gh-102511: Speed up os.path.splitroot() with native helpers (GH-118089)

This commit is contained in:
Nice Zombies 2024-04-25 11:07:38 +02:00 committed by GitHub
parent e38b43c213
commit 10bb90ed49
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
8 changed files with 337 additions and 108 deletions

View file

@ -167,56 +167,76 @@ def splitdrive(p):
return drive, root + tail
def splitroot(p):
"""Split a pathname into drive, root and tail. The drive is defined
exactly as in splitdrive(). On Windows, the root may be a single path
separator or an empty string. The tail contains anything after the root.
For example:
try:
from nt import _path_splitroot_ex
except ImportError:
def splitroot(p):
"""Split a pathname into drive, root and tail. The drive is defined
exactly as in splitdrive(). On Windows, the root may be a single path
separator or an empty string. The tail contains anything after the root.
For example:
splitroot('//server/share/') == ('//server/share', '/', '')
splitroot('C:/Users/Barney') == ('C:', '/', 'Users/Barney')
splitroot('C:///spam///ham') == ('C:', '/', '//spam///ham')
splitroot('Windows/notepad') == ('', '', 'Windows/notepad')
"""
p = os.fspath(p)
if isinstance(p, bytes):
sep = b'\\'
altsep = b'/'
colon = b':'
unc_prefix = b'\\\\?\\UNC\\'
empty = b''
else:
sep = '\\'
altsep = '/'
colon = ':'
unc_prefix = '\\\\?\\UNC\\'
empty = ''
normp = p.replace(altsep, sep)
if normp[:1] == sep:
if normp[1:2] == sep:
# UNC drives, e.g. \\server\share or \\?\UNC\server\share
# Device drives, e.g. \\.\device or \\?\device
start = 8 if normp[:8].upper() == unc_prefix else 2
index = normp.find(sep, start)
if index == -1:
return p, empty, empty
index2 = normp.find(sep, index + 1)
if index2 == -1:
return p, empty, empty
return p[:index2], p[index2:index2 + 1], p[index2 + 1:]
splitroot('//server/share/') == ('//server/share', '/', '')
splitroot('C:/Users/Barney') == ('C:', '/', 'Users/Barney')
splitroot('C:///spam///ham') == ('C:', '/', '//spam///ham')
splitroot('Windows/notepad') == ('', '', 'Windows/notepad')
"""
p = os.fspath(p)
if isinstance(p, bytes):
sep = b'\\'
altsep = b'/'
colon = b':'
unc_prefix = b'\\\\?\\UNC\\'
empty = b''
else:
# Relative path with root, e.g. \Windows
return empty, p[:1], p[1:]
elif normp[1:2] == colon:
if normp[2:3] == sep:
# Absolute drive-letter path, e.g. X:\Windows
return p[:2], p[2:3], p[3:]
sep = '\\'
altsep = '/'
colon = ':'
unc_prefix = '\\\\?\\UNC\\'
empty = ''
normp = p.replace(altsep, sep)
if normp[:1] == sep:
if normp[1:2] == sep:
# UNC drives, e.g. \\server\share or \\?\UNC\server\share
# Device drives, e.g. \\.\device or \\?\device
start = 8 if normp[:8].upper() == unc_prefix else 2
index = normp.find(sep, start)
if index == -1:
return p, empty, empty
index2 = normp.find(sep, index + 1)
if index2 == -1:
return p, empty, empty
return p[:index2], p[index2:index2 + 1], p[index2 + 1:]
else:
# Relative path with root, e.g. \Windows
return empty, p[:1], p[1:]
elif normp[1:2] == colon:
if normp[2:3] == sep:
# Absolute drive-letter path, e.g. X:\Windows
return p[:2], p[2:3], p[3:]
else:
# Relative path with drive, e.g. X:Windows
return p[:2], empty, p[2:]
else:
# Relative path with drive, e.g. X:Windows
return p[:2], empty, p[2:]
else:
# Relative path, e.g. Windows
return empty, empty, p
# Relative path, e.g. Windows
return empty, empty, p
else:
def splitroot(p):
"""Split a pathname into drive, root and tail. The drive is defined
exactly as in splitdrive(). On Windows, the root may be a single path
separator or an empty string. The tail contains anything after the root.
For example:
splitroot('//server/share/') == ('//server/share', '/', '')
splitroot('C:/Users/Barney') == ('C:', '/', 'Users/Barney')
splitroot('C:///spam///ham') == ('C:', '/', '//spam///ham')
splitroot('Windows/notepad') == ('', '', 'Windows/notepad')
"""
p = os.fspath(p)
if isinstance(p, bytes):
drive, root, tail = _path_splitroot_ex(os.fsdecode(p))
return os.fsencode(drive), os.fsencode(root), os.fsencode(tail)
return _path_splitroot_ex(p)
# Split a path in head (everything up to the last '/') and tail (the