|
""" path.py - An object representing a path to a file or directory.
Example:
from path import path d = path('/home/guido/bin') for f in d.files('*.py'): f.chmod(0755)
This module requires Python 2.2 or later.
URL: http://www.jorendorff.com/articles/python/path Author: Jason Orendorff <jason@jorendorff.com> (and others - see the url!) Date: 29 Feb 2004 """
# TODO # - Bug in write_text(). It doesn't support Universal newline mode. # - Better error message in listdir() when self isn't a # directory. (On Windows, the error message really sucks.) # - Make sure everything has a good docstring. # - Add methods for regex find and replace. # - guess_content_type() method? # - Perhaps support arguments to touch(). # - Could add split() and join() methods that generate warnings. # - Note: __add__() technically has a bug, I think, where # it doesn't play nice with other types that implement # __radd__(). Test this.
# Pre-2.3 support. Are unicode filenames supported? except AttributeError: pass
# Pre-2.3 workaround for basestring. except NameError: basestring = (str, unicode)
# Universal newline support
""" Represents a filesystem path.
For documentation on individual methods, consult their counterparts in os.path. """
# --- Special Python methods.
return 'path(%s)' % _base.__repr__(self)
# Adding a path and a string yields a path. return path(_base(self) + more)
return path(other + _base(self))
# The / operator joins paths. """ fp.__div__(rel) == fp / rel == fp.joinpath(rel)
Join two path components, adding a separator character if needed. """ return path(os.path.join(self, rel))
# Make the / operator work even when true division is enabled.
""" Return the current working directory as a path object. """ return path(os.getcwd())
# --- Operations on path strings.
""" Clean up a filename by calling expandvars(), expanduser(), and normpath() on it.
This is commonly everything needed to clean up a filename read from a configuration file, for example. """ return self.expandvars().expanduser().normpath()
base, ext = os.path.splitext(self.name) return base
f, ext = os.path.splitext(_base(self)) return ext
drive, r = os.path.splitdrive(self) return path(drive)
dirname, None, None, """ This path's parent directory, as a new path object.
For example, path('/usr/local/lib/libpython.so').parent == path('/usr/local/lib') """)
basename, None, None, """ The name of this file or directory without the full path.
For example, path('/usr/local/lib/libpython.so').name == 'libpython.so' """)
_get_namebase, None, None, """ The same as path.name, but with one file extension stripped off.
For example, path('/home/guido/python.tar.gz').name == 'python.tar.gz', but path('/home/guido/python.tar.gz').namebase == 'python.tar' """)
_get_ext, None, None, """ The file extension, for example '.py'. """)
_get_drive, None, None, """ The drive specifier, for example 'C:'. This is always empty on systems that don't use drive specifiers. """)
""" p.splitpath() -> Return (p.parent, p.name). """ parent, child = os.path.split(self) return path(parent), child
""" p.splitdrive() -> Return (p.drive, <the rest of p>).
Split the drive specifier from this path. If there is no drive specifier, p.drive is empty, so the return value is simply (path(''), p). This is always the case on Unix. """ drive, rel = os.path.splitdrive(self) return path(drive), rel
""" p.splitext() -> Return (p.stripext(), p.ext).
Split the filename extension from this path and return the two parts. Either part may be empty.
The extension is everything from '.' to the end of the last path segment. This has the property that if (a, b) == p.splitext(), then a + b == p. """ # Cast to plain string using _base because Python 2.2 # implementations of os.path.splitext use "for c in path:..." # which means something different when applied to a path # object. filename, ext = os.path.splitext(_base(self)) return path(filename), ext
""" p.stripext() -> Remove one file extension from the path.
For example, path('/home/guido/python.tar.gz').stripext() returns path('/home/guido/python.tar'). """ return self.splitext()[0]
unc, rest = os.path.splitunc(self) return path(unc), rest
unc, r = os.path.splitunc(self) return path(unc)
_get_uncshare, None, None, """ The UNC mount point for this path. This is empty for paths on local drives. """)
""" Join two or more path components, adding a separator character (os.sep) if needed. Returns a new path object. """ return path(os.path.join(self, *args))
""" Return a list of the path components in this path.
The first item in the list will be a path. Its value will be either os.curdir, os.pardir, empty, or the root directory of this path (for example, '/' or 'C:\\'). The other items in the list will be strings.
path.path.joinpath(*result) will yield the original path. """ parts = [] loc = self while loc != os.curdir and loc != os.pardir: prev = loc loc, child = prev.splitpath() if loc == prev: break parts.append(child) parts.append(loc) parts.reverse() return parts
""" Return this path as a relative path, based from the current working directory. """ cwd = path(os.getcwd()) return cwd.relpathto(self)
""" Return a relative path from self to dest.
If there is no relative path from self to dest, for example if they reside on different drives in Windows, then this returns dest.abspath(). """ origin = self.abspath() dest = path(dest).abspath()
orig_list = origin.normcase().splitall() # Don't normcase dest! We want to preserve the case. dest_list = dest.splitall()
if orig_list[0] != os.path.normcase(dest_list[0]): # Can't get here from there. return dest
# Find the location where the two paths start to differ. i = 0 for start_seg, dest_seg in zip(orig_list, dest_list): if start_seg != os.path.normcase(dest_seg): break i += 1
# Now i is the point where the two paths diverge. # Need a certain number of "os.pardir"s to work up # from the origin to the point of divergence. segments = [os.pardir] * (len(orig_list) - i) # Need to add the diverging part of dest_list. segments += dest_list[i:] if len(segments) == 0: # If they happen to be identical, use os.curdir. return path(os.curdir) else: return path(os.path.join(*segments))
# --- Listing, searching, walking, and matching
""" D.listdir() -> List of items in this directory.
Use D.files() or D.dirs() instead if you want a listing of just files or just subdirectories.
The elements of the list are path objects.
With the optional 'pattern' argument, this only lists items whose names match the given pattern. """ names = os.listdir(self) if pattern is not None: names = fnmatch.filter(names, pattern) return [self / child for child in names]
""" D.dirs() -> List of this directory's subdirectories.
The elements of the list are path objects. This does not walk recursively into subdirectories (but see path.walkdirs).
With the optional 'pattern' argument, this only lists directories whose names match the given pattern. For example, d.dirs('build-*'). """ return [p for p in self.listdir(pattern) if p.isdir()]
""" D.files() -> List of the files in this directory.
The elements of the list are path objects. This does not walk into subdirectories (see path.walkfiles).
With the optional 'pattern' argument, this only lists files whose names match the given pattern. For example, d.files('*.pyc'). """
return [p for p in self.listdir(pattern) if p.isfile()]
""" D.walk() -> iterator over files and subdirs, recursively.
The iterator yields path objects naming each child item of this directory and its descendants. This requires that D.isdir().
This performs a depth-first traversal of the directory tree. Each directory is returned just before all its children. """ for child in self.listdir(): if pattern is None or child.fnmatch(pattern): yield child if child.isdir(): for item in child.walk(pattern): yield item
""" D.walkdirs() -> iterator over subdirs, recursively.
With the optional 'pattern' argument, this yields only directories whose names match the given pattern. For example, mydir.walkdirs('*test') yields only directories with names ending in 'test'. """ for child in self.dirs(): if pattern is None or child.fnmatch(pattern): yield child for subsubdir in child.walkdirs(pattern): yield subsubdir
""" D.walkfiles() -> iterator over files in D, recursively.
The optional argument, pattern, limits the results to files with names that match the pattern. For example, mydir.walkfiles('*.tmp') yields only files with the .tmp extension. """ for child in self.listdir(): if child.isfile(): if pattern is None or child.fnmatch(pattern): yield child elif child.isdir(): for f in child.walkfiles(pattern): yield f
""" Return True if self.name matches the given pattern.
pattern - A filename pattern with wildcards, for example '*.py'. """ return fnmatch.fnmatch(self.name, pattern)
""" Return a list of path objects that match the pattern.
pattern - a path relative to this directory, with wildcards.
For example, path('/users').glob('*/bin/*') returns a list of all the files users have in their bin directories. """ return map(path, glob.glob(_base(self / pattern)))
# --- Reading or writing an entire file at once.
""" Open this file. Return a file object. """ return file(self, mode)
""" Open this file, read all bytes, return them as a string. """ f = self.open('rb') try: return f.read() finally: f.close()
""" Open this file and write the given bytes to it.
Default behavior is to overwrite any existing file. Call this with write_bytes(bytes, append=True) to append instead. """ if append: mode = 'ab' else: mode = 'wb' f = self.open(mode) try: f.write(bytes) finally: f.close()
""" Open this file, read it in, return the content as a string.
This uses 'U' mode in Python 2.3 and later, so '\r\n' and '\r' are automatically translated to '\n'.
Optional arguments:
encoding - The Unicode encoding (or character set) of the file. If present, the content of the file is decoded and returned as a unicode object; otherwise it is returned as an 8-bit str. errors - How to handle Unicode errors; see help(str.decode) for the options. Default is 'strict'. """ if encoding is None: # 8-bit f = self.open(_textmode) try: return f.read() finally: f.close() else: # Unicode f = codecs.open(self, 'r', encoding, errors) # (Note - Can't use 'U' mode here, since codecs.open # doesn't support 'U' mode, even in Python 2.3.) try: t = f.read() finally: f.close() return t.replace(u'\r\n', u'\n').replace(u'\r', u'\n')
""" Write the given text to this file.
The default behavior is to overwrite any existing file; to append instead, use the 'append=True' keyword argument.
There are two differences between path.write_text() and path.write_bytes(): Unicode handling and newline handling.
--- Unicode
If 'text' isn't Unicode, this essentially just does open(self, 'w').write(text). The 'encoding' and 'errors' arguments are ignored.
If 'text' is Unicode, it is first converted to bytes using the specified 'encoding' (or the default encoding if 'encoding' isn't specified). The 'errors' argument applies only to this conversion.
--- Newlines
write_text() converts from programmer-friendly newlines (always '\n') to platform-specific newlines (see os.linesep; on Windows, for example, the end-of-line marker is '\r\n'). This applies to Unicode text the same as to 8-bit text.
Because of this conversion, the text should only contain plain newlines ('\n'), just like the return value of path.text(). If the text contains the characters '\r\n', it may be written as '\r\r\n' or '\r\r' depending on your platform. (This is exactly the same as when you open a file for writing with fopen(filename, "w") in C or file(filename, 'w') in Python.) """ if isinstance(text, unicode): text = text.replace(u'\n', os.linesep) if encoding is None: encoding = sys.getdefaultencoding() bytes = text.encode(encoding, errors) self.write_bytes(bytes, append) else: if append: mode = 'a' else: mode = 'w' f = self.open(mode) try: f.write(text) finally: f.close()
""" Open this file, read all lines, return them in a list.
Optional arguments: encoding - The Unicode encoding (or character set) of the file. The default is None, meaning the content of the file is read as 8-bit characters and returned as a list of (non-Unicode) str objects. errors - How to handle Unicode errors; see help(str.decode) for the options. Default is 'strict' retain - If true, retain newline characters; but all newline character combinations ('\r', '\n', '\r\n') are translated to '\n'. If false, newline characters are stripped off. Default is True.
This uses 'U' mode in Python 2.3 and later. """ if encoding is None and retain: f = self.open(_textmode) try: return f.readlines() finally: f.close() else: return self.text(encoding, errors).splitlines(retain)
linesep=os.linesep): """ Overwrite this file with the given lines of text.
lines - A list of strings. encoding - A Unicode encoding to use. This applies only if 'lines' contains any Unicode strings. errors - How to handle errors in Unicode encoding. This also applies only to Unicode strings. linesep - A character sequence that will be added at the end of every line that doesn't already have it. """ f = self.open('wb') try: for line in lines: if not line.endswith(linesep): line += linesep if isinstance(line, unicode): if encoding is None: encoding = sys.getdefaultencoding() line = line.encode(encoding, errors=errors) f.write(line) finally: f.close()
# --- Methods for querying the filesystem.
samefile = os.path.samefile
getatime, None, None, """ Last access time of the file. """)
getmtime, None, None, """ Last-modified time of the file. """)
getctime, None, None, """ Creation time of the file. """)
getsize, None, None, """ Size of the file, in bytes. """)
""" Return true if current user has access to this path.
mode - One of the constants os.F_OK, os.R_OK, os.W_OK, os.X_OK """ return os.access(self, mode)
""" Perform a stat() system call on this path. """ return os.stat(self)
""" Like path.stat(), but do not follow symbolic links. """ return os.lstat(self)
def statvfs(self): """ Perform a statvfs() system call on this path. """ return os.statvfs(self)
def pathconf(self, name): return os.pathconf(self, name)
# --- Modifying operations on files and directories
""" Set the access and modified times of this file. """ os.utime(self, times)
os.chmod(self, mode)
def chown(self, uid, gid): os.chown(self, uid, gid)
os.rename(self, new)
os.renames(self, new)
# --- Create/delete operations on directories
os.mkdir(self, mode)
os.makedirs(self, mode)
os.rmdir(self)
os.removedirs(self)
# --- Modifying operations on files
""" Set the access/modified times of this file to the current time. Create the file if it does not exist. """ fd = os.open(self, os.O_WRONLY | os.O_CREAT, 0666) os.close(fd) os.utime(self, None)
os.remove(self)
os.unlink(self)
# --- Links
def link(self, newpath): """ Create a hard link at 'newpath', pointing to this file. """ os.link(self, newpath)
def symlink(self, newlink): """ Create a symbolic link at 'newlink', pointing here. """ os.symlink(self, newlink)
def readlink(self): """ Return the path to which this symbolic link points.
The result may be an absolute or a relative path. """ return path(os.readlink(self))
def readlinkabs(self): """ Return the path to which this symbolic link points.
The result is always an absolute path. """ p = self.readlink() if p.isabs(): return p else: return (self.parent / p).abspath()
# --- High-level functions from shutil
# --- Special stuff from os
def chroot(self): os.chroot(self)
os.startfile(self)
|