blob: 9fd2e39b53720bfe406948de7d2e9d6648501a75 [file] [log] [blame]
# Copyright (C) 2010 Google Inc. All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions are
# met:
#
# * Redistributions of source code must retain the above copyright
# notice, this list of conditions and the following disclaimer.
# * Redistributions in binary form must reproduce the above
# copyright notice, this list of conditions and the following disclaimer
# in the documentation and/or other materials provided with the
# distribution.
# * Neither the name of Google Inc. nor the names of its
# contributors may be used to endorse or promote products derived from
# this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
"""Wrapper object for functions that access the filesystem.
A FileSystem object can be used to represent dependency on the
filesystem, and can be replaced with a MockFileSystem in tests.
"""
from __future__ import unicode_literals
import codecs
import errno
import glob
import hashlib
import logging
import os
import shutil
import stat
import subprocess
import sys
import tempfile
import time
try:
import exceptions
except ImportError:
# In py3, exceptions were moved into builtins
import builtins as exceptions
_log = logging.getLogger(__name__)
class FileSystem(object):
"""FileSystem interface for blinkpy.
Unless otherwise noted, all paths are allowed to be either absolute
or relative.
"""
sep = os.sep
pardir = os.pardir
WINDOWS_MAX_PATH = 260
def _path_for_access(self, path):
"""Ensures a path can be used to access the file.
Pass a path through this method when and only when the path is about to
be accessed via a syscall (e.g. open()); DO NOT use this method if the
path is to be manipulated by (most of) the functions in os.path, etc.
This method currently only works around one issue: the maximum path
length on Windows. If the current platform is Windows and the given path
is longer than MAX_PATH (260), the path will be converted to a UNC path
by first making the path absolute and then prepending the UNC magic
prefix '\\?\'. Otherwise, the method is a no-op.
(https://msdn.microsoft.com/en-us/library/aa365247.aspx#maxpath)
"""
if sys.platform == 'win32' and len(path) >= self.WINDOWS_MAX_PATH:
assert not path.startswith(r'\\'), "must not already be UNC"
return r'\\?\%s' % (self.abspath(path), )
return path
def abspath(self, path):
return os.path.abspath(path)
def realpath(self, path):
return os.path.realpath(path)
def path_to_module(self, module_name):
"""Returns the absolute path of a module."""
# FIXME: This is the only use of sys in this file. It's possible that
# this function should move elsewhere.
# __file__ is not always an absolute path in Python <3.4
# (https://bugs.python.org/issue18416).
return self.abspath(sys.modules[module_name].__file__)
def expanduser(self, path):
return os.path.expanduser(path)
def basename(self, path):
return os.path.basename(path)
def chdir(self, path):
return os.chdir(path)
def copyfile(self, source, destination):
# shutil.copyfile() uses open() underneath, which supports UNC paths.
shutil.copyfile(
self._path_for_access(source), self._path_for_access(destination))
def dirname(self, path):
return os.path.dirname(path)
def exists(self, path):
return os.path.exists(path)
def files_under(self, path, dirs_to_skip=None, file_filter=None):
"""Walks the filesystem tree under the given path in top-down order.
Args:
dirs_to_skip: A list of directories to skip over during the
traversal (e.g., .svn, resources, etc.).
file_filter: If not None, the filter will be invoked with the
filesystem object and the dirname and basename of each file
found. The file is included in the result if the callback
returns True.
Returns:
A list of all files under the given path in top-down order.
"""
dirs_to_skip = dirs_to_skip or []
def filter_all(fs, dirpath, basename):
return True
file_filter = file_filter or filter_all
files = []
if self.isfile(path):
if file_filter(self, self.dirname(path), self.basename(path)):
files.append(path)
return files
if self.basename(path) in dirs_to_skip:
return []
for (dirpath, dirnames, filenames) in os.walk(path):
for d in dirs_to_skip:
if d in dirnames:
dirnames.remove(d)
for filename in filenames:
if file_filter(self, dirpath, filename):
files.append(self.join(dirpath, filename))
return files
def getcwd(self):
return os.getcwd()
def glob(self, path):
return glob.glob(path)
def isabs(self, path):
return os.path.isabs(path)
def isfile(self, path):
return os.path.isfile(path)
def isdir(self, path):
return os.path.isdir(path)
def join(self, *comps):
# TODO(robertma): UNC paths are not currently supported, but can be done
# with os.path.splitunc().
return os.path.join(*comps)
def listdir(self, path):
return os.listdir(path)
def walk(self, top, topdown=True, onerror=None, followlinks=False):
return os.walk(
top, topdown=topdown, onerror=onerror, followlinks=followlinks)
def mkdtemp(self, **kwargs):
"""Creates and returns a uniquely-named directory.
This is like tempfile.mkdtemp, but if used in a with statement
the directory will self-delete at the end of the block (if the
directory is empty; non-empty directories raise errors). The
directory can be safely deleted inside the block as well, if so
desired.
Note that the object returned is not a string and does not support all
of the string methods. If you need a string, coerce the object to a
string and go from there.
"""
class TemporaryDirectory(object):
def __init__(self, **kwargs):
self._kwargs = kwargs
self._directory_path = tempfile.mkdtemp(**self._kwargs)
def __str__(self):
return self._directory_path
def __enter__(self):
return self._directory_path
def __exit__(self, type, value, traceback):
# Only self-delete if necessary.
# FIXME: Should we delete non-empty directories?
if os.path.exists(self._directory_path):
os.rmdir(self._directory_path)
return TemporaryDirectory(**kwargs)
def maybe_make_directory(self, *path):
"""Creates the specified directory if it doesn't already exist."""
try:
# os.makedirs() supports UNC paths:
# https://docs.python.org/2/library/os.html#os.makedirs
os.makedirs(self._path_for_access(self.join(*path)))
except OSError as error:
if error.errno != errno.EEXIST:
raise
def move(self, source, destination):
shutil.move(source, destination)
def mtime(self, path):
return os.stat(path).st_mtime
def normpath(self, path):
return os.path.normpath(path)
def open_binary_tempfile(self, suffix=''):
"""Creates, opens, and returns a binary temp file.
Returns a tuple of the file and the name.
"""
temp_fd, temp_name = tempfile.mkstemp(suffix)
f = os.fdopen(temp_fd, 'wb')
return f, temp_name
def open_binary_file_for_reading(self, path):
return file(self._path_for_access(path), 'rb')
def open_binary_file_for_writing(self, path):
return file(self._path_for_access(path), 'wb')
def read_binary_file(self, path):
"""Returns the contents of the file as a byte string."""
with self.open_binary_file_for_reading(path) as f:
return f.read()
def write_binary_file(self, path, contents):
with self.open_binary_file_for_writing(path) as f:
f.write(contents)
def open_text_tempfile(self, suffix=''):
"""Creates, opens, and returns a text temp file.
Returns a tuple of the file and the name.
"""
temp_fd, temp_name = tempfile.mkstemp(suffix)
# Close the OS fd opened by mkstemp as we will reopen the file with an
# explict encoding.
os.close(temp_fd)
f = codecs.open(temp_name, 'w', 'utf8')
return f, temp_name
def open_text_file_for_reading(self, path):
# Note: There appears to be an issue with the returned file objects not
# being seekable. See:
# http://stackoverflow.com/questions/1510188/can-seek-and-tell-work-with-utf-8-encoded-documents-in-python
return codecs.open(self._path_for_access(path), 'r', 'utf8')
def open_text_file_for_writing(self, path):
return codecs.open(self._path_for_access(path), 'w', 'utf8')
def read_text_file(self, path):
"""Returns the contents of the file as a Unicode string.
The file is read assuming it is a UTF-8 encoded file with no BOM.
"""
with self.open_text_file_for_reading(path) as f:
return f.read()
def write_text_file(self, path, contents):
"""Writes the contents to the file at the given location.
The file is written encoded as UTF-8 with no BOM.
"""
with self.open_text_file_for_writing(path) as f:
f.write(contents)
def sha1(self, path):
contents = self.read_binary_file(path)
return hashlib.sha1(contents).hexdigest()
def relpath(self, path, start='.'):
return os.path.relpath(path, start)
class _WindowsError(exceptions.OSError):
"""Fake exception for Linux and Mac."""
def remove(self, path, osremove=os.remove, retry=True):
"""Removes a file.
On Windows, if a process was recently killed and it held on to a file,
the OS will hold on to the file for a short while. This makes attempts
to delete the file fail. To work around that, this method will retry
for a few seconds until Windows is done with the file.
"""
try:
exceptions.WindowsError
except AttributeError:
exceptions.WindowsError = FileSystem._WindowsError
retry_timeout_sec = 3.0
sleep_interval = 0.1
while True:
try:
# The default os.remove() supports UNC paths on Windows.
osremove(self._path_for_access(path))
return True
except exceptions.WindowsError:
time.sleep(sleep_interval)
retry_timeout_sec -= sleep_interval
if retry_timeout_sec < 0 and not retry:
raise
def rmtree(self, path, ignore_errors=True, onerror=None):
"""Deletes the directory rooted at path, whether empty or not."""
if sys.platform == 'win32':
assert not path.startswith(r'\\'), 'root path cannot be UNC-style'
path_abs = self.abspath(path)
# Ensure the root of the tree being rmtree'd is not a long path.
# We can't convert it to a long path (using _path_for_access),
# because long paths are not supported in 'rmdir' on Windows 7.
assert len(path_abs) < self.WINDOWS_MAX_PATH, \
'root path is too long'
# Ensure (hopefully) that the quoting done on the next line is safe.
assert '"' not in path_abs, 'path contains a quotation mark (")'
# Create a shell command to call rmdir. (Note rmdir is a shell
# built-in, so it cannot be called as an executable.)
cmd = 'rmdir /s /q "{}"'.format(path_abs)
try:
subprocess.check_call(cmd, shell=True)
except subprocess.CalledProcessError as e:
if not ignore_errors:
if onerror:
# Unfortunately we can't know the exact file that failed,
# so we conjure up error info from the top level.
onerror('FileSystem.rmtree', path, sys.exc_info())
else:
raise e
else:
# shutil.rmtree() uses os.path.join() which doesn't support UNC paths.
shutil.rmtree(path, ignore_errors=ignore_errors, onerror=onerror)
def remove_contents(self, dirname):
"""Attempts to remove the contents of a directory tree.
Args:
dirname (string): Directory to remove the contents of.
Returns:
bool: True if the directory is now empty.
"""
return _remove_contents(self, dirname)
def copytree(self, source, destination):
# shutil.copytree() uses os.path.join() which doesn't support UNC paths.
shutil.copytree(source, destination)
def split(self, path):
"""Return (dirname, basename + '.' + ext)"""
return os.path.split(path)
def splitext(self, path):
"""Return (dirname + os.sep + basename, '.' + ext)"""
return os.path.splitext(path)
def make_executable(self, file_path):
os.chmod(
file_path, stat.S_IRUSR | stat.S_IWUSR | stat.S_IXUSR
| stat.S_IRGRP | stat.S_IXGRP)
def symlink(self, source, link_name):
"""Create a symbolic link. Unix only."""
os.symlink(source, link_name)
def sanitize_filename(self, filename, replacement='_'):
"""Replaces all illegal characters in a filename with a given character.
The list of illegal characters covers all restrictions on Linux, Mac and
Windows.
Args:
filename: A basename (or a part of the basename) of a file to
sanitize. It cannot be a path because slashes will be replaced.
replacement: A character to replace all illegal characters with.
Returns:
The sanitized filename.
"""
return _sanitize_filename(filename, replacement)
# _remove_contents is implemented in terms of other FileSystem functions. To
# allow it to be reused on the MockFileSystem object, we define it here and
# then call it in both FileSystem and MockFileSystem classes.
def _remove_contents(fs, dirname, sleep=time.sleep):
# We try multiple times, because on Windows a process which is
# currently closing could still have a file open in the directory.
_log.info('Removing contents of %s', dirname)
errors = []
def onerror(func, path, exc_info):
errors.append(path)
_log.exception('Failed at %s %s: %r', func, path, exc_info)
attempts = 0
while attempts < 5:
del errors[:]
for name in fs.listdir(dirname):
fullname = fs.join(dirname, name)
isdir = True
try:
isdir = fs.isdir(fullname)
except os.error:
onerror(fs.isdir, fullname, sys.exc_info())
continue
if isdir:
try:
_log.debug('Removing directory %s', fullname)
fs.rmtree(fullname, ignore_errors=False, onerror=onerror)
except os.error:
onerror(fs.rmtree, fullname, sys.exc_info())
continue
else:
try:
_log.debug('Removing file %s', fullname)
fs.remove(fullname, retry=False)
except os.error:
onerror(fs.remove, fullname, sys.exc_info())
continue
if not errors:
break
_log.warning('Contents removal failed, retrying in 1 second.')
attempts += 1
sleep(1)
# Check the path is gone.
if not fs.listdir(dirname):
return True
_log.warning('Unable to remove %s', dirname)
for dirpath, dirnames, filenames in fs.walk(
dirname, onerror=onerror, topdown=False):
for fname in filenames:
_log.warning('File %s still in output dir.', fs.join(
dirpath, fname))
for dname in dirnames:
_log.warning('Dir %s still in output dir.', fs.join(
dirpath, dname))
return False
# Same as _remove_contents, to reuse it in MockFileSystem.
def _sanitize_filename(filename, replacement):
# The list comes from restrictions on Windows:
# https://support.microsoft.com/lo-la/help/905231/information-about-the-characters-that-you-cannot-use-in-site-names--fo
# It also includes all illegal characters on Mac and Linux.
illegal_filename_chars = r'~#%&*{}\:<>?/|"'
for char in illegal_filename_chars:
filename = filename.replace(char, replacement)
return filename