Fluent Python

Chapter 10. Sequence Hacking, Hashing, and Slicing

Don’t check whether it is-a duck: check whether it quacks-like-a duck, walks-like-a duck, etc, etc, depending on exactly what subset of duck-like behavior you need to play your language-games with. (comp.lang.python, Jul. 26, 2000)

Alex Martelli

In this chapter, we will create a class to represent a multidimensional Vector class—a significant step up from the two-dimensional Vector2d of Chapter 9. Vector will behave like a standard Python immutable flat sequence. Its elements will be floats, and it will support the following by the end of this chapter:

• Basic sequence protocol: __len__ and __getitem__.

• Safe representation of instances with many items.

• Proper slicing support, producing new Vector instances.

• Aggregate hashing taking into account every contained element value.

• Custom formatting language extension.

We’ll also implement dynamic attribute access with __getattr__ as a way of replacing the read-only properties we used in Vector2d—although this is not typical of sequence types.

The code-intensive presentation will be interrupted by a conceptual discussion about the idea of protocols as an informal interface. We’ll talk about how protocols and duck typing are related, and its practical implications when you create your own types.

Let’s get started.

Vector Applications Beyond Three Dimensions

Who needs a vector with 1,000 dimensions? Hint: not 3D artists! However, n-dimensional vectors (with large values of ...