A :class:`importlib.resources.abc.Traversable` interface for zip files.
Implements many of the features users enjoy from
:class:`pathlib.Path`.
Consider a zip file with this structure::
.
├── a.txt
└── b
├── c.txt
└── d
└── e.txt
>>> data = io.BytesIO()
>>> zf = ZipFile(data, 'w')
>>> zf.writestr('a.txt', 'content of a')
>>> zf.writestr('b/c.txt', 'content of c')
>>> zf.writestr('b/d/e.txt', 'content of e')
>>> zf.filename = 'mem/abcde.zip'
Path accepts the zipfile object itself or a filename
>>> path = Path(zf)
From there, several path operations are available.
Directory iteration (including the zip file itself):
>>> a, b = path.iterdir()
>>> a
Path('mem/abcde.zip', 'a.txt')
>>> b
Path('mem/abcde.zip', 'b/')
name property:
>>> b.name
'b'
join with divide operator:
>>> c = b / 'c.txt'
>>> c
Path('mem/abcde.zip', 'b/c.txt')
>>> c.name
'c.txt'
Read text:
>>> c.read_text(encoding='utf-8')
'content of c'
existence:
>>> c.exists()
True
>>> (b / 'missing.txt').exists()
False
Coercion to string:
>>> import os
>>> str(c).replace(os.sep, posixpath.sep)
'mem/abcde.zip/b/c.txt'
At the root, ``name``, ``filename``, and ``parent``
resolve to the zipfile.
>>> str(path)
'mem/abcde.zip'
>>> path.name
'abcde.zip'
>>> path.filename == pathlib.Path('mem/abcde.zip')
True
>>> str(path.parent)
'mem'
If the zipfile has no filename, such attributes are not
valid and accessing them will raise an Exception.
>>> zf.filename = None
>>> path.name
Traceback (most recent call last):
...
TypeError: ...
>>> path.filename
Traceback (most recent call last):
...
TypeError: ...
>>> path.parent
Traceback (most recent call last):
...
TypeError: ...
# workaround python/cpython#106763
>>> pass