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
  16 __all__ = ["normcase","isabs","join","splitdrive","split","splitext",
  17            "basename","dirname","commonprefix","getsize","getmtime",
  18            "getatime","getctime","islink","exists","isdir","isfile","ismount",
  19            "walk","expanduser","expandvars","normpath","abspath",
  20            "samefile","sameopenfile","samestat",
  21            "curdir","pardir","sep","pathsep","defpath","altsep","extsep",
  22            "realpath","supports_unicode_filenames"]
  23
  24 # strings representing various path-related bits and pieces
  25 curdir = '.'
  26 pardir = '..'
  27 extsep = '.'
  28 sep = '/'
  29 pathsep = ':'
  30 defpath = ':/bin:/usr/bin'
  31 altsep = None
  32
  33 # Normalize the case of a pathname.  Trivial in Posix, string.lower on Mac.
  34 # On MS-DOS this may also turn slashes into backslashes; however, other
  35 # normalizations (such as optimizing '../' away) are not allowed
  36 # (another function should be defined to do that).
  37
  38 def normcase(s):
  39     """Normalize case of pathname.  Has no effect under Posix"""
  40     return s
  41
  42
  43 # Return whether a path is absolute.
  44 # Trivial in Posix, harder on the Mac or MS-DOS.
  45
  46 def isabs(s):
  47     """Test whether a path is absolute"""
  48     return s[:1] == '/'
  49
  50
  51 # Join pathnames.
  52 # Ignore the previous parts if a part is absolute.
  53 # Insert a '/' unless the first part is empty or already ends in '/'.
  54
  55 def join(a, *p):
  56     """Join two or more pathname components, inserting '/' as needed"""
  57     path = a
  58     for b in p:
  59         if b[:1] == '/':
  60             path = b
  61         elif path == '' or path[-1:] == '/':
  62             path = path + b
  63         else:
  64             path = path + '/' + b
  65     return path
  66
  67
  68 # Split a path in head (everything up to the last '/') and tail (the
  69 # rest).  If the path ends in '/', tail will be empty.  If there is no
  70 # '/' in the path, head  will be empty.
  71 # Trailing '/'es are stripped from head unless it is the root.
  72
  73 def split(p):
  74     """Split a pathname.  Returns tuple "(head, tail)" where "tail" is
  75     everything after the final slash.  Either part may be empty."""
  76     i = p.rfind('/') + 1
  77     head, tail = p[:i], p[i:]
  78     if head and head != '/'*len(head):
  79         while head[-1] == '/':
  80             head = head[:-1]
  81     return head, tail
  82
  83
  84 # Split a path in root and extension.
  85 # The extension is everything starting at the last dot in the last
  86 # pathname component; the root is everything before that.
  87 # It is always true that root + ext == p.
  88
  89 def splitext(p):
  90     """Split the extension from a pathname.  Extension is everything from the
  91     last dot to the end.  Returns "(root, ext)", either part may be empty."""
  92     i = p.rfind('.')
  93     if i<=p.rfind('/'):
  94         return p, ''
  95     else:
  96         return p[:i], p[i:]
  97
  98
  99 # Split a pathname into a drive specification and the rest of the
 100 # path.  Useful on DOS/Windows/NT; on Unix, the drive is always empty.
 101
 102 def splitdrive(p):
 103     """Split a pathname into drive and path. On Posix, drive is always
 104     empty."""
 105     return '', p
 106
 107
 108 # Return the tail (basename) part of a path.
 109
 110 def basename(p):
 111     """Returns the final component of a pathname"""
 112     return split(p)[1]
 113
 114
 115 # Return the head (dirname) part of a path.
 116
 117 def dirname(p):
 118     """Returns the directory component of a pathname"""
 119     return split(p)[0]
 120
 121
 122 # Return the longest prefix of all list elements.
 123
 124 def commonprefix(m):
 125     "Given a list of pathnames, returns the longest common leading component"
 126     if not m: return ''
 127     prefix = m[0]
 128     for item in m:
 129         for i in range(len(prefix)):
 130             if prefix[:i+1] != item[:i+1]:
 131                 prefix = prefix[:i]
 132                 if i == 0: return ''
 133                 break
 134     return prefix
 135
 136
 137 # Get size, mtime, atime of files.
 138
 139 def getsize(filename):
 140     """Return the size of a file, reported by os.stat()."""
 141     return os.stat(filename).st_size
 142
 143 def getmtime(filename):
 144     """Return the last modification time of a file, reported by os.stat()."""
 145     return os.stat(filename).st_mtime
 146
 147 def getatime(filename):
 148     """Return the last access time of a file, reported by os.stat()."""
 149     return os.stat(filename).st_atime
 150
 151 def getctime(filename):
 152     """Return the creation time of a file, reported by os.stat()."""
 153     return os.stat(filename).st_ctime
 154
 155 # Is a path a symbolic link?
 156 # This will always return false on systems where os.lstat doesn't exist.
 157
 158 def islink(path):
 159     """Test whether a path is a symbolic link"""
 160     try:
 161         st = os.lstat(path)
 162     except (os.error, AttributeError):
 163         return False
 164     return stat.S_ISLNK(st.st_mode)
 165
 166
 167 # Does a path exist?
 168 # This is false for dangling symbolic links.
 169
 170 def exists(path):
 171     """Test whether a path exists.  Returns False for broken symbolic links"""
 172     try:
 173         st = os.stat(path)
 174     except os.error:
 175         return False
 176     return True
 177
 178
 179 # Is a path a directory?
 180 # This follows symbolic links, so both islink() and isdir() can be true
 181 # for the same path.
 182
 183 def isdir(path):
 184     """Test whether a path is a directory"""
 185     try:
 186         st = os.stat(path)
 187     except os.error:
 188         return False
 189     return stat.S_ISDIR(st.st_mode)
 190
 191
 192 # Is a path a regular file?
 193 # This follows symbolic links, so both islink() and isfile() can be true
 194 # for the same path.
 195
 196 def isfile(path):
 197     """Test whether a path is a regular file"""
 198     try:
 199         st = os.stat(path)
 200     except os.error:
 201         return False
 202     return stat.S_ISREG(st.st_mode)
 203
 204
 205 # Are two filenames really pointing to the same file?
 206
 207 def samefile(f1, f2):
 208     """Test whether two pathnames reference the same actual file"""
 209     s1 = os.stat(f1)
 210     s2 = os.stat(f2)
 211     return samestat(s1, s2)
 212
 213
 214 # Are two open files really referencing the same file?
 215 # (Not necessarily the same file descriptor!)
 216
 217 def sameopenfile(fp1, fp2):
 218     """Test whether two open file objects reference the same file"""
 219     s1 = os.fstat(fp1)
 220     s2 = os.fstat(fp2)
 221     return samestat(s1, s2)
 222
 223
 224 # Are two stat buffers (obtained from stat, fstat or lstat)
 225 # describing the same file?
 226
 227 def samestat(s1, s2):
 228     """Test whether two stat buffers reference the same file"""
 229     return s1.st_ino == s2.st_ino and \
 230            s1.st_dev == s2.st_dev
 231
 232
 233 # Is a path a mount point?
 234 # (Does this work for all UNIXes?  Is it even guaranteed to work by Posix?)
 235
 236 def ismount(path):
 237     """Test whether a path is a mount point"""
 238     try:
 239         s1 = os.stat(path)
 240         s2 = os.stat(join(path, '..'))
 241     except os.error:
 242         return False # It doesn't exist -- so not a mount point :-)
 243     dev1 = s1.st_dev
 244     dev2 = s2.st_dev
 245     if dev1 != dev2:
 246         return True     # path/.. on a different device as path
 247     ino1 = s1.st_ino
 248     ino2 = s2.st_ino
 249     if ino1 == ino2:
 250         return True     # path/.. is the same i-node as path
 251     return False
 252
 253
 254 # Directory tree walk.
 255 # For each directory under top (including top itself, but excluding
 256 # '.' and '..'), func(arg, dirname, filenames) is called, where
 257 # dirname is the name of the directory and filenames is the list
 258 # of files (and subdirectories etc.) in the directory.
 259 # The func may modify the filenames list, to implement a filter,
 260 # or to impose a different order of visiting.
 261
 262 def walk(top, func, arg):
 263     """Directory tree walk with callback function.
 264
 265     For each directory in the directory tree rooted at top (including top
 266     itself, but excluding '.' and '..'), call func(arg, dirname, fnames).
 267     dirname is the name of the directory, and fnames a list of the names of
 268     the files and subdirectories in dirname (excluding '.' and '..').  func
 269     may modify the fnames list in-place (e.g. via del or slice assignment),
 270     and walk will only recurse into the subdirectories whose names remain in
 271     fnames; this can be used to implement a filter, or to impose a specific
 272     order of visiting.  No semantics are defined for, or required of, arg,
 273     beyond that arg is always passed to func.  It can be used, e.g., to pass
 274     a filename pattern, or a mutable object designed to accumulate
 275     statistics.  Passing None for arg is common."""
 276
 277     try:
 278         names = os.listdir(top)
 279     except os.error:
 280         return
 281     func(arg, top, names)
 282     for name in names:
 283         name = join(top, name)
 284         try:
 285             st = os.lstat(name)
 286         except os.error:
 287             continue
 288         if stat.S_ISDIR(st.st_mode):
 289             walk(name, func, arg)
 290
 291
 292 # Expand paths beginning with '~' or '~user'.
 293 # '~' means $HOME; '~user' means that user's home directory.
 294 # If the path doesn't begin with '~', or if the user or $HOME is unknown,
 295 # the path is returned unchanged (leaving error reporting to whatever
 296 # function is called with the expanded path as argument).
 297 # See also module 'glob' for expansion of *, ? and [...] in pathnames.
 298 # (A function should also be defined to do full *sh-style environment
 299 # variable expansion.)
 300
 301 def expanduser(path):
 302     """Expand ~ and ~user constructions.  If user or $HOME is unknown,
 303     do nothing."""
 304     if path[:1] != '~':
 305         return path
 306     i, n = 1, len(path)
 307     while i < n and path[i] != '/':
 308         i = i + 1
 309     if i == 1:
 310         if not 'HOME' in os.environ:
 311             import pwd
 312             userhome = pwd.getpwuid(os.getuid())[5]
 313         else:
 314             userhome = os.environ['HOME']
 315     else:
 316         import pwd
 317         try:
 318             pwent = pwd.getpwnam(path[1:i])
 319         except KeyError:
 320             return path
 321         userhome = pwent[5]
 322     if userhome[-1:] == '/': i = i + 1
 323     return userhome + path[i:]
 324
 325
 326 # Expand paths containing shell variable substitutions.
 327 # This expands the forms $variable and ${variable} only.
 328 # Non-existent variables are left unchanged.
 329
 330 _varprog = None
 331
 332 def expandvars(path):
 333     """Expand shell variables of form $var and ${var}.  Unknown variables
 334     are left unchanged."""
 335     global _varprog
 336     if '$' not in path:
 337         return path
 338     if not _varprog:
 339         import re
 340         _varprog = re.compile(r'\$(\w+|\{[^}]*\})')
 341     i = 0
 342     while True:
 343         m = _varprog.search(path, i)
 344         if not m:
 345             break
 346         i, j = m.span(0)
 347         name = m.group(1)
 348         if name[:1] == '{' and name[-1:] == '}':
 349             name = name[1:-1]
 350         if name in os.environ:
 351             tail = path[j:]
 352             path = path[:i] + os.environ[name]
 353             i = len(path)
 354             path = path + tail
 355         else:
 356             i = j
 357     return path
 358
 359
 360 # Normalize a path, e.g. A//B, A/./B and A/foo/../B all become A/B.
 361 # It should be understood that this may change the meaning of the path
 362 # if it contains symbolic links!
 363
 364 def normpath(path):
 365     """Normalize path, eliminating double slashes, etc."""
 366     if path == '':
 367         return '.'
 368     initial_slashes = path.startswith('/')
 369     # POSIX allows one or two initial slashes, but treats three or more
 370     # as single slash.
 371     if (initial_slashes and
 372         path.startswith('//') and not path.startswith('///')):
 373         initial_slashes = 2
 374     comps = path.split('/')
 375     new_comps = []
 376     for comp in comps:
 377         if comp in ('', '.'):
 378             continue
 379         if (comp != '..' or (not initial_slashes and not new_comps) or
 380              (new_comps and new_comps[-1] == '..')):
 381             new_comps.append(comp)
 382         elif new_comps:
 383             new_comps.pop()
 384     comps = new_comps
 385     path = '/'.join(comps)
 386     if initial_slashes:
 387         path = '/'*initial_slashes + path
 388     return path or '.'
 389
 390
 391 def abspath(path):
 392     """Return an absolute path."""
 393     if not isabs(path):
 394         path = join(os.getcwd(), path)
 395     return normpath(path)
 396
 397
 398 # Return a canonical path (i.e. the absolute location of a file on the
 399 # filesystem).
 400
 401 def realpath(filename):
 402     """Return the canonical path of the specified filename, eliminating any
 403 symbolic links encountered in the path."""
 404     filename = abspath(filename)
 405
 406     bits = ['/'] + filename.split('/')[1:]
 407     for i in range(2, len(bits)+1):
 408         component = join(*bits[0:i])
 409         if islink(component):
 410             resolved = os.readlink(component)
 411             (dir, file) = split(component)
 412             resolved = normpath(join(dir, resolved))
 413             newpath = join(*([resolved] + bits[i:]))
 414             return realpath(newpath)
 415
 416     return filename
 417
 418 supports_unicode_filenames = False