让我就这一需求提出自己的看法:

from typing import Tuple, Union, Optional, Generator


StrOrInt = Union[str, int]


# On Python 3.6, string concatenation is REALLY fast
# Tested myself, and this fella also tested:
# https://blog.ganssle.io/articles/2019/11/string-concat.html
def griter(s: str) -> Generator[StrOrInt, None, None]:
    last_was_digit: Optional[bool] = None
    cluster: str = ""
    for c in s:
        if last_was_digit is None:
            last_was_digit = c.isdigit()
            cluster += c
            continue
        if c.isdigit() != last_was_digit:
            if last_was_digit:
                yield int(cluster)
            else:
                yield cluster
            last_was_digit = c.isdigit()
            cluster = ""
        cluster += c
    if last_was_digit:
        yield int(cluster)
    else:
        yield cluster
    return


def grouper(s: str) -> Tuple[StrOrInt, ...]:
    return tuple(griter(s))

现在如果我们有这样的列表:

filelist = [
    'File3', 'File007', 'File3a', 'File10', 'File11', 'File1', 'File4', 'File5',
    'File9', 'File8', 'File8b1', 'File8b2', 'File8b11', 'File6'
]

我们可以简单地使用key= kwarg来进行自然排序:

>>> sorted(filelist, key=grouper)
['File1', 'File3', 'File3a', 'File4', 'File5', 'File6', 'File007', 'File8', 
'File8b1', 'File8b2', 'File8b11', 'File9', 'File10', 'File11']

当然，这里的缺点是，就像现在一样，该函数将对大写字母在小写字母之前进行排序。

我将把不区分大小写的grouper的实现留给读者:-)

试试这个:

import re

def natural_sort(l): 
    convert = lambda text: int(text) if text.isdigit() else text.lower()
    alphanum_key = lambda key: [convert(c) for c in re.split('([0-9]+)', key)]
    return sorted(l, key=alphanum_key)

输出:

['elm0', 'elm1', 'Elm2', 'elm9', 'elm10', 'Elm11', 'Elm12', 'elm13']

代码改编自这里:排序人类:自然排序顺序。

>>> import re
>>> sorted(lst, key=lambda x: int(re.findall(r'\d+$', x)[0]))
['elm0', 'elm1', 'Elm2', 'elm9', 'elm10', 'Elm11', 'Elm12', 'elm13']

data = ['elm13', 'elm9', 'elm0', 'elm1', 'Elm11', 'Elm2', 'elm10']

让我们分析一下数据。所有元素的数字容量为2。在常见的字面部分“elm”中有3个字母。

所以，元素的最大长度是5。我们可以增加这个值以确保(例如，增加到8)。

记住这一点，我们有一个简单的解决方案:

data.sort(key=lambda x: '{0:0>8}'.format(x).lower())

没有正则表达式和外部库!

print(data)

>>> ['elm0', 'elm1', 'Elm2', 'elm9', 'elm10', 'Elm11', 'elm13']

解释:

for elm in data:
    print('{0:0>8}'.format(elm).lower())

>>>
0000elm0
0000elm1
0000elm2
0000elm9
000elm10
000elm11
000elm13

一个紧凑的解决方案，基于将字符串转换为List[Tuple(str, int)]。

Code

def string_to_pairs(s, pairs=re.compile(r"(\D*)(\d*)").findall):
    return [(text.lower(), int(digits or 0)) for (text, digits) in pairs(s)[:-1]]

示范

sorted(['Elm11', 'Elm12', 'Elm2', 'elm0', 'elm1', 'elm10', 'elm13', 'elm9'], key=string_to_pairs)

输出:

['elm0', 'elm1', 'Elm2', 'elm9', 'elm10', 'Elm11', 'Elm12', 'elm13']

测试

转换

assert string_to_pairs("") == []
assert string_to_pairs("123") == [("", 123)]
assert string_to_pairs("abc") == [("abc", 0)]
assert string_to_pairs("123abc") == [("", 123), ("abc", 0)]
assert string_to_pairs("abc123") == [("abc", 123)]
assert string_to_pairs("123abc456") == [("", 123), ("abc", 456)]
assert string_to_pairs("abc123efg") == [("abc", 123), ("efg", 0)]

排序

# Some extracts from the test suite of the natsort library. Permalink:
# https://github.com/SethMMorton/natsort/blob/e3c32f5638bf3a0e9a23633495269bea0e75d379/tests/test_natsorted.py

sort_data = [
    (  # same as test_natsorted_can_sort_as_unsigned_ints_which_is_default()
        ["a50", "a51.", "a50.31", "a-50", "a50.4", "a5.034e1", "a50.300"],
        ["a5.034e1", "a50", "a50.4", "a50.31", "a50.300", "a51.", "a-50"],
    ),
    (  # same as test_natsorted_numbers_in_ascending_order()
        ["a2", "a5", "a9", "a1", "a4", "a10", "a6"],
        ["a1", "a2", "a4", "a5", "a6", "a9", "a10"],
    ),
    (  # same as test_natsorted_can_sort_as_version_numbers()
        ["1.9.9a", "1.11", "1.9.9b", "1.11.4", "1.10.1"],
        ["1.9.9a", "1.9.9b", "1.10.1", "1.11", "1.11.4"],
    ),
    (  # different from test_natsorted_handles_filesystem_paths()
        [
            "/p/Folder (10)/file.tar.gz",
            "/p/Folder (1)/file (1).tar.gz",
            "/p/Folder/file.x1.9.tar.gz",
            "/p/Folder (1)/file.tar.gz",
            "/p/Folder/file.x1.10.tar.gz",
        ],
        [
            "/p/Folder (1)/file (1).tar.gz",
            "/p/Folder (1)/file.tar.gz",
            "/p/Folder (10)/file.tar.gz",
            "/p/Folder/file.x1.9.tar.gz",
            "/p/Folder/file.x1.10.tar.gz",
        ],
    ),
    (  # same as test_natsorted_path_extensions_heuristic()
        [
            "Try.Me.Bug - 09 - One.Two.Three.[text].mkv",
            "Try.Me.Bug - 07 - One.Two.5.[text].mkv",
            "Try.Me.Bug - 08 - One.Two.Three[text].mkv",
        ],
        [
            "Try.Me.Bug - 07 - One.Two.5.[text].mkv",
            "Try.Me.Bug - 08 - One.Two.Three[text].mkv",
            "Try.Me.Bug - 09 - One.Two.Three.[text].mkv",
        ],
    ),
    (  # same as ns.IGNORECASE for test_natsorted_supports_case_handling()
        ["Apple", "corn", "Corn", "Banana", "apple", "banana"],
        ["Apple", "apple", "Banana", "banana", "corn", "Corn"],
    ),

]

for (given, expected) in sort_data:
    assert sorted(given, key=string_to_pairs) == expected

奖金

如果字符串混合了非ascii文本和数字，您可能会对将string_to_pairs()与我在其他地方给出的函数remove_diacritics()组合感兴趣。

下面是马克·拜尔回答的一个更加python化的版本:

import re

def natural_sort_key(s, _nsre=re.compile('([0-9]+)')):
    return [int(text) if text.isdigit() else text.lower()
            for text in _nsre.split(s)]

现在这个函数可以在任何使用它的函数中用作键，比如list。Sort, sorted, max，等等。

作为lambda:

lambda s: [int(t) if t.isdigit() else t.lower() for t in re.split('(\d+)', s)]

完全可重复的演示代码:

import re
natsort = lambda s: [int(t) if t.isdigit() else t.lower() for t in re.split('(\d+)', s)]
L = ["a1", "a10", "a11", "a2", "a22", "a3"]   
print(sorted(L, key=natsort))  
# ['a1', 'a2', 'a3', 'a10', 'a11', 'a22']