Error Handling

Raising exceptions with raise

You can trigger exceptions yourself with the raise keyword. This is the standard Python way to signal that an input is invalid or a precondition was not met.

So far you've reacted to errors. Now you'll trigger them yourself. The raise keyword tells Python: stop, this situation is invalid.

Basic syntax: pick an exception class and pass a message.

raise ValueError("something is wrong")

Python

A function that refuses negative input:

def square(n):
	if n < 0:
		raise ValueError("no negatives")
	return n * n
print(square(4))

Python

Output

When the bad input arrives, the function raises and never returns:

def square(n):
	if n < 0:
		raise ValueError("no negatives")
	return n * n
try:
	print(square(-1))
except ValueError as e:
	print(str(e))

Python

Output

What will be the output?

def check(n):
	if n == 0:
		raise ValueError("zero")
	return n
try:
	check(0)
except ValueError as e:
	print(str(e))

Python

Pick the exception class that fits the problem. ValueError signals a bad value, TypeError signals the wrong kind of object.

A type check using TypeError:

def shout(text):
	if type(text) is not str:
		raise TypeError("need str")
	return text.upper()
try:
	shout(123)
except TypeError as e:
	print(str(e))

Python

Output

What will be the output?

def half(n):
	if type(n) is not int:
		raise TypeError("int only")
	return n / 2
try:
	half("5")
except TypeError as e:
	print(str(e))

Python

Inside an except block, a bare raise re-raises the exception you just caught. This is useful for logging the error and still letting it bubble up.

Re-raise after recording what happened:

log = []
try:
	try:
		raise ValueError("oops")
	except ValueError:
		log.append("seen")
		raise
except ValueError:
	log.append("again")
print(log)

Python

Output

What will be the output?

def divide(a, b):
	if b == 0:
		raise ZeroDivisionError("no")
	return a / b
try:
	divide(10, 0)
except ZeroDivisionError as e:
	print(str(e))

Python

Raised exceptions travel up through every function call until something catches them, or the program exits with a traceback.

The exception propagates from inner() through outer() and is caught at the top:

def inner():
	raise ValueError("from inner")
def outer():
	inner()
try:
	outer()
except ValueError as e:
	print(str(e))

Python

Output

What will be the output?

def check(age):
	if age < 0:
		raise ValueError("neg")
	return age
try:
	check(-5)
except ValueError as e:
	result = str(e)
print(result)

Python

What will be the output?

def need_str(v):
	if type(v) is not str:
		raise TypeError("nope")
	return len(v)
try:
	print(need_str("hi"))
except TypeError as e:
	print(str(e))

Python

What will be the output?

def divisible(n, by):
	if by == 0:
		raise ValueError("div zero")
	return n % by == 0
try:
	print(divisible(10, 0))
except ValueError as e:
	print(str(e))

Python