我可以使用哪个Python库从路径中提取文件名,而不管操作系统或路径格式是什么?

例如,我希望所有这些路径都返回c:

a/b/c/
a/b/c
\a\b\c
\a\b\c\
a\b\c
a/b/../../a/b/c/
a/b/../../a/b/c

当前回答

如果您的文件路径不是以“/”结尾,且目录以“/”分隔,则使用以下代码。众所周知,path通常不以“/”结尾。

import os
path_str = "/var/www/index.html"
print(os.path.basename(path_str))

但在某些情况下,像url以“/”结尾,然后使用以下代码

import os
path_str = "/home/some_str/last_str/"
split_path = path_str.rsplit("/",1)
print(os.path.basename(split_path[0]))

但是当你的路径被“\”分开时,你通常在Windows路径中找到,然后你可以使用以下代码

import os
path_str = "c:\\var\www\index.html"
print(os.path.basename(path_str))

import os
path_str = "c:\\home\some_str\last_str\\"
split_path = path_str.rsplit("\\",1)
print(os.path.basename(split_path[0]))

您可以通过检查操作系统类型将两者组合成一个函数并返回结果。

其他回答

os.path.split 这是你要找的函数吗

head, tail = os.path.split("/tmp/d/a.dat")

>>> print(tail)
a.dat
>>> print(head)
/tmp/d
import os
file_location = '/srv/volume1/data/eds/eds_report.csv'
file_name = os.path.basename(file_location )  #eds_report.csv
location = os.path.dirname(file_location )    #/srv/volume1/data/eds

我在Windows和Ubuntu (WSL)上使用此方法,它只使用“import os”即可工作(我): 基本上,replace()根据当前操作系统平台设置正确的路径分隔符。

如果路径以斜杠'/'结束,那么它不是一个文件而是一个目录,因此它返回一个空字符串。

import os

my_fullpath = r"D:\MY_FOLDER\TEST\20201108\20201108_073751.DNG"
os.path.basename(my_fullpath.replace('\\',os.sep))

my_fullpath = r"/MY_FOLDER/TEST/20201108/20201108_073751.DNG"
os.path.basename(my_fullpath.replace('\\',os.sep))

my_fullpath = r"/MY_FOLDER/TEST/20201108/"
os.path.basename(my_fullpath.replace('\\',os.sep))

my_fullpath = r"/MY_FOLDER/TEST/20201108"
os.path.basename(my_fullpath.replace('\\',os.sep))

在Windows(左)和Ubuntu(通过WSL,右)上:

这是一个仅适用于regex的解决方案,它似乎适用于任何OS上的任何OS路径。

不需要其他模块,也不需要预处理:

import re

def extract_basename(path):
  """Extracts basename of a given path. Should Work with any OS Path on any OS"""
  basename = re.search(r'[^\\/]+(?=[\\/]?$)', path)
  if basename:
    return basename.group(0)


paths = ['a/b/c/', 'a/b/c', '\\a\\b\\c', '\\a\\b\\c\\', 'a\\b\\c',
         'a/b/../../a/b/c/', 'a/b/../../a/b/c']

print([extract_basename(path) for path in paths])
# ['c', 'c', 'c', 'c', 'c', 'c', 'c']


extra_paths = ['C:\\', 'alone', '/a/space in filename', 'C:\\multi\nline']

print([extract_basename(path) for path in extra_paths])
# ['C:', 'alone', 'space in filename', 'multi\nline']

更新:

If you only want a potential filename, if present (i.e., /a/b/ is a dir and so is c:\windows\), change the regex to: r'[^\\/]+(?![\\/])$' . For the "regex challenged," this changes the positive forward lookahead for some sort of slash to a negative forward lookahead, causing pathnames that end with said slash to return nothing instead of the last sub-directory in the pathname. Of course there is no guarantee that the potential filename actually refers to a file and for that os.path.is_dir() or os.path.is_file() would need to be employed.

这将匹配如下:

/a/b/c/             # nothing, pathname ends with the dir 'c'
c:\windows\         # nothing, pathname ends with the dir 'windows'
c:hello.txt         # matches potential filename 'hello.txt'
~it_s_me/.bashrc    # matches potential filename '.bashrc'
c:\windows\system32 # matches potential filename 'system32', except
                    # that is obviously a dir. os.path.is_dir()
                    # should be used to tell us for sure

正则表达式可以在这里测试。

有一个函数返回你想要的结果

import os
print(os.path.basename(your_path))

警告:在POSIX系统上使用os.path.basename()从windows样式的路径(例如。"C:\\my\\file.txt"),则会返回整个路径。

下面的例子来自交互式python shell在Linux主机上运行:

Python 3.8.2 (default, Mar 13 2020, 10:14:16)
[GCC 9.3.0] on Linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import os
>>> filepath = "C:\\my\\path\\to\\file.txt" # A Windows style file path.
>>> os.path.basename(filepath)
'C:\\my\\path\\to\\file.txt'