traverseda
/
cpython
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Issue #15275: Clean up and speed up the ntpath module.

This commit is contained in:
Serhiy Storchaka 2014-07-23 20:43:13 +03:00
parent f886697659
commit 8518b79a8d
2 changed files with 79 additions and 93 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

170
Lib/ntpath.py
View File

@ -32,48 +32,12 @@ if 'ce' in sys.builtin_module_names:
defpath = '\\Windows' defpath = '\\Windows'
devnull = 'nul' devnull = 'nul'
def _get_empty(path):
if isinstance(path, bytes):
return b''
else:
return ''
def _get_sep(path):
if isinstance(path, bytes):
return b'\\'
else:
return '\\'
def _get_altsep(path):
if isinstance(path, bytes):
return b'/'
else:
return '/'
def _get_bothseps(path): def _get_bothseps(path):
if isinstance(path, bytes): if isinstance(path, bytes):
return b'\\/' return b'\\/'
else: else:
return '\\/' return '\\/'
def _get_dot(path):
if isinstance(path, bytes):
return b'.'
else:
return '.'
def _get_colon(path):
if isinstance(path, bytes):
return b':'
else:
return ':'
def _get_special(path):
if isinstance(path, bytes):
return (b'\\\\.\\', b'\\\\?\\')
else:
return ('\\\\.\\', '\\\\?\\')
# Normalize the case of a pathname and map slashes to backslashes. # Normalize the case of a pathname and map slashes to backslashes.
# Other normalizations (such as optimizing '../' away) are not done # Other normalizations (such as optimizing '../' away) are not done
# (this is done by normpath). # (this is done by normpath).
@ -82,10 +46,16 @@ def normcase(s):
"""Normalize case of pathname. """Normalize case of pathname.
Makes all characters lowercase and all slashes into backslashes.""" Makes all characters lowercase and all slashes into backslashes."""
if not isinstance(s, (bytes, str)): try:
raise TypeError("normcase() argument must be str or bytes, " if isinstance(s, bytes):
"not '{}'".format(s.__class__.__name__)) return s.replace(b'/', b'\\').lower()
return s.replace(_get_altsep(s), _get_sep(s)).lower() else:
return s.replace('/', '\\').lower()
except (TypeError, AttributeError):
if not isinstance(s, (bytes, str)):
raise TypeError("normcase() argument must be str or bytes, "
"not %r" % s.__class__.__name__) from None
raise
# Return whether a path is absolute. # Return whether a path is absolute.
@ -97,14 +67,19 @@ def normcase(s):
def isabs(s): def isabs(s):
"""Test whether a path is absolute""" """Test whether a path is absolute"""
s = splitdrive(s)[1] s = splitdrive(s)[1]
return len(s) > 0 and s[:1] in _get_bothseps(s) return len(s) > 0 and s[0] in _get_bothseps(s)
# Join two (or more) paths. # Join two (or more) paths.
def join(path, *paths): def join(path, *paths):
sep = _get_sep(path) if isinstance(path, bytes):
seps = _get_bothseps(path) sep = b'\\'
colon = _get_colon(path) seps = b'\\/'
colon = b':'
else:
sep = '\\'
seps = '\\/'
colon = ':'
result_drive, result_path = splitdrive(path) result_drive, result_path = splitdrive(path)
for p in paths: for p in paths:
p_drive, p_path = splitdrive(p) p_drive, p_path = splitdrive(p)
@ -155,10 +130,16 @@ def splitdrive(p):
Paths cannot contain both a drive letter and a UNC path. Paths cannot contain both a drive letter and a UNC path.
""" """
empty = _get_empty(p) if len(p) >= 2:
if len(p) > 1: if isinstance(p, bytes):
sep = _get_sep(p) sep = b'\\'
normp = p.replace(_get_altsep(p), sep) altsep = b'/'
colon = b':'
else:
sep = '\\'
altsep = '/'
colon = ':'
normp = p.replace(altsep, sep)
if (normp[0:2] == sep*2) and (normp[2:3] != sep): if (normp[0:2] == sep*2) and (normp[2:3] != sep):
# is a UNC path: # is a UNC path:
# vvvvvvvvvvvvvvvvvvvv drive letter or UNC path # vvvvvvvvvvvvvvvvvvvv drive letter or UNC path
@ -166,18 +147,18 @@ def splitdrive(p):
# directory ^^^^^^^^^^^^^^^ # directory ^^^^^^^^^^^^^^^
index = normp.find(sep, 2) index = normp.find(sep, 2)
if index == -1: if index == -1:
return empty, p return p[:0], p
index2 = normp.find(sep, index + 1) index2 = normp.find(sep, index + 1)
# a UNC path can't have two slashes in a row # a UNC path can't have two slashes in a row
# (after the initial two) # (after the initial two)
if index2 == index + 1: if index2 == index + 1:
return empty, p return p[:0], p
if index2 == -1: if index2 == -1:
index2 = len(p) index2 = len(p)
return p[:index2], p[index2:] return p[:index2], p[index2:]
if normp[1:2] == _get_colon(p): if normp[1:2] == colon:
return p[:2], p[2:] return p[:2], p[2:]
return empty, p return p[:0], p
# Parse UNC paths # Parse UNC paths
@ -221,10 +202,7 @@ def split(p):
i -= 1 i -= 1
head, tail = p[:i], p[i:] # now tail has no slashes head, tail = p[:i], p[i:] # now tail has no slashes
# remove trailing slashes from head, unless it's all slashes # remove trailing slashes from head, unless it's all slashes
head2 = head head = head.rstrip(seps) or head
while head2 and head2[-1:] in seps:
head2 = head2[:-1]
head = head2 or head
return d + head, tail return d + head, tail
@ -234,8 +212,10 @@ def split(p):
# It is always true that root + ext == p. # It is always true that root + ext == p.
def splitext(p): def splitext(p):
return genericpath._splitext(p, _get_sep(p), _get_altsep(p), if isinstance(p, bytes):
_get_dot(p)) return genericpath._splitext(p, b'\\', b'/', b'.')
else:
return genericpath._splitext(p, '\\', '/', '.')
splitext.__doc__ = genericpath._splitext.__doc__ splitext.__doc__ = genericpath._splitext.__doc__
@ -343,7 +323,7 @@ def expanduser(path):
userhome = join(drive, os.environ['HOMEPATH']) userhome = join(drive, os.environ['HOMEPATH'])
if isinstance(path, bytes): if isinstance(path, bytes):
userhome = userhome.encode(sys.getfilesystemencoding()) userhome = os.fsencode(userhome)
if i != 1: #~user if i != 1: #~user
userhome = join(dirname(userhome), path[1:i]) userhome = join(dirname(userhome), path[1:i])
@ -369,13 +349,14 @@ def expandvars(path):
Unknown variables are left unchanged.""" Unknown variables are left unchanged."""
if isinstance(path, bytes): if isinstance(path, bytes):
if ord('$') not in path and ord('%') not in path: if b'$' not in path and b'%' not in path:
return path return path
import string import string
varchars = bytes(string.ascii_letters + string.digits + '_-', 'ascii') varchars = bytes(string.ascii_letters + string.digits + '_-', 'ascii')
quote = b'\'' quote = b'\''
percent = b'%' percent = b'%'
brace = b'{' brace = b'{'
rbrace = b'}'
dollar = b'$' dollar = b'$'
environ = getattr(os, 'environb', None) environ = getattr(os, 'environb', None)
else: else:
@ -386,6 +367,7 @@ def expandvars(path):
quote = '\'' quote = '\''
percent = '%' percent = '%'
brace = '{' brace = '{'
rbrace = '}'
dollar = '$' dollar = '$'
environ = os.environ environ = os.environ
res = path[:0] res = path[:0]
@ -432,15 +414,9 @@ def expandvars(path):
path = path[index+2:] path = path[index+2:]
pathlen = len(path) pathlen = len(path)
try: try:
if isinstance(path, bytes): index = path.index(rbrace)
index = path.index(b'}')
else:
index = path.index('}')
except ValueError: except ValueError:
if isinstance(path, bytes): res += dollar + brace + path
res += b'${' + path
else:
res += '${' + path
index = pathlen - 1 index = pathlen - 1
else: else:
var = path[:index] var = path[:index]
@ -450,10 +426,7 @@ def expandvars(path):
else: else:
value = environ[var] value = environ[var]
except KeyError: except KeyError:
if isinstance(path, bytes): value = dollar + brace + var + rbrace
value = b'${' + var + b'}'
else:
value = '${' + var + '}'
res += value res += value
else: else:
var = path[:0] var = path[:0]
@ -485,16 +458,25 @@ def expandvars(path):
def normpath(path): def normpath(path):
"""Normalize path, eliminating double slashes, etc.""" """Normalize path, eliminating double slashes, etc."""
sep = _get_sep(path) if isinstance(path, bytes):
dotdot = _get_dot(path) * 2 sep = b'\\'
special_prefixes = _get_special(path) altsep = b'/'
curdir = b'.'
pardir = b'..'
special_prefixes = (b'\\\\.\\', b'\\\\?\\')
else:
sep = '\\'
altsep = '/'
curdir = '.'
pardir = '..'
special_prefixes = ('\\\\.\\', '\\\\?\\')
if path.startswith(special_prefixes): if path.startswith(special_prefixes):
# in the case of paths with these prefixes: # in the case of paths with these prefixes:
# \\.\ -> device names # \\.\ -> device names
# \\?\ -> literal paths # \\?\ -> literal paths
# do not do any normalization, but return the path unchanged # do not do any normalization, but return the path unchanged
return path return path
path = path.replace(_get_altsep(path), sep) path = path.replace(altsep, sep)
prefix, path = splitdrive(path) prefix, path = splitdrive(path)
# collapse initial backslashes # collapse initial backslashes
@ -505,13 +487,13 @@ def normpath(path):
comps = path.split(sep) comps = path.split(sep)
i = 0 i = 0
while i < len(comps): while i < len(comps):
if not comps[i] or comps[i] == _get_dot(path): if not comps[i] or comps[i] == curdir:
del comps[i] del comps[i]
elif comps[i] == dotdot: elif comps[i] == pardir:
if i > 0 and comps[i-1] != dotdot: if i > 0 and comps[i-1] != pardir:
del comps[i-1:i+1] del comps[i-1:i+1]
i -= 1 i -= 1
elif i == 0 and prefix.endswith(_get_sep(path)): elif i == 0 and prefix.endswith(sep):
del comps[i] del comps[i]
else: else:
i += 1 i += 1
@ -519,7 +501,7 @@ def normpath(path):
i += 1 i += 1
# If the path is now empty, substitute '.' # If the path is now empty, substitute '.'
if not prefix and not comps: if not prefix and not comps:
comps.append(_get_dot(path)) comps.append(curdir)
return prefix + sep.join(comps) return prefix + sep.join(comps)
@ -559,12 +541,19 @@ realpath = abspath
supports_unicode_filenames = (hasattr(sys, "getwindowsversion") and supports_unicode_filenames = (hasattr(sys, "getwindowsversion") and
sys.getwindowsversion()[3] >= 2) sys.getwindowsversion()[3] >= 2)
def relpath(path, start=curdir): def relpath(path, start=None):
"""Return a relative version of a path""" """Return a relative version of a path"""
sep = _get_sep(path) if isinstance(path, bytes):
sep = b'\\'
curdir = b'.'
pardir = b'..'
else:
sep = '\\'
curdir = '.'
pardir = '..'
if start is curdir: if start is None:
start = _get_dot(path) start = curdir
if not path: if not path:
raise ValueError("no path specified") raise ValueError("no path specified")
@ -574,9 +563,8 @@ def relpath(path, start=curdir):
start_drive, start_rest = splitdrive(start_abs) start_drive, start_rest = splitdrive(start_abs)
path_drive, path_rest = splitdrive(path_abs) path_drive, path_rest = splitdrive(path_abs)
if normcase(start_drive) != normcase(path_drive): if normcase(start_drive) != normcase(path_drive):
error = "path is on mount '{0}', start on mount '{1}'".format( raise ValueError("path is on mount %r, start on mount %r" % (
path_drive, start_drive) path_drive, start_drive))
raise ValueError(error)
start_list = [x for x in start_rest.split(sep) if x] start_list = [x for x in start_rest.split(sep) if x]
path_list = [x for x in path_rest.split(sep) if x] path_list = [x for x in path_rest.split(sep) if x]
@ -587,13 +575,9 @@ def relpath(path, start=curdir):
break break
i += 1 i += 1
if isinstance(path, bytes):
pardir = b'..'
else:
pardir = '..'
rel_list = [pardir] * (len(start_list)-i) + path_list[i:] rel_list = [pardir] * (len(start_list)-i) + path_list[i:]
if not rel_list: if not rel_list:
return _get_dot(path) return curdir
return join(*rel_list) return join(*rel_list)

2
Misc/NEWS
View File

@ -108,6 +108,8 @@ Core and Builtins
Library Library
------- -------
- Issue #15275: Clean up and speed up the ntpath module.
- Issue #21888: plistlib's load() and loads() now work if the fmt parameter is - Issue #21888: plistlib's load() and loads() now work if the fmt parameter is
specified. specified.