Lib/posixpath.py

   1 """Common operations on Posix pathnames.
   2
   3 Instead of importing this module directly, import os and refer to
   4 this module as os.path.  The "os.path" name is an alias for this
   5 module on Posix systems; on other systems (e.g. Mac, Windows),
   6 os.path provides the same operations in a manner specific to that
   7 platform, and is an alias to another module (e.g. macpath, ntpath).
   8
   9 Some of this can actually be useful on non-Posix systems too, e.g.
  10 for manipulation of the pathname component of URLs.
  11 """
  12
  13 import os
  14 import stat
  15 from genericpath import *
  16
  17 __all__ = ["normcase","isabs","join","splitdrive","split","splitext",
  18            "basename","dirname","commonprefix","getsize","getmtime",
  19            "getatime","getctime","islink","exists","lexists","isdir","isfile",
  20            "ismount","walk","expanduser","expandvars","normpath","abspath",
  21            "samefile","sameopenfile","samestat",
  22            "curdir","pardir","sep","pathsep","defpath","altsep","extsep",
  23            "devnull","realpath","supports_unicode_filenames"]
  24
  25 # strings representing various path-related bits and pieces
  26 curdir = '.'
  27 pardir = '..'
  28 extsep = '.'
  29 sep = '/'
  30 pathsep = ':'
  31 defpath = ':/bin:/usr/bin'
  32 altsep = None
  33 devnull = '/dev/null'
  34
  35 # Normalize the case of a pathname.  Trivial in Posix, string.lower on Mac.
  36 # On MS-DOS this may also turn slashes into backslashes; however, other
  37 # normalizations (such as optimizing '../' away) are not allowed
  38 # (another function should be defined to do that).
  39
  40 def normcase(s):
  41     """Normalize case of pathname.  Has no effect under Posix"""
  42     return s
  43
  44
  45 # Return whether a path is absolute.
  46 # Trivial in Posix, harder on the Mac or MS-DOS.
  47
  48 def isabs(s):
  49     """Test whether a path is absolute"""
  50     return s.startswith('/')
  51
  52
  53 # Join pathnames.
  54 # Ignore the previous parts if a part is absolute.
  55 # Insert a '/' unless the first part is empty or already ends in '/'.
  56
  57 def join(a, *p):
  58     """Join two or more pathname components, inserting '/' as needed"""
  59     path = a
  60     for b in p:
  61         if b.startswith('/'):
  62             path = b
  63         elif path == '' or path.endswith('/'):
  64             path +=  b
  65         else:
  66             path += '/' + b
  67     return path
  68
  69
  70 # Split a path in head (everything up to the last '/') and tail (the
  71 # rest).  If the path ends in '/', tail will be empty.  If there is no
  72 # '/' in the path, head  will be empty.
  73 # Trailing '/'es are stripped from head unless it is the root.
  74
  75 def split(p):
  76     """Split a pathname.  Returns tuple "(head, tail)" where "tail" is
  77     everything after the final slash.  Either part may be empty."""
  78     i = p.rfind('/') + 1
  79     head, tail = p[:i], p[i:]
  80     if head and head != '/'*len(head):
  81         head = head.rstrip('/')
  82     return head, tail
  83
  84
  85 # Split a path in root and extension.
  86 # The extension is everything starting at the last dot in the last
  87 # pathname component; the root is everything before that.
  88 # It is always true that root + ext == p.
  89
  90 def splitext(p):
  91     """Split the extension from a pathname.  Extension is everything from the
  92     last dot to the end.  Returns "(root, ext)", either part may be empty."""
  93     i = p.rfind('.')
  94     if i<=p.rfind('/'):
  95         return p, ''
  96     else:
  97         return p[:i], p[i:]
  98
  99
 100 # Split a pathname into a drive specification and the rest of the
 101 # path.  Useful on DOS/Windows/NT; on Unix, the drive is always empty.
 102
 103 def splitdrive(p):
 104     """Split a pathname into drive and path. On Posix, drive is always
 105     empty."""
 106     return '', p
 107
 108
 109 # Return the tail (basename) part of a path, same as split(path)[1].
 110
 111 def basename(p):
 112     """Returns the final component of a pathname"""
 113     i = p.rfind('/') + 1
 114     return p[i:]
 115
 116
 117 # Return the head (dirname) part of a path, same as split(path)[0].
 118
 119 def dirname(p):
 120     """Returns the directory component of a pathname"""
 121     i = p.rfind('/') + 1
 122     head = p[:i]
 123     if head and head != '/'*len(head):
 124         head = head.rstrip('/')
 125     return head
 126
 127
 128 # Is a path a symbolic link?
 129 # This will always return false on systems where os.lstat doesn't exist.
 130
 131 def islink(path):
 132     """Test whether a path is a symbolic link"""
 133     try:
 134         st = os.lstat(path)
 135     except (os.error, AttributeError):
 136         return False
 137     return stat.S_ISLNK(st.st_mode)
 138
 139 # Being true for dangling symbolic links is also useful.
 140
 141 def lexists(path):
 142     """Test whether a path exists.  Returns True for broken symbolic links"""
 143     try:
 144         st = os.lstat(path)
 145     except os.error:
 146         return False
 147     return True
 148
 149
 150 # Are two filenames really pointing to the same file?
 151
 152 def samefile(f1, f2):
 153     """Test whether two pathnames reference the same actual file"""
 154     s1 = os.stat(f1)
 155     s2 = os.stat(f2)
 156     return samestat(s1, s2)
 157
 158
 159 # Are two open files really referencing the same file?
 160 # (Not necessarily the same file descriptor!)
 161
 162 def sameopenfile(fp1, fp2):
 163     """Test whether two open file objects reference the same file"""
 164     s1 = os.fstat(fp1)
 165     s2 = os.fstat(fp2)
 166     return samestat(s1, s2)
 167
 168
 169 # Are two stat buffers (obtained from stat, fstat or lstat)
 170 # describing the same file?
 171
 172 def samestat(s1, s2):
 173     """Test whether two stat buffers reference the same file"""
 174     return s1.st_ino == s2.st_ino and \
 175            s1.st_dev == s2.st_dev
 176
 177
 178 # Is a path a mount point?
 179 # (Does this work for all UNIXes?  Is it even guaranteed to work by Posix?)
 180
 181 def ismount(path):
 182     """Test whether a path is a mount point"""
 183     try:
 184         s1 = os.stat(path)
 185         s2 = os.stat(join(path, '..'))
 186     except os.error:
 187         return False # It doesn't exist -- so not a mount point :-)
 188     dev1 = s1.st_dev
 189     dev2 = s2.st_dev
 190     if dev1 != dev2:
 191         return True     # path/.. on a different device as path
 192     ino1 = s1.st_ino
 193     ino2 = s2.st_ino
 194     if ino1 == ino2:
 195         return True     # path/.. is the same i-node as path
 196     return False
 197
 198
 199 # Directory tree walk.
 200 # For each directory under top (including top itself, but excluding
 201 # '.' and '..'), func(arg, dirname, filenames) is called, where
 202 # dirname is the name of the directory and filenames is the list
 203 # of files (and subdirectories etc.) in the directory.
 204 # The func may modify the filenames list, to implement a filter,
 205 # or to impose a different order of visiting.
 206
 207 def walk(top, func, arg):
 208     """Directory tree walk with callback function.
 209
 210     For each directory in the directory tree rooted at top (including top
 211     itself, but excluding '.' and '..'), call func(arg, dirname, fnames).
 212     dirname is the name of the directory, and fnames a list of the names of
 213     the files and subdirectories in dirname (excluding '.' and '..').  func
 214     may modify the fnames list in-place (e.g. via del or slice assignment),
 215     and walk will only recurse into the subdirectories whose names remain in
 216     fnames; this can be used to implement a filter, or to impose a specific
 217     order of visiting.  No semantics are defined for, or required of, arg,
 218     beyond that arg is always passed to func.  It can be used, e.g., to pass
 219     a filename pattern, or a mutable object designed to accumulate
 220     statistics.  Passing None for arg is common."""
 221
 222     try:
 223         names = os.listdir(top)
 224     except os.error:
 225         return
 226     func(arg, top, names)
 227     for name in names:
 228         name = join(top, name)
 229         try:
 230             st = os.lstat(name)
 231         except os.error:
 232             continue
 233         if stat.S_ISDIR(st.st_mode):
 234             walk(name, func, arg)
 235
 236
 237 # Expand paths beginning with '~' or '~user'.
 238 # '~' means $HOME; '~user' means that user's home directory.
 239 # If the path doesn't begin with '~', or if the user or $HOME is unknown,
 240 # the path is returned unchanged (leaving error reporting to whatever
 241 # function is called with the expanded path as argument).
 242 # See also module 'glob' for expansion of *, ? and [...] in pathnames.
 243 # (A function should also be defined to do full *sh-style environment
 244 # variable expansion.)
 245
 246 def expanduser(path):
 247     """Expand ~ and ~user constructions.  If user or $HOME is unknown,
 248     do nothing."""
 249     if not path.startswith('~'):
 250         return path
 251     i = path.find('/', 1)
 252     if i < 0:
 253         i = len(path)
 254     if i == 1:
 255         if 'HOME' not in os.environ:
 256             import pwd
 257             userhome = pwd.getpwuid(os.getuid()).pw_dir
 258         else:
 259             userhome = os.environ['HOME']
 260     else:
 261         import pwd
 262         try:
 263             pwent = pwd.getpwnam(path[1:i])
 264         except KeyError:
 265             return path
 266         userhome = pwent.pw_dir
 267     userhome = userhome.rstrip('/')
 268     return userhome + path[i:]
 269
 270
 271 # Expand paths containing shell variable substitutions.
 272 # This expands the forms $variable and ${variable} only.
 273 # Non-existent variables are left unchanged.
 274
 275 _varprog = None
 276
 277 def expandvars(path):
 278     """Expand shell variables of form $var and ${var}.  Unknown variables
 279     are left unchanged."""
 280     global _varprog
 281     if '$' not in path:
 282         return path
 283     if not _varprog:
 284         import re
 285         _varprog = re.compile(r'\$(\w+|\{[^}]*\})')
 286     i = 0
 287     while True:
 288         m = _varprog.search(path, i)
 289         if not m:
 290             break
 291         i, j = m.span(0)
 292         name = m.group(1)
 293         if name.startswith('{') and name.endswith('}'):
 294             name = name[1:-1]
 295         if name in os.environ:
 296             tail = path[j:]
 297             path = path[:i] + os.environ[name]
 298             i = len(path)
 299             path += tail
 300         else:
 301             i = j
 302     return path
 303
 304
 305 # Normalize a path, e.g. A//B, A/./B and A/foo/../B all become A/B.
 306 # It should be understood that this may change the meaning of the path
 307 # if it contains symbolic links!
 308
 309 def normpath(path):
 310     """Normalize path, eliminating double slashes, etc."""
 311     if path == '':
 312         return '.'
 313     initial_slashes = path.startswith('/')
 314     # POSIX allows one or two initial slashes, but treats three or more
 315     # as single slash.
 316     if (initial_slashes and
 317         path.startswith('//') and not path.startswith('///')):
 318         initial_slashes = 2
 319     comps = path.split('/')
 320     new_comps = []
 321     for comp in comps:
 322         if comp in ('', '.'):
 323             continue
 324         if (comp != '..' or (not initial_slashes and not new_comps) or
 325              (new_comps and new_comps[-1] == '..')):
 326             new_comps.append(comp)
 327         elif new_comps:
 328             new_comps.pop()
 329     comps = new_comps
 330     path = '/'.join(comps)
 331     if initial_slashes:
 332         path = '/'*initial_slashes + path
 333     return path or '.'
 334
 335
 336 def abspath(path):
 337     """Return an absolute path."""
 338     if not isabs(path):
 339         path = join(os.getcwd(), path)
 340     return normpath(path)
 341
 342
 343 # Return a canonical path (i.e. the absolute location of a file on the
 344 # filesystem).
 345
 346 def realpath(filename):
 347     """Return the canonical path of the specified filename, eliminating any
 348 symbolic links encountered in the path."""
 349     if isabs(filename):
 350         bits = ['/'] + filename.split('/')[1:]
 351     else:
 352         bits = [''] + filename.split('/')
 353
 354     for i in range(2, len(bits)+1):
 355         component = join(*bits[0:i])
 356         # Resolve symbolic links.
 357         if islink(component):
 358             resolved = _resolve_link(component)
 359             if resolved is None:
 360                 # Infinite loop -- return original component + rest of the path
 361                 return abspath(join(*([component] + bits[i:])))
 362             else:
 363                 newpath = join(*([resolved] + bits[i:]))
 364                 return realpath(newpath)
 365
 366     return abspath(filename)
 367
 368
 369 def _resolve_link(path):
 370     """Internal helper function.  Takes a path and follows symlinks
 371     until we either arrive at something that isn't a symlink, or
 372     encounter a path we've seen before (meaning that there's a loop).
 373     """
 374     paths_seen = []
 375     while islink(path):
 376         if path in paths_seen:
 377             # Already seen this path, so we must have a symlink loop
 378             return None
 379         paths_seen.append(path)
 380         # Resolve where the link points to
 381         resolved = os.readlink(path)
 382         if not isabs(resolved):
 383             dir = dirname(path)
 384             path = normpath(join(dir, resolved))
 385         else:
 386             path = normpath(resolved)
 387     return path
 388
 389 supports_unicode_filenames = False