combined patches from http://bugs.python.org/issue3187

Lib/posixpath.py

 """Common operations on Posix pathnames.

[GvR, 2008/09/30 18:10:41]

Also note that os.path.dirname() and os.path.basename(), which sound like thin
wrappers around os.path.split(), are somehow not implemented that way and
currently fail; I think these should be fixed too (maybe by deferring to
os.path.split()).

 
 Instead of importing this module directly, import os and refer to
 this module as os.path.  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).
 
 Some of this can actually be useful on non-Posix systems too, e.g.
 for manipulation of the pathname component of URLs.
 """
 
 import os
+import sys
 import stat
 import genericpath
 from genericpath import *
 
 __all__ = ["normcase","isabs","join","splitdrive","split","splitext",
            "basename","dirname","commonprefix","getsize","getmtime",
            "getatime","getctime","islink","exists","lexists","isdir","isfile",
            "ismount", "expanduser","expandvars","normpath","abspath",
            "samefile","sameopenfile","samestat",
            "curdir","pardir","sep","pathsep","defpath","altsep","extsep",
            "devnull","realpath","supports_unicode_filenames","relpath"]
 
-# strings representing various path-related bits and pieces
+# Strings representing various path-related bits and pieces.
+# These are primarily for export; internally, they are hardcoded.
 curdir = '.'
 pardir = '..'
 extsep = '.'
 sep = '/'
 pathsep = ':'
 defpath = ':/bin:/usr/bin'
 altsep = None
 devnull = '/dev/null'
 
+def _get_sep(path):
+    if isinstance(path, bytes):
+        return b'/'
+    else:
+        return '/'
+
 # 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):
     """Normalize case of pathname.  Has no effect under Posix"""
+    # TODO: on Mac OS X, this should really return s.lower().
     return s
 
 
 # Return whether a path is absolute.
 # Trivial in Posix, harder on the Mac or MS-DOS.
 
 def isabs(s):
     """Test whether a path is absolute"""
-    return s.startswith('/')
+    sep = _get_sep(s)
+    return s.startswith(sep)
 
 
 # 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):
     """Join two or more pathname components, inserting '/' as needed.
     If any component is an absolute path, all previous path components
     will be discarded."""
-    if isinstance(a, bytes):
-        sep = b'/'
-    else:
-        sep = '/'
+    sep = _get_sep(a)
     path = a
     for b in p:
         if b.startswith(sep):
             path = b
         elif not path or path.endswith(sep):
             path +=  b
         else:
             path += sep + 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):
     """Split a pathname.  Returns tuple "(head, tail)" where "tail" is
     everything after the final slash.  Either part may be empty."""
-    if isinstance(p, bytes):
-        sep = b'/'
-    else:
-        sep = '/'
+    sep = _get_sep(p)
     i = p.rfind(sep) + 1
     head, tail = p[:i], p[i:]
     if head and head != sep*len(head):
         head = head.rstrip(sep)
     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):
-    return genericpath._splitext(p, sep, altsep, extsep)
+    if isinstance(p, bytes):
+        sep = b'/'
+        extsep = b'.'
+    else:
+        sep = '/'
+        extsep = '.'
+    return genericpath._splitext(p, sep, None, extsep)
 splitext.__doc__ = genericpath._splitext.__doc__
 
 # 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):
     """Split a pathname into drive and path. On Posix, drive is always
     empty."""
-    return '', p
+    return p[:0], p
 
 
 # Return the tail (basename) part of a path, same as split(path)[1].
 
 def basename(p):
     """Returns the final component of a pathname"""
-    i = p.rfind('/') + 1
+    sep = _get_sep(p)
+    i = p.rfind(sep) + 1
     return p[i:]
 
 
 # Return the head (dirname) part of a path, same as split(path)[0].
 
 def dirname(p):
     """Returns the directory component of a pathname"""
-    i = p.rfind('/') + 1
+    sep = _get_sep(p)
+    i = p.rfind(sep) + 1
     head = p[:i]
-    if head and head != '/'*len(head):
-        head = head.rstrip('/')
+    if head and head != sep*len(head):
+        head = head.rstrip(sep)
     return head
 
 
 # Is a path a symbolic link?
 # This will always return false on systems where os.lstat doesn't exist.
 
 def islink(path):
     """Test whether a path is a symbolic link"""
     try:
         st = os.lstat(path)
@@ skipping 40 matching lines @@
            s1.st_dev == s2.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):
     """Test whether a path is a mount point"""
     try:
         s1 = os.lstat(path)
-        s2 = os.lstat(join(path, '..'))
+        if isinstance(path, bytes):
+            parent = join(path, b'..')
+        else:
+            parent = join(path, '..')
+        s2 = os.lstat(parent)
     except os.error:
         return False # It doesn't exist -- so not a mount point :-)
     dev1 = s1.st_dev
     dev2 = s2.st_dev
     if dev1 != dev2:
         return True     # path/.. on a different device as path
     ino1 = s1.st_ino
     ino2 = s2.st_ino
     if ino1 == ino2:
         return True     # path/.. is the same i-node as path
     return False
 
 
 # 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):
     """Expand ~ and ~user constructions.  If user or $HOME is unknown,
     do nothing."""
-    if not path.startswith('~'):
+    if isinstance(path, bytes):
+        tilde = b'~'
+    else:
+        tilde = '~'
+    if not path.startswith(tilde):
         return path
-    i = path.find('/', 1)
+    sep = _get_sep(path)
+    i = path.find(sep, 1)
     if i < 0:
         i = len(path)
     if i == 1:
         if 'HOME' not in os.environ:
             import pwd
             userhome = pwd.getpwuid(os.getuid()).pw_dir
         else:
             userhome = os.environ['HOME']
     else:
         import pwd
+        name = path[1:i]
+        if isinstance(name, bytes):
+            name = str(name, 'ASCII')
         try:
-            pwent = pwd.getpwnam(path[1:i])
+            pwent = pwd.getpwnam(name)
         except KeyError:
             return path
         userhome = pwent.pw_dir
-    userhome = userhome.rstrip('/')
+    if isinstance(path, bytes):
+        userhome = userhome.encode(sys.getfilesystemencoding())
+    userhome = userhome.rstrip(sep)
     return userhome + path[i:]
 
 
 # Expand paths containing shell variable substitutions.
 # This expands the forms $variable and ${variable} only.
 # Non-existent variables are left unchanged.
 
 _varprog = None
+_varprogb = None
 
 def expandvars(path):
     """Expand shell variables of form $var and ${var}.  Unknown variables
     are left unchanged."""
-    global _varprog
-    if '$' not in path:
-        return path
-    if not _varprog:
-        import re
-        _varprog = re.compile(r'\$(\w+|\{[^}]*\})', re.ASCII)
+    global _varprog, _varprogb
+    if isinstance(path, bytes):
+        if b'$' not in path:
+            return path
+        if not _varprogb:
+            import re
+            _varprogb = re.compile(br'\$(\w+|\{[^}]*\})', re.ASCII)
+        search = _varprogb.search
+        start = b'{'
+        end = b'}'
+    else:
+        if '$' not in path:
+            return path
+        if not _varprog:
+            import re
+            _varprog = re.compile(r'\$(\w+|\{[^}]*\})', re.ASCII)
+        search = _varprog.search
+        start = '{'
+        end = '}'
     i = 0
     while True:
-        m = _varprog.search(path, i)
+        m = search(path, i)
         if not m:
             break
         i, j = m.span(0)
         name = m.group(1)
-        if name.startswith('{') and name.endswith('}'):
+        if name.startswith(start) and name.endswith(end):
             name = name[1:-1]
+        if isinstance(name, bytes):
+            name = str(name, 'ASCII')
         if name in os.environ:
             tail = path[j:]
-            path = path[:i] + os.environ[name]
+            value = os.environ[name]
+            if isinstance(path, bytes):
+                value = value.encode('ASCII')
+            path = path[:i] + value
             i = len(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):
     """Normalize path, eliminating double slashes, etc."""
-    if path == '':
-        return '.'
-    initial_slashes = path.startswith('/')
+    if isinstance(path, bytes):
+        sep = b'/'
+        empty = b''
+        dot = b'.'
+        dotdot = b'..'
+    else:
+        sep = '/'
+        empty = ''
+        dot = '.'
+        dotdot = '..'
+    if path == empty:
+        return dot
+    initial_slashes = path.startswith(sep)
     # POSIX allows one or two initial slashes, but treats three or more
     # as single slash.
     if (initial_slashes and
-        path.startswith('//') and not path.startswith('///')):
+        path.startswith(sep*2) and not path.startswith(sep*3)):
         initial_slashes = 2
-    comps = path.split('/')
+    comps = path.split(sep)
     new_comps = []
     for comp in comps:
-        if comp in ('', '.'):
+        if comp in (empty, dot):
             continue
-        if (comp != '..' or (not initial_slashes and not new_comps) or
-             (new_comps and new_comps[-1] == '..')):
+        if (comp != dotdot or (not initial_slashes and not new_comps) or
+             (new_comps and new_comps[-1] == dotdot)):
             new_comps.append(comp)
         elif new_comps:
             new_comps.pop()
     comps = new_comps
-    path = '/'.join(comps)
+    path = sep.join(comps)
     if initial_slashes:
-        path = '/'*initial_slashes + path
-    return path or '.'
+        path = sep*initial_slashes + path
+    return path or dot
 
 
 def abspath(path):
     """Return an absolute path."""
     if not isabs(path):
-        path = join(os.getcwd(), path)
+        if isinstance(path, bytes):
+            cwd = os.getcwdb()
+        else:
+            cwd = os.getcwd()
+        path = join(cwd, path)
     return normpath(path)
 
 
 # Return a canonical path (i.e. the absolute location of a file on the
 # filesystem).
 
 def realpath(filename):
     """Return the canonical path of the specified filename, eliminating any
 symbolic links encountered in the path."""
+    if isinstance(filename, bytes):
+        sep = b'/'
+        empty = b''
+    else:
+        sep = '/'
+        empty = ''
     if isabs(filename):
-        bits = ['/'] + filename.split('/')[1:]
-    else:
-        bits = [''] + filename.split('/')
+        bits = [sep] + filename.split(sep)[1:]
+    else:
+        bits = [empty] + filename.split(sep)
 
     for i in range(2, len(bits)+1):
         component = join(*bits[0:i])
         # Resolve symbolic links.
         if islink(component):
             resolved = _resolve_link(component)
             if resolved is None:
                 # Infinite loop -- return original component + rest of the path
                 return abspath(join(*([component] + bits[i:])))
             else:
@@ skipping 18 matching lines @@
         resolved = os.readlink(path)
         if not isabs(resolved):
             dir = dirname(path)
             path = normpath(join(dir, resolved))
         else:
             path = normpath(resolved)
     return path
 
 supports_unicode_filenames = False
 
-def relpath(path, start=curdir):
+def relpath(path, start=None):
     """Return a relative version of a path"""
 
     if not path:
         raise ValueError("no path specified")
+
+    if isinstance(path, bytes):
+        curdir = b'.'
+        sep = b'/'
+        pardir = b'..'
+    else:
+        curdir = '.'
+        sep = '/'
+        pardir = '..'
+
+    if start is None:
+        start = curdir
 
     start_list = abspath(start).split(sep)
     path_list = abspath(path).split(sep)
 
     # Work out how much of the filepath is shared by start and path.
     i = len(commonprefix([start_list, path_list]))
 
     rel_list = [pardir] * (len(start_list)-i) + path_list[i:]
     if not rel_list:
         return curdir
     return join(*rel_list)