70
mode = os.lstat(f)[ST_MODE]
101
_directory_kind = 'directory'
104
stat.S_IFDIR:_directory_kind,
105
stat.S_IFCHR:'chardev',
106
stat.S_IFBLK:'block',
109
stat.S_IFLNK:'symlink',
110
stat.S_IFSOCK:'socket',
114
def file_kind_from_stat_mode(stat_mode, _formats=_formats, _unknown='unknown'):
115
"""Generate a file kind from a stat mode. This is used in walkdirs.
117
Its performance is critical: Do not mutate without careful benchmarking.
120
return _formats[stat_mode & 0170000]
125
def file_kind(f, _lstat=os.lstat, _mapper=file_kind_from_stat_mode):
127
return _mapper(_lstat(f).st_mode)
129
if getattr(e, 'errno', None) == errno.ENOENT:
130
raise errors.NoSuchFile(f)
135
"""Return the current umask"""
136
# Assume that people aren't messing with the umask while running
137
# XXX: This is not thread safe, but there is no way to get the
138
# umask without setting it
89
144
def kind_marker(kind):
90
145
if kind == 'file':
92
elif kind == 'directory':
147
elif kind == _directory_kind:
94
149
elif kind == 'symlink':
97
raise BzrError('invalid file kind %r' % kind)
101
if hasattr(os, 'lstat'):
107
if e.errno == errno.ENOENT:
110
raise BzrError("lstat/stat of (%r): %r" % (f, e))
152
raise errors.BzrError('invalid file kind %r' % kind)
154
lexists = getattr(os.path, 'lexists', None)
158
if getattr(os, 'lstat') is not None:
164
if e.errno == errno.ENOENT:
167
raise errors.BzrError("lstat/stat of (%r): %r" % (f, e))
170
def fancy_rename(old, new, rename_func, unlink_func):
171
"""A fancy rename, when you don't have atomic rename.
173
:param old: The old path, to rename from
174
:param new: The new path, to rename to
175
:param rename_func: The potentially non-atomic rename function
176
:param unlink_func: A way to delete the target file if the full rename succeeds
179
# sftp rename doesn't allow overwriting, so play tricks:
181
base = os.path.basename(new)
182
dirname = os.path.dirname(new)
183
tmp_name = u'tmp.%s.%.9f.%d.%s' % (base, time.time(), os.getpid(), rand_chars(10))
184
tmp_name = pathjoin(dirname, tmp_name)
186
# Rename the file out of the way, but keep track if it didn't exist
187
# We don't want to grab just any exception
188
# something like EACCES should prevent us from continuing
189
# The downside is that the rename_func has to throw an exception
190
# with an errno = ENOENT, or NoSuchFile
193
rename_func(new, tmp_name)
194
except (errors.NoSuchFile,), e:
197
# RBC 20060103 abstraction leakage: the paramiko SFTP clients rename
198
# function raises an IOError with errno is None when a rename fails.
199
# This then gets caught here.
200
if e.errno not in (None, errno.ENOENT, errno.ENOTDIR):
203
if (getattr(e, 'errno', None) is None
204
or e.errno not in (errno.ENOENT, errno.ENOTDIR)):
211
# This may throw an exception, in which case success will
213
rename_func(old, new)
217
# If the file used to exist, rename it back into place
218
# otherwise just delete it from the tmp location
220
unlink_func(tmp_name)
222
rename_func(tmp_name, new)
225
# In Python 2.4.2 and older, os.path.abspath and os.path.realpath
226
# choke on a Unicode string containing a relative path if
227
# os.getcwd() returns a non-sys.getdefaultencoding()-encoded
229
_fs_enc = sys.getfilesystemencoding() or 'utf-8'
230
def _posix_abspath(path):
231
# jam 20060426 rather than encoding to fsencoding
232
# copy posixpath.abspath, but use os.getcwdu instead
233
if not posixpath.isabs(path):
234
path = posixpath.join(getcwd(), path)
235
return posixpath.normpath(path)
238
def _posix_realpath(path):
239
return posixpath.realpath(path.encode(_fs_enc)).decode(_fs_enc)
242
def _win32_fixdrive(path):
243
"""Force drive letters to be consistent.
245
win32 is inconsistent whether it returns lower or upper case
246
and even if it was consistent the user might type the other
247
so we force it to uppercase
248
running python.exe under cmd.exe return capital C:\\
249
running win32 python inside a cygwin shell returns lowercase c:\\
251
drive, path = _nt_splitdrive(path)
252
return drive.upper() + path
255
def _win32_abspath(path):
256
# Real _nt_abspath doesn't have a problem with a unicode cwd
257
return _win32_fixdrive(_nt_abspath(unicode(path)).replace('\\', '/'))
260
def _win32_realpath(path):
261
# Real _nt_realpath doesn't have a problem with a unicode cwd
262
return _win32_fixdrive(_nt_realpath(unicode(path)).replace('\\', '/'))
265
def _win32_pathjoin(*args):
266
return _nt_join(*args).replace('\\', '/')
269
def _win32_normpath(path):
270
return _win32_fixdrive(_nt_normpath(unicode(path)).replace('\\', '/'))
274
return _win32_fixdrive(os.getcwdu().replace('\\', '/'))
277
def _win32_mkdtemp(*args, **kwargs):
278
return _win32_fixdrive(tempfile.mkdtemp(*args, **kwargs).replace('\\', '/'))
281
def _win32_rename(old, new):
282
"""We expect to be able to atomically replace 'new' with old.
284
On win32, if new exists, it must be moved out of the way first,
288
fancy_rename(old, new, rename_func=os.rename, unlink_func=os.unlink)
290
if e.errno in (errno.EPERM, errno.EACCES, errno.EBUSY, errno.EINVAL):
291
# If we try to rename a non-existant file onto cwd, we get
292
# EPERM or EACCES instead of ENOENT, this will raise ENOENT
293
# if the old path doesn't exist, sometimes we get EACCES
294
# On Linux, we seem to get EBUSY, on Mac we get EINVAL
300
return unicodedata.normalize('NFKC', os.getcwdu())
303
# Default is to just use the python builtins, but these can be rebound on
304
# particular platforms.
305
abspath = _posix_abspath
306
realpath = _posix_realpath
307
pathjoin = os.path.join
308
normpath = os.path.normpath
311
dirname = os.path.dirname
312
basename = os.path.basename
313
# These were already imported into local scope
314
# mkdtemp = tempfile.mkdtemp
315
# rmtree = shutil.rmtree
317
MIN_ABS_PATHLENGTH = 1
320
if sys.platform == 'win32':
321
abspath = _win32_abspath
322
realpath = _win32_realpath
323
pathjoin = _win32_pathjoin
324
normpath = _win32_normpath
325
getcwd = _win32_getcwd
326
mkdtemp = _win32_mkdtemp
327
rename = _win32_rename
329
MIN_ABS_PATHLENGTH = 3
331
def _win32_delete_readonly(function, path, excinfo):
332
"""Error handler for shutil.rmtree function [for win32]
333
Helps to remove files and dirs marked as read-only.
335
type_, value = excinfo[:2]
336
if function in (os.remove, os.rmdir) \
337
and type_ == OSError \
338
and value.errno == errno.EACCES:
344
def rmtree(path, ignore_errors=False, onerror=_win32_delete_readonly):
345
"""Replacer for shutil.rmtree: could remove readonly dirs/files"""
346
return shutil.rmtree(path, ignore_errors, onerror)
347
elif sys.platform == 'darwin':
351
def get_terminal_encoding():
352
"""Find the best encoding for printing to the screen.
354
This attempts to check both sys.stdout and sys.stdin to see
355
what encoding they are in, and if that fails it falls back to
356
bzrlib.user_encoding.
357
The problem is that on Windows, locale.getpreferredencoding()
358
is not the same encoding as that used by the console:
359
http://mail.python.org/pipermail/python-list/2003-May/162357.html
361
On my standard US Windows XP, the preferred encoding is
362
cp1252, but the console is cp437
364
output_encoding = getattr(sys.stdout, 'encoding', None)
365
if not output_encoding:
366
input_encoding = getattr(sys.stdin, 'encoding', None)
367
if not input_encoding:
368
output_encoding = bzrlib.user_encoding
369
mutter('encoding stdout as bzrlib.user_encoding %r', output_encoding)
371
output_encoding = input_encoding
372
mutter('encoding stdout as sys.stdin encoding %r', output_encoding)
374
mutter('encoding stdout as sys.stdout encoding %r', output_encoding)
375
return output_encoding
112
378
def normalizepath(f):
113
if hasattr(os.path, 'realpath'):
379
if getattr(os.path, 'realpath', None) is not None:
117
383
[p,e] = os.path.split(f)
118
384
if e == "" or e == "." or e == "..":
121
return os.path.join(F(p), e)
387
return pathjoin(F(p), e)
123
if os.name == "posix":
124
# In Python 2.4.2 and older, os.path.abspath and os.path.realpath
125
# choke on a Unicode string containing a relative path if
126
# os.getcwd() returns a non-sys.getdefaultencoding()-encoded
128
_fs_enc = sys.getfilesystemencoding()
130
return os.path.abspath(path.encode(_fs_enc)).decode(_fs_enc)
132
return os.path.realpath(path.encode(_fs_enc)).decode(_fs_enc)
134
# We need to use the Unicode-aware os.path.abspath and
135
# os.path.realpath on Windows systems.
136
abspath = os.path.abspath
137
realpath = os.path.realpath
139
390
def backup_file(fn):
140
391
"""Copy a file to a backup.
480
804
s.insert(0, tail)
482
# XXX This should raise a NotChildPath exception, as its not tied
484
raise NotBranchError("path %r is not within branch %r" % (rp, base))
486
return os.sep.join(s)
806
raise errors.PathNotChild(rp, base)
814
def safe_unicode(unicode_or_utf8_string):
815
"""Coerce unicode_or_utf8_string into unicode.
817
If it is unicode, it is returned.
818
Otherwise it is decoded from utf-8. If a decoding error
819
occurs, it is wrapped as a If the decoding fails, the exception is wrapped
820
as a BzrBadParameter exception.
822
if isinstance(unicode_or_utf8_string, unicode):
823
return unicode_or_utf8_string
825
return unicode_or_utf8_string.decode('utf8')
826
except UnicodeDecodeError:
827
raise errors.BzrBadParameterNotUnicode(unicode_or_utf8_string)
830
_platform_normalizes_filenames = False
831
if sys.platform == 'darwin':
832
_platform_normalizes_filenames = True
835
def normalizes_filenames():
836
"""Return True if this platform normalizes unicode filenames.
838
Mac OSX does, Windows/Linux do not.
840
return _platform_normalizes_filenames
843
def _accessible_normalized_filename(path):
844
"""Get the unicode normalized path, and if you can access the file.
846
On platforms where the system normalizes filenames (Mac OSX),
847
you can access a file by any path which will normalize correctly.
848
On platforms where the system does not normalize filenames
849
(Windows, Linux), you have to access a file by its exact path.
851
Internally, bzr only supports NFC/NFKC normalization, since that is
852
the standard for XML documents.
854
So return the normalized path, and a flag indicating if the file
855
can be accessed by that path.
858
return unicodedata.normalize('NFKC', unicode(path)), True
861
def _inaccessible_normalized_filename(path):
862
__doc__ = _accessible_normalized_filename.__doc__
864
normalized = unicodedata.normalize('NFKC', unicode(path))
865
return normalized, normalized == path
868
if _platform_normalizes_filenames:
869
normalized_filename = _accessible_normalized_filename
871
normalized_filename = _inaccessible_normalized_filename
874
def terminal_width():
875
"""Return estimated terminal width."""
876
if sys.platform == 'win32':
877
import bzrlib.win32console
878
return bzrlib.win32console.get_console_size()[0]
881
import struct, fcntl, termios
882
s = struct.pack('HHHH', 0, 0, 0, 0)
883
x = fcntl.ioctl(1, termios.TIOCGWINSZ, s)
884
width = struct.unpack('HHHH', x)[1]
889
width = int(os.environ['COLUMNS'])
898
def supports_executable():
899
return sys.platform != "win32"
902
def set_or_unset_env(env_variable, value):
903
"""Modify the environment, setting or removing the env_variable.
905
:param env_variable: The environment variable in question
906
:param value: The value to set the environment to. If None, then
907
the variable will be removed.
908
:return: The original value of the environment variable.
910
orig_val = os.environ.get(env_variable)
912
if orig_val is not None:
913
del os.environ[env_variable]
915
if isinstance(value, unicode):
916
value = value.encode(bzrlib.user_encoding)
917
os.environ[env_variable] = value
921
_validWin32PathRE = re.compile(r'^([A-Za-z]:[/\\])?[^:<>*"?\|]*$')
924
def check_legal_path(path):
925
"""Check whether the supplied path is legal.
926
This is only required on Windows, so we don't test on other platforms
929
if sys.platform != "win32":
931
if _validWin32PathRE.match(path) is None:
932
raise errors.IllegalPath(path)
935
def walkdirs(top, prefix=""):
936
"""Yield data about all the directories in a tree.
938
This yields all the data about the contents of a directory at a time.
939
After each directory has been yielded, if the caller has mutated the list
940
to exclude some directories, they are then not descended into.
942
The data yielded is of the form:
943
((directory-relpath, directory-path-from-top),
944
[(relpath, basename, kind, lstat), ...]),
945
- directory-relpath is the relative path of the directory being returned
946
with respect to top. prefix is prepended to this.
947
- directory-path-from-root is the path including top for this directory.
948
It is suitable for use with os functions.
949
- relpath is the relative path within the subtree being walked.
950
- basename is the basename of the path
951
- kind is the kind of the file now. If unknown then the file is not
952
present within the tree - but it may be recorded as versioned. See
954
- lstat is the stat data *if* the file was statted.
955
- planned, not implemented:
956
path_from_tree_root is the path from the root of the tree.
958
:param prefix: Prefix the relpaths that are yielded with 'prefix'. This
959
allows one to walk a subtree but get paths that are relative to a tree
961
:return: an iterator over the dirs.
963
#TODO there is a bit of a smell where the results of the directory-
964
# summary in this, and the path from the root, may not agree
965
# depending on top and prefix - i.e. ./foo and foo as a pair leads to
966
# potentially confusing output. We should make this more robust - but
967
# not at a speed cost. RBC 20060731
970
_directory = _directory_kind
971
_listdir = os.listdir
972
pending = [(prefix, "", _directory, None, top)]
975
currentdir = pending.pop()
976
# 0 - relpath, 1- basename, 2- kind, 3- stat, 4-toppath
979
relroot = currentdir[0] + '/'
982
for name in sorted(_listdir(top)):
983
abspath = top + '/' + name
984
statvalue = lstat(abspath)
985
dirblock.append((relroot + name, name,
986
file_kind_from_stat_mode(statvalue.st_mode),
988
yield (currentdir[0], top), dirblock
989
# push the user specified dirs from dirblock
990
for dir in reversed(dirblock):
991
if dir[2] == _directory:
995
def copy_tree(from_path, to_path, handlers={}):
996
"""Copy all of the entries in from_path into to_path.
998
:param from_path: The base directory to copy.
999
:param to_path: The target directory. If it does not exist, it will
1001
:param handlers: A dictionary of functions, which takes a source and
1002
destinations for files, directories, etc.
1003
It is keyed on the file kind, such as 'directory', 'symlink', or 'file'
1004
'file', 'directory', and 'symlink' should always exist.
1005
If they are missing, they will be replaced with 'os.mkdir()',
1006
'os.readlink() + os.symlink()', and 'shutil.copy2()', respectively.
1008
# Now, just copy the existing cached tree to the new location
1009
# We use a cheap trick here.
1010
# Absolute paths are prefixed with the first parameter
1011
# relative paths are prefixed with the second.
1012
# So we can get both the source and target returned
1013
# without any extra work.
1015
def copy_dir(source, dest):
1018
def copy_link(source, dest):
1019
"""Copy the contents of a symlink"""
1020
link_to = os.readlink(source)
1021
os.symlink(link_to, dest)
1023
real_handlers = {'file':shutil.copy2,
1024
'symlink':copy_link,
1025
'directory':copy_dir,
1027
real_handlers.update(handlers)
1029
if not os.path.exists(to_path):
1030
real_handlers['directory'](from_path, to_path)
1032
for dir_info, entries in walkdirs(from_path, prefix=to_path):
1033
for relpath, name, kind, st, abspath in entries:
1034
real_handlers[kind](abspath, relpath)
1037
def path_prefix_key(path):
1038
"""Generate a prefix-order path key for path.
1040
This can be used to sort paths in the same way that walkdirs does.
1042
return (dirname(path) , path)
1045
def compare_paths_prefix_order(path_a, path_b):
1046
"""Compare path_a and path_b to generate the same order walkdirs uses."""
1047
key_a = path_prefix_key(path_a)
1048
key_b = path_prefix_key(path_b)
1049
return cmp(key_a, key_b)
1052
_cached_user_encoding = None
1055
def get_user_encoding():
1056
"""Find out what the preferred user encoding is.
1058
This is generally the encoding that is used for command line parameters
1059
and file contents. This may be different from the terminal encoding
1060
or the filesystem encoding.
1062
:return: A string defining the preferred user encoding
1064
global _cached_user_encoding
1065
if _cached_user_encoding is not None:
1066
return _cached_user_encoding
1068
if sys.platform == 'darwin':
1069
# work around egregious python 2.4 bug
1070
sys.platform = 'posix'
1074
sys.platform = 'darwin'
1079
_cached_user_encoding = locale.getpreferredencoding()
1080
except locale.Error, e:
1081
sys.stderr.write('bzr: warning: %s\n'
1082
' Could not determine what text encoding to use.\n'
1083
' This error usually means your Python interpreter\n'
1084
' doesn\'t support the locale set by $LANG (%s)\n'
1085
" Continuing with ascii encoding.\n"
1086
% (e, os.environ.get('LANG')))
1088
if _cached_user_encoding is None:
1089
_cached_user_encoding = 'ascii'
1090
return _cached_user_encoding
1093
def recv_all(socket, bytes):
1094
"""Receive an exact number of bytes.
1096
Regular Socket.recv() may return less than the requested number of bytes,
1097
dependning on what's in the OS buffer. MSG_WAITALL is not available
1098
on all platforms, but this should work everywhere. This will return
1099
less than the requested amount if the remote end closes.
1101
This isn't optimized and is intended mostly for use in testing.
1104
while len(b) < bytes:
1105
new = socket.recv(bytes - len(b))
1111
def dereference_path(path):
1112
"""Determine the real path to a file.
1114
All parent elements are dereferenced. But the file itself is not
1116
:param path: The original path. May be absolute or relative.
1117
:return: the real path *to* the file
1119
parent, base = os.path.split(path)
1120
# The pathjoin for '.' is a workaround for Python bug #1213894.
1121
# (initial path components aren't dereferenced)
1122
return pathjoin(realpath(pathjoin('.', parent)), base)
added
removed
bzrlib/osutils.py
