blob: 996080175b0b651582be415d4fc04de4a7322b91 [file] [log] [blame]
# Module 'posixpath' -- common operations on Posix pathnames.
# Some of this can actually be useful on non-Posix systems too, e.g.
# for manipulation of the pathname component of URLs.
# The "os.path" name is an alias for this module on Posix systems;
# on other systems (e.g. Mac, Windows), os.path provides the same
# operations in a manner specific to that platform, and is an alias
# to another module (e.g. macpath, ntpath).
import os
import stat
# Normalize the case of a pathname. Trivial in Posix, string.lower on Mac.
# On MS-DOS this may also turn slashes into backslashes; however, other
# normalizations (such as optimizing '../' away) are not allowed
# (another function should be defined to do that).
def normcase(s):
return s
# Return wheter a path is absolute.
# Trivial in Posix, harder on the Mac or MS-DOS.
def isabs(s):
return s[:1] == '/'
# Join pathnames.
# Ignore the previous parts if a part is absolute.
# Insert a '/' unless the first part is empty or already ends in '/'.
def join(a, *p):
path = a
for b in p:
if b[:1] == '/':
path = b
elif path == '' or path[-1:] == '/':
path = path + b
else:
path = path + '/' + b
return path
# Split a path in head (everything up to the last '/') and tail (the
# rest). If the path ends in '/', tail will be empty. If there is no
# '/' in the path, head will be empty.
# Trailing '/'es are stripped from head unless it is the root.
def split(p):
import string
i = string.rfind(p, '/') + 1
head, tail = p[:i], p[i:]
if head and head <> '/'*len(head):
while head[-1] == '/':
head = head[:-1]
return head, tail
# Split a path in root and extension.
# The extension is everything starting at the last dot in the last
# pathname component; the root is everything before that.
# It is always true that root + ext == p.
def splitext(p):
root, ext = '', ''
for c in p:
if c == '/':
root, ext = root + ext + c, ''
elif c == '.':
if ext:
root, ext = root + ext, c
else:
ext = c
elif ext:
ext = ext + c
else:
root = root + c
return root, ext
# Split a pathname into a drive specification and the rest of the
# path. Useful on DOS/Windows/NT; on Unix, the drive is always empty.
def splitdrive(p):
return '', p
# Return the tail (basename) part of a path.
def basename(p):
return split(p)[1]
# Return the head (dirname) part of a path.
def dirname(p):
return split(p)[0]
# Return the longest prefix of all list elements.
def commonprefix(m):
if not m: return ''
prefix = m[0]
for item in m:
for i in range(len(prefix)):
if prefix[:i+1] <> item[:i+1]:
prefix = prefix[:i]
if i == 0: return ''
break
return prefix
# Is a path a symbolic link?
# This will always return false on systems where os.lstat doesn't exist.
def islink(path):
try:
st = os.lstat(path)
except (os.error, AttributeError):
return 0
return stat.S_ISLNK(st[stat.ST_MODE])
# Does a path exist?
# This is false for dangling symbolic links.
def exists(path):
try:
st = os.stat(path)
except os.error:
return 0
return 1
# Is a path a directory?
# This follows symbolic links, so both islink() and isdir() can be true
# for the same path.
def isdir(path):
try:
st = os.stat(path)
except os.error:
return 0
return stat.S_ISDIR(st[stat.ST_MODE])
# Is a path a regular file?
# This follows symbolic links, so both islink() and isfile() can be true
# for the same path.
def isfile(path):
try:
st = os.stat(path)
except os.error:
return 0
return stat.S_ISREG(st[stat.ST_MODE])
# Are two filenames really pointing to the same file?
def samefile(f1, f2):
s1 = os.stat(f1)
s2 = os.stat(f2)
return samestat(s1, s2)
# Are two open files really referencing the same file?
# (Not necessarily the same file descriptor!)
def sameopenfile(fp1, fp2):
s1 = os.fstat(fp1)
s2 = os.fstat(fp2)
return samestat(s1, s2)
# Are two stat buffers (obtained from stat, fstat or lstat)
# describing the same file?
def samestat(s1, s2):
return s1[stat.ST_INO] == s2[stat.ST_INO] and \
s1[stat.ST_DEV] == s2[stat.ST_DEV]
# Is a path a mount point?
# (Does this work for all UNIXes? Is it even guaranteed to work by Posix?)
def ismount(path):
try:
s1 = os.stat(path)
s2 = os.stat(join(path, '..'))
except os.error:
return 0 # It doesn't exist -- so not a mount point :-)
dev1 = s1[stat.ST_DEV]
dev2 = s2[stat.ST_DEV]
if dev1 != dev2:
return 1 # path/.. on a different device as path
ino1 = s1[stat.ST_INO]
ino2 = s2[stat.ST_INO]
if ino1 == ino2:
return 1 # path/.. is the same i-node as path
return 0
# Directory tree walk.
# For each directory under top (including top itself, but excluding
# '.' and '..'), func(arg, dirname, filenames) is called, where
# dirname is the name of the directory and filenames is the list
# files files (and subdirectories etc.) in the directory.
# The func may modify the filenames list, to implement a filter,
# or to impose a different order of visiting.
def walk(top, func, arg):
try:
names = os.listdir(top)
except os.error:
return
func(arg, top, names)
exceptions = ('.', '..')
for name in names:
if name not in exceptions:
name = join(top, name)
if isdir(name) and not islink(name):
walk(name, func, arg)
# Expand paths beginning with '~' or '~user'.
# '~' means $HOME; '~user' means that user's home directory.
# If the path doesn't begin with '~', or if the user or $HOME is unknown,
# the path is returned unchanged (leaving error reporting to whatever
# function is called with the expanded path as argument).
# See also module 'glob' for expansion of *, ? and [...] in pathnames.
# (A function should also be defined to do full *sh-style environment
# variable expansion.)
def expanduser(path):
if path[:1] <> '~':
return path
i, n = 1, len(path)
while i < n and path[i] <> '/':
i = i+1
if i == 1:
if not os.environ.has_key('HOME'):
return path
userhome = os.environ['HOME']
else:
import pwd
try:
pwent = pwd.getpwnam(path[1:i])
except KeyError:
return path
userhome = pwent[5]
if userhome[-1:] == '/': i = i+1
return userhome + path[i:]
# Expand paths containing shell variable substitutions.
# This expands the forms $variable and ${variable} only.
# Non-existant variables are left unchanged.
_varprog = None
def expandvars(path):
global _varprog
if '$' not in path:
return path
if not _varprog:
import re
_varprog = re.compile(r'\$(\w+|\{[^}]*\})')
i = 0
while 1:
m = _varprog.search(path, i)
if not m:
break
i, j = m.span(0)
name = m.group(1)
if name[:1] == '{' and name[-1:] == '}':
name = name[1:-1]
if os.environ.has_key(name):
tail = path[j:]
path = path[:i] + os.environ[name]
i = len(path)
path = path + tail
else:
i = j
return path
# Normalize a path, e.g. A//B, A/./B and A/foo/../B all become A/B.
# It should be understood that this may change the meaning of the path
# if it contains symbolic links!
def normpath(path):
import string
# Treat initial slashes specially
slashes = ''
while path[:1] == '/':
slashes = slashes + '/'
path = path[1:]
comps = string.splitfields(path, '/')
i = 0
while i < len(comps):
if comps[i] == '.':
del comps[i]
elif comps[i] == '..' and i > 0 and \
comps[i-1] not in ('', '..'):
del comps[i-1:i+1]
i = i-1
elif comps[i] == '' and i > 0 and comps[i-1] <> '':
del comps[i]
else:
i = i+1
# If the path is now empty, substitute '.'
if not comps and not slashes:
comps.append('.')
return slashes + string.joinfields(comps, '/')