为什么在 Python 的标准库中没有排序的容器？

小开

There is a heapq in the standard library, it is not exactly sorted, but kind of. There is also a blist package, but it is not in the standard library.

小开

Python lists are ordered. If you sort them, they stay that way. In Python 2.7 an OrderedDict type was added to maintain an explicitly ordereded dictionary.

Python also has sets (a collection in which the members must be unique), but by definition they are unordered. Sorting a set just returns a list.

小开

最佳答案

It's a conscious design decision on Guido's part (he was even somewhat reluctant regarding the addition of the collections module). His goal is to preserve "one obvious way to do it" when it comes to the selection of data types for applications.

The basic concept is that if a user is sophisticated enough to realise that the builtin types aren't the right solution for their problem, then they're also up to the task of finding an appropriate third party library.

Given that list+sorting, list+heapq and list+bisect cover many of the use cases that would otherwise rely on inherently sorted data structures, and packages like blist exist, there isn't a huge drive to add more complexity in this space to the standard library.

In some ways, it is similar to the fact that there's no multi-dimensional array in the standard library, instead ceding that task to the NumPy folks.

小开

There is also the blist module that contains a sortedset data type:

sortedset(iterable=(), key=None)


>>> from blist import sortedset
>>> my_set = sortedset([3,7,2,2])
sortedset([2, 3, 7]

小开

Not exactly a "sorted container", but you might be interested in the standard library's bisect module, which "provides support for maintaining a list in sorted order without having to sort the list after each insertion".

小开

There's also a python sortedcontainers module that implements sorted list, dict, and set types. It's very similar to blist but implemented in pure-Python and in most cases faster.

>>> from sortedcontainers import SortedSet
>>> ss = SortedSet([3, 7, 2, 2])
>>> ss
SortedSet([2, 3, 7])

It also has functionality uncommon to other packages:

>>> from sortedcontainers import SortedDict
>>> sd = SortedDict((num, num) for num in range(100000))
>>> sd.iloc[-5] # Lookup the fifth-to-last key.
99995

Disclosure: I am the author of the sortedcontainers module.