“是”运算符在Python中做了什么？

Question

I was (maybe wrongfully) thinking that is operator is doing id() comparison.

>>> x = 10
>>> y = 10
>>> id(x)
1815480092
>>> id(y)
1815480092
>>> x is y
True

However, with val is not None , it seems like that it's not that simple.

>>> id(not None)
2001680
>>> id(None)
2053536
>>> val = 10
>>> id(val)
1815480092
>>> val is not None
True

Then, what does 'is' operator do? Is it just object id comparison just I conjectured? If so, val is not None is interpreted in Python as not (val is None) ?

Answer 1

You missed that is not is an operator too .

Without is , the regular not operator returns a boolean:

>>> not None
True

not None is thus the inverse boolean 'value' of None . In a boolean context None is false:

>>> bool(None)
False

so not None is boolean True .

Both None and True are objects too, and both have a memory address (the value id() returns for the CPython implementation of Python):

>>> id(True)
4440103488
>>> id(not None)
4440103488
>>> id(None)
4440184448

is tests if two references are pointing to the same object ; if something is the same object, it'll have the same id() as well. is returns a boolean value, True or False .

is not is the inverse of the is operator. It is the equivalent of not (op1 is op2) , in one operator. It should not be read as op1 is (not op2) here:

>>> 1 is not None     # is 1 a different object from None?
True
>>> 1 is (not None)   # is 1 the same object as True?
False

Answer 2

As a complement to Martijn Pieters answer, None , True and False are singletons. There is only one instance of each of these values so it is safe to use is to test them..

>>> id(True)
8851952
>>> id(False)
8851920
>>> id(None)
8559264
>>> id(not None)
8851952
>>> id(1 == 0)
8851920
>>> id(1 == 1)
8851952

Answer 3

Python tries to emulate the English language syntax to make it more human readable, but in this instance the operator precedence is a little confusing.

>>> val is not None
True

In English, we are asking if val does not have a value that is represented by the same object as None . In your above example val=10 , so the answer is (correctly) True .

However, from a logical, syntactical viewpoint you have broken down the statement word-for-word and assumed it to be:

>>> val is (not None)
False

which, when enclosed in appropriate patentheses, returns your expected result ( False ).

As @Martijn Pieters points out, is not is an operator in its own right , and it the only operator at work here.

If you wanted to write it as a non ambiguous statement using only operators that contain no spaces, you could write:

>>> not (val is None)
True

But that is not how you would 'say' the intended statement in English, or possibly even write it in pseudo-code.