GH-127807: pathlib ABCs: remove a few private attributes (#127851)

From `PurePathBase` delete `_globber`, `_stack` and `_pattern_str`, and
from `PathBase` delete `_glob_selector`. This helps avoid an unpleasant
surprise for a users who try to use these names.

Author: barneygale

Lib/pathlib/_abc.py

@@ -25,6 +25,23 @@ def _is_case_sensitive(parser):
     return parser.normcase('Aa') == 'Aa'
 
 
+def _explode_path(path):
+    """
+    Split the path into a 2-tuple (anchor, parts), where *anchor* is the
+    uppermost parent of the path (equivalent to path.parents[-1]), and
+    *parts* is a reversed list of parts following the anchor.
+    """
+    split = path.parser.split
+    path = str(path)
+    parent, name = split(path)
+    names = []
+    while path != parent:
+        names.append(name)
+        path = parent
+        parent, name = split(path)
+    return path, names
+
+
 class PathGlobber(_GlobberBase):
     """
     Class providing shell-style globbing for path objects.
@@ -50,7 +67,6 @@ class PurePathBase:
 
     __slots__ = ()
     parser = posixpath
-    _globber = PathGlobber
 
     def with_segments(self, *pathsegments):
         """Construct a new path object from any number of path-like objects.
@@ -82,7 +98,7 @@ def root(self):
     @property
     def anchor(self):
         """The concatenation of the drive and root, or ''."""
-        return self._stack[0]
+        return _explode_path(self)[0]
 
     @property
     def name(self):
@@ -160,8 +176,8 @@ def relative_to(self, other, *, walk_up=False):
         """
         if not isinstance(other, PurePathBase):
             other = self.with_segments(other)
-        anchor0, parts0 = self._stack
-        anchor1, parts1 = other._stack
+        anchor0, parts0 = _explode_path(self)
+        anchor1, parts1 = _explode_path(other)
         if anchor0 != anchor1:
             raise ValueError(f"{str(self)!r} and {str(other)!r} have different anchors")
         while parts0 and parts1 and parts0[-1] == parts1[-1]:
@@ -183,8 +199,8 @@ def is_relative_to(self, other):
         """
         if not isinstance(other, PurePathBase):
             other = self.with_segments(other)
-        anchor0, parts0 = self._stack
-        anchor1, parts1 = other._stack
+        anchor0, parts0 = _explode_path(self)
+        anchor1, parts1 = _explode_path(other)
         if anchor0 != anchor1:
             return False
         while parts0 and parts1 and parts0[-1] == parts1[-1]:
@@ -199,7 +215,7 @@ def is_relative_to(self, other):
     def parts(self):
         """An object providing sequence-like access to the
         components in the filesystem path."""
-        anchor, parts = self._stack
+        anchor, parts = _explode_path(self)
         if anchor:
             parts.append(anchor)
         return tuple(reversed(parts))
@@ -224,23 +240,6 @@ def __rtruediv__(self, key):
         except TypeError:
             return NotImplemented
 
-    @property
-    def _stack(self):
-        """
-        Split the path into a 2-tuple (anchor, parts), where *anchor* is the
-        uppermost parent of the path (equivalent to path.parents[-1]), and
-        *parts* is a reversed list of parts following the anchor.
-        """
-        split = self.parser.split
-        path = str(self)
-        parent, name = split(path)
-        names = []
-        while path != parent:
-            names.append(name)
-            path = parent
-            parent, name = split(path)
-        return path, names
-
     @property
     def parent(self):
         """The logical parent of the path."""
@@ -268,11 +267,6 @@ def is_absolute(self):
         a drive)."""
         return self.parser.isabs(str(self))
 
-    @property
-    def _pattern_str(self):
-        """The path expressed as a string, for use in pattern-matching."""
-        return str(self)
-
     def match(self, path_pattern, *, case_sensitive=None):
         """
         Return True if this path matches the given pattern. If the pattern is
@@ -293,7 +287,7 @@ def match(self, path_pattern, *, case_sensitive=None):
             return False
         if len(path_parts) > len(pattern_parts) and path_pattern.anchor:
             return False
-        globber = self._globber(sep, case_sensitive)
+        globber = PathGlobber(sep, case_sensitive)
         for path_part, pattern_part in zip(path_parts, pattern_parts):
             match = globber.compile(pattern_part)
             if match(path_part) is None:
@@ -309,9 +303,9 @@ def full_match(self, pattern, *, case_sensitive=None):
             pattern = self.with_segments(pattern)
         if case_sensitive is None:
             case_sensitive = _is_case_sensitive(self.parser)
-        globber = self._globber(pattern.parser.sep, case_sensitive, recursive=True)
-        match = globber.compile(pattern._pattern_str)
-        return match(self._pattern_str) is not None
+        globber = PathGlobber(pattern.parser.sep, case_sensitive, recursive=True)
+        match = globber.compile(str(pattern))
+        return match(str(self)) is not None
 
 
 
@@ -463,29 +457,25 @@ def iterdir(self):
         """
         raise NotImplementedError
 
-    def _glob_selector(self, parts, case_sensitive, recurse_symlinks):
-        if case_sensitive is None:
-            case_sensitive = _is_case_sensitive(self.parser)
-            case_pedantic = False
-        else:
-            # The user has expressed a case sensitivity choice, but we don't
-            # know the case sensitivity of the underlying filesystem, so we
-            # must use scandir() for everything, including non-wildcard parts.
-            case_pedantic = True
-        recursive = True if recurse_symlinks else _no_recurse_symlinks
-        globber = self._globber(self.parser.sep, case_sensitive, case_pedantic, recursive)
-        return globber.selector(parts)
-
     def glob(self, pattern, *, case_sensitive=None, recurse_symlinks=True):
         """Iterate over this subtree and yield all existing files (of any
         kind, including directories) matching the given relative pattern.
         """
         if not isinstance(pattern, PurePathBase):
             pattern = self.with_segments(pattern)
-        anchor, parts = pattern._stack
+        anchor, parts = _explode_path(pattern)
         if anchor:
             raise NotImplementedError("Non-relative patterns are unsupported")
-        select = self._glob_selector(parts, case_sensitive, recurse_symlinks)
+        if case_sensitive is None:
+            case_sensitive = _is_case_sensitive(self.parser)
+            case_pedantic = False
+        elif case_sensitive == _is_case_sensitive(self.parser):
+            case_pedantic = False
+        else:
+            case_pedantic = True
+        recursive = True if recurse_symlinks else _no_recurse_symlinks
+        globber = PathGlobber(self.parser.sep, case_sensitive, case_pedantic, recursive)
+        select = globber.selector(parts)
         return select(self)
 
     def rglob(self, pattern, *, case_sensitive=None, recurse_symlinks=True):

Lib/pathlib/_local.py

@@ -5,7 +5,7 @@
 import posixpath
 import sys
 from errno import EINVAL, EXDEV
-from glob import _StringGlobber
+from glob import _StringGlobber, _no_recurse_symlinks
 from itertools import chain
 from stat import S_ISSOCK, S_ISBLK, S_ISCHR, S_ISFIFO
 from _collections_abc import Sequence
@@ -112,7 +112,6 @@ class PurePath(PurePathBase):
         '_hash',
     )
     parser = os.path
-    _globber = _StringGlobber
 
     def __new__(cls, *args, **kwargs):
         """Construct a PurePath from one or several strings and or existing
@@ -513,13 +512,22 @@ def as_uri(self):
         from urllib.parse import quote_from_bytes
         return prefix + quote_from_bytes(os.fsencode(path))
 
-    @property
-    def _pattern_str(self):
-        """The path expressed as a string, for use in pattern-matching."""
+    def full_match(self, pattern, *, case_sensitive=None):
+        """
+        Return True if this path matches the given glob-style pattern. The
+        pattern is matched against the entire path.
+        """
+        if not isinstance(pattern, PurePathBase):
+            pattern = self.with_segments(pattern)
+        if case_sensitive is None:
+            case_sensitive = self.parser is posixpath
+
         # The string representation of an empty path is a single dot ('.'). Empty
         # paths shouldn't match wildcards, so we change it to the empty string.
-        path_str = str(self)
-        return '' if path_str == '.' else path_str
+        path = str(self) if self.parts else ''
+        pattern = str(pattern) if pattern.parts else ''
+        globber = _StringGlobber(self.parser.sep, case_sensitive, recursive=True)
+        return globber.compile(pattern)(path) is not None
 
 # Subclassing os.PathLike makes isinstance() checks slower,
 # which in turn makes Path construction slower. Register instead!
@@ -749,8 +757,18 @@ def glob(self, pattern, *, case_sensitive=None, recurse_symlinks=False):
         kind, including directories) matching the given relative pattern.
         """
         sys.audit("pathlib.Path.glob", self, pattern)
+        if case_sensitive is None:
+            case_sensitive = self.parser is posixpath
+            case_pedantic = False
+        else:
+            # The user has expressed a case sensitivity choice, but we don't
+            # know the case sensitivity of the underlying filesystem, so we
+            # must use scandir() for everything, including non-wildcard parts.
+            case_pedantic = True
         parts = self._parse_pattern(pattern)
-        select = self._glob_selector(parts[::-1], case_sensitive, recurse_symlinks)
+        recursive = True if recurse_symlinks else _no_recurse_symlinks
+        globber = _StringGlobber(self.parser.sep, case_sensitive, case_pedantic, recursive)
+        select = globber.selector(parts[::-1])
         root = str(self)
         paths = select(root)