GH-127456: pathlib ABCs: add protocol for path parser (#127494)

Change the default value of `PurePathBase.parser` from `ParserBase()` to
`posixpath`. As a result, user subclasses of `PurePathBase` and `PathBase`
use POSIX path syntax by default, which is very often desirable.

Move `pathlib._abc.ParserBase` to `pathlib._types.Parser`, and convert it
to a runtime-checkable protocol.

Co-authored-by: Bénédikt Tran <10796600+picnixz@users.noreply.github.com>

2 files changed, 24 insertions, 54 deletions

diff --git a/Lib/pathlib/_abc.py b/Lib/pathlib/_abc.py
index 309eab2..f68685f 100644
--- a/Lib/pathlib/_abc.py
+++ b/Lib/pathlib/_abc.py
@@ -13,6 +13,7 @@ resemble pathlib's PurePath and Path respectively.
 
 import functools
 import operator
+import posixpath
 from errno import EINVAL
 from glob import _GlobberBase, _no_recurse_symlinks
 from stat import S_ISDIR, S_ISLNK, S_ISREG, S_ISSOCK, S_ISBLK, S_ISCHR, S_ISFIFO
@@ -33,59 +34,6 @@ def _is_case_sensitive(parser):
     return parser.normcase('Aa') == 'Aa'
 
 
-
-class ParserBase:
-    """Base class for path parsers, which do low-level path manipulation.
-
-    Path parsers provide a subset of the os.path API, specifically those
-    functions needed to provide PurePathBase functionality. Each PurePathBase
-    subclass references its path parser via a 'parser' class attribute.
-
-    Every method in this base class raises an UnsupportedOperation exception.
-    """
-
-    @classmethod
-    def _unsupported_msg(cls, attribute):
-        return f"{cls.__name__}.{attribute} is unsupported"
-
-    @property
-    def sep(self):
-        """The character used to separate path components."""
-        raise UnsupportedOperation(self._unsupported_msg('sep'))
-
-    def join(self, path, *paths):
-        """Join path segments."""
-        raise UnsupportedOperation(self._unsupported_msg('join()'))
-
-    def split(self, path):
-        """Split the path into a pair (head, tail), where *head* is everything
-        before the final path separator, and *tail* is everything after.
-        Either part may be empty.
-        """
-        raise UnsupportedOperation(self._unsupported_msg('split()'))
-
-    def splitdrive(self, path):
-        """Split the path into a 2-item tuple (drive, tail), where *drive* is
-        a device name or mount point, and *tail* is everything after the
-        drive. Either part may be empty."""
-        raise UnsupportedOperation(self._unsupported_msg('splitdrive()'))
-
-    def splitext(self, path):
-        """Split the path into a pair (root, ext), where *ext* is empty or
-        begins with a period and contains at most one period,
-        and *root* is everything before the extension."""
-        raise UnsupportedOperation(self._unsupported_msg('splitext()'))
-
-    def normcase(self, path):
-        """Normalize the case of the path."""
-        raise UnsupportedOperation(self._unsupported_msg('normcase()'))
-
-    def isabs(self, path):
-        """Returns whether the path is absolute, i.e. unaffected by the
-        current directory or drive."""
-        raise UnsupportedOperation(self._unsupported_msg('isabs()'))
-
-
 class PathGlobber(_GlobberBase):
     """
     Class providing shell-style globbing for path objects.
@@ -115,7 +63,7 @@ class PurePathBase:
         # the `__init__()` method.
         '_raw_paths',
     )
-    parser = ParserBase()
+    parser = posixpath
     _globber = PathGlobber
 
     def __init__(self, *args):
diff --git a/Lib/pathlib/_types.py b/Lib/pathlib/_types.py
new file mode 100644
index 0000000..60df94d
--- /dev/null
+++ b/Lib/pathlib/_types.py
@@ -0,0 +1,22 @@
+"""
+Protocols for supporting classes in pathlib.
+"""
+from typing import Protocol, runtime_checkable
+
+
+@runtime_checkable
+class Parser(Protocol):
+    """Protocol for path parsers, which do low-level path manipulation.
+
+    Path parsers provide a subset of the os.path API, specifically those
+    functions needed to provide PurePathBase functionality. Each PurePathBase
+    subclass references its path parser via a 'parser' class attribute.
+    """
+
+    sep: str
+    def join(self, path: str, *paths: str) -> str: ...
+    def split(self, path: str) -> tuple[str, str]: ...
+    def splitdrive(self, path: str) -> tuple[str, str]: ...
+    def splitext(self, path: str) -> tuple[str, str]: ...
+    def normcase(self, path: str) -> str: ...
+    def isabs(self, path: str) -> bool: ...