Now we understand how code like this works.

message = input("Enter something: ")

if message != '':

print("You entered:", message)

else:

print("You didn't enter anything!")

But most Python programmers would write that code like this

instead:

message = input("Enter something: ")

if message:

print("You entered:", message)

else:

print("You didn't enter anything!")

What the heck was that? We did `if message`, but `message`

was a string, not True or False!

Python converted our message to a Boolean and then checked if

the Boolean it ended up with was True. But when will it be true?

We can convert things to Booleans like Python did by doing

`bool(things)`. Let's try that with strings.

>>> bool('hello')

>>> bool('there')

>>> bool('True')

>>> bool('False') # this isn't special in any way

>>>

The `if message:` actually did the same thing as `if bool(message)`,

which is same as `if bool(message) == True:`. Usually we just don't

write the `==True` part anywhere because we don't need it.

As we can see, the Boolean value of most strings is True. The

only string that has a false Boolean value is the empty string,

`''` or `""`:

>>> bool('')

>>>

Most other things are also treated as False if they're empty and

True if they're not empty.

>>> bool([1, 2, 3])

>>> bool([])

>>> bool((1, 2, 3))

>>> bool(())

>>> bool({'a': 1, 'b': 2})

>>> bool({})

>>>

None and zero are also falsy, but positive and negative numbers

are treated as True.

>>> bool(None)

>>> bool(0)

>>> bool(0.0)

>>> bool(1)

>>> bool(-1)

>>>

Most other things are also treated as True.

>>> bool(OSError)

>>> bool(print)

>>>

It's recommended to rely on the Boolean value when we're doing

something with things like lists and tuples. This way our code

will work even if it gets a value of a different type than we

were expected it to get originally.

For example, this code doesn't work right if we give it

something else than a list. It thinks that empty tuples,

strings and dictionaries aren't empty just because they aren't

empty lists:

>>> def is_this_empty(thing):

... if thing == []:

... print("It's empty!")

... else:

... print("It's not empty.")

...

>>> is_this_empty([1, 2, 3])

It's not empty.

>>> is_this_empty([])

It's empty!

>>> is_this_empty(())

It's not empty.

>>> is_this_empty('')

It's not empty.

>>> is_this_empty({})

It's not empty.

>>>

We could improve the code by checking against different empty

things.

>>> def is_this_empty(thing):

... if thing == [] or thing == () or thing == '' or thing == {}:

... print("It's empty!")

... else:

... print("It's not empty.")

...

>>>

But Python has many other data types that can be empty and we

haven't talked about in this tutorial. Trying to check all of

them would be pointless because functions like this already

work with all of them:

>>> def is_this_empty(thing):

... if thing:

... print("It's not empty.")

... else:

... print("It's empty!")

...

>>>

There's also cases when we should not rely on the Boolean value.

When we're doing things with numbers and None it's best to

simply compare to None or zero. Like this:

if number != 0:

print("number is not zero")

if value is not None:

print("value is not None")

Not like this:

if number:

print("number is not zero")

if value:

print("value is not None")

We used `is not` instead of `!=` in the first example because

the official style guide recommends it. The reason is that it's

possible to create a value that isn't really None but seems like

None when we compare it with None using `==` or `!=`, and we want

to make sure that we don't treat values like that as None.

So here's how we should check if something is None:

if not value: ... # not good if we want to check if it's None

if value == None: ... # better

if value is None: ... # best

- `if thing:` does the same thing as `if bool(thing):`.

- `bool()` of most things is True, but `bool()` of None, zero and

most empty things are False.

- Use `is` and `is not` when comparing to None, `==` and `!=` when

comparing to numbers and rely on the Boolean value otherwise.