items (old article)
• Looping over selected items
• Functions
• Declaring functions
• Calling functions
• Function objects and lambdas
• Function decorators
• With statements
• Exception handling
• String functions
• Built-in functions
• Optimisation
• Optimisation good practice
• Low level code optimisation
• Structural optimisation