New submission from William Chargin : The operator override dunder methods, like `__add__`, are described in the “Data model” docs here: https://docs.python.org/3.10/reference/datamodel.html#object.__add__ Those docs say:

For instance, to evaluate the expression `x + y`, where `x` is an instance of a class that has an `__add__()` method, `x.__add__(y)` is called.

But this is not correct: `x + y` always uses `type(x).__add__`, which may be different from `x.__add__` if `__add__` has been set directly on the instance. Demonstration: ``` class C: def __add__(self, other): return "from class" c = C() print(c + c) # prints "from class" c.__add__ = lambda other: "from instance" print(c.__add__(c)) # prints "from instance" print(type(c).__add__(c, c)) # prints "from class" print(c + c) # prints "from class"! ``` The same issue appears in the reversed operator dunder (`__radd__`, et al.) docs below. I have a patch that I can submit as a PR; just need a bpo number. ---------- assignee: docs@python components: Documentation messages: 376316 nosy: docs@python, wchargin priority: normal severity: normal status: open title: docs: operator dunder (`__add__`, et al.) invocations described incorrectly type: enhancement versions: Python 3.10, Python 3.6, Python 3.7, Python 3.8, Python 3.9 _______________________________________ Python tracker https://bugs.python.org/issue41706 _______________________________________