16
16
# along with this program; if not, write to the Free Software
17
17
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
19
from shutil import copyfile
20
from stat import (S_ISREG, S_ISDIR, S_ISLNK, ST_MODE, ST_SIZE,
21
S_ISCHR, S_ISBLK, S_ISFIFO, S_ISSOCK)
19
22
from cStringIO import StringIO
21
from ntpath import (abspath as _nt_abspath,
23
normpath as _nt_normpath,
24
realpath as _nt_realpath,
25
splitdrive as _nt_splitdrive,
28
from os import listdir
33
from shutil import copyfile
35
from stat import (S_ISREG, S_ISDIR, S_ISLNK, ST_MODE, ST_SIZE,
36
S_ISCHR, S_ISBLK, S_ISFIFO, S_ISSOCK)
45
from bzrlib.errors import (BzrError,
46
BzrBadParameterNotUnicode,
51
from bzrlib.symbol_versioning import (deprecated_function,
33
from bzrlib.errors import BzrError, PathNotChild
53
34
from bzrlib.trace import mutter
56
# On win32, O_BINARY is used to indicate the file should
57
# be opened in binary mode, rather than text mode.
58
# On other platforms, O_BINARY doesn't exist, because
59
# they always open in binary mode, so it is okay to
60
# OR with 0 on those platforms
61
O_BINARY = getattr(os, 'O_BINARY', 0)
64
37
def make_readonly(filename):
65
38
"""Make a filename read-only."""
66
39
mod = os.stat(filename).st_mode
96
_directory_kind = 'directory'
99
stat.S_IFDIR:_directory_kind,
100
stat.S_IFCHR:'chardev',
101
stat.S_IFBLK:'block',
104
stat.S_IFLNK:'symlink',
105
stat.S_IFSOCK:'socket',
109
def file_kind_from_stat_mode(stat_mode, _formats=_formats, _unknown='unknown'):
110
"""Generate a file kind from a stat mode. This is used in walkdirs.
112
Its performance is critical: Do not mutate without careful benchmarking.
115
return _formats[stat_mode & 0170000]
120
def file_kind(f, _lstat=os.lstat, _mapper=file_kind_from_stat_mode):
122
return _mapper(_lstat(f).st_mode)
124
if getattr(e, 'errno', None) == errno.ENOENT:
125
raise bzrlib.errors.NoSuchFile(f)
130
"""Return the current umask"""
131
# Assume that people aren't messing with the umask while running
132
# XXX: This is not thread safe, but there is no way to get the
133
# umask without setting it
70
mode = os.lstat(f)[ST_MODE]
139
89
def kind_marker(kind):
140
90
if kind == 'file':
142
elif kind == _directory_kind:
92
elif kind == 'directory':
144
94
elif kind == 'symlink':
147
97
raise BzrError('invalid file kind %r' % kind)
149
lexists = getattr(os.path, 'lexists', None)
153
if getattr(os, 'lstat') is not None:
159
if e.errno == errno.ENOENT:
162
raise BzrError("lstat/stat of (%r): %r" % (f, e))
165
def fancy_rename(old, new, rename_func, unlink_func):
166
"""A fancy rename, when you don't have atomic rename.
168
:param old: The old path, to rename from
169
:param new: The new path, to rename to
170
:param rename_func: The potentially non-atomic rename function
171
:param unlink_func: A way to delete the target file if the full rename succeeds
174
# sftp rename doesn't allow overwriting, so play tricks:
176
base = os.path.basename(new)
177
dirname = os.path.dirname(new)
178
tmp_name = u'tmp.%s.%.9f.%d.%s' % (base, time.time(), os.getpid(), rand_chars(10))
179
tmp_name = pathjoin(dirname, tmp_name)
181
# Rename the file out of the way, but keep track if it didn't exist
182
# We don't want to grab just any exception
183
# something like EACCES should prevent us from continuing
184
# The downside is that the rename_func has to throw an exception
185
# with an errno = ENOENT, or NoSuchFile
188
rename_func(new, tmp_name)
189
except (NoSuchFile,), e:
192
# RBC 20060103 abstraction leakage: the paramiko SFTP clients rename
193
# function raises an IOError with errno is None when a rename fails.
194
# This then gets caught here.
195
if e.errno not in (None, errno.ENOENT, errno.ENOTDIR):
198
if (getattr(e, 'errno', None) is None
199
or e.errno not in (errno.ENOENT, errno.ENOTDIR)):
206
# This may throw an exception, in which case success will
208
rename_func(old, new)
212
# If the file used to exist, rename it back into place
213
# otherwise just delete it from the tmp location
215
unlink_func(tmp_name)
217
rename_func(tmp_name, new)
220
# In Python 2.4.2 and older, os.path.abspath and os.path.realpath
221
# choke on a Unicode string containing a relative path if
222
# os.getcwd() returns a non-sys.getdefaultencoding()-encoded
224
_fs_enc = sys.getfilesystemencoding()
225
def _posix_abspath(path):
226
# jam 20060426 rather than encoding to fsencoding
227
# copy posixpath.abspath, but use os.getcwdu instead
228
if not posixpath.isabs(path):
229
path = posixpath.join(getcwd(), path)
230
return posixpath.normpath(path)
233
def _posix_realpath(path):
234
return posixpath.realpath(path.encode(_fs_enc)).decode(_fs_enc)
237
def _win32_fixdrive(path):
238
"""Force drive letters to be consistent.
240
win32 is inconsistent whether it returns lower or upper case
241
and even if it was consistent the user might type the other
242
so we force it to uppercase
243
running python.exe under cmd.exe return capital C:\\
244
running win32 python inside a cygwin shell returns lowercase c:\\
246
drive, path = _nt_splitdrive(path)
247
return drive.upper() + path
250
def _win32_abspath(path):
251
# Real _nt_abspath doesn't have a problem with a unicode cwd
252
return _win32_fixdrive(_nt_abspath(unicode(path)).replace('\\', '/'))
255
def _win32_realpath(path):
256
# Real _nt_realpath doesn't have a problem with a unicode cwd
257
return _win32_fixdrive(_nt_realpath(unicode(path)).replace('\\', '/'))
260
def _win32_pathjoin(*args):
261
return _nt_join(*args).replace('\\', '/')
264
def _win32_normpath(path):
265
return _win32_fixdrive(_nt_normpath(unicode(path)).replace('\\', '/'))
269
return _win32_fixdrive(os.getcwdu().replace('\\', '/'))
272
def _win32_mkdtemp(*args, **kwargs):
273
return _win32_fixdrive(tempfile.mkdtemp(*args, **kwargs).replace('\\', '/'))
276
def _win32_rename(old, new):
277
"""We expect to be able to atomically replace 'new' with old.
279
On win32, if new exists, it must be moved out of the way first,
283
fancy_rename(old, new, rename_func=os.rename, unlink_func=os.unlink)
285
if e.errno in (errno.EPERM, errno.EACCES, errno.EBUSY, errno.EINVAL):
286
# If we try to rename a non-existant file onto cwd, we get
287
# EPERM or EACCES instead of ENOENT, this will raise ENOENT
288
# if the old path doesn't exist, sometimes we get EACCES
289
# On Linux, we seem to get EBUSY, on Mac we get EINVAL
295
return unicodedata.normalize('NFKC', os.getcwdu())
298
# Default is to just use the python builtins, but these can be rebound on
299
# particular platforms.
300
abspath = _posix_abspath
301
realpath = _posix_realpath
302
pathjoin = os.path.join
303
normpath = os.path.normpath
305
mkdtemp = tempfile.mkdtemp
307
dirname = os.path.dirname
308
basename = os.path.basename
309
rmtree = shutil.rmtree
311
MIN_ABS_PATHLENGTH = 1
314
if sys.platform == 'win32':
315
abspath = _win32_abspath
316
realpath = _win32_realpath
317
pathjoin = _win32_pathjoin
318
normpath = _win32_normpath
319
getcwd = _win32_getcwd
320
mkdtemp = _win32_mkdtemp
321
rename = _win32_rename
323
MIN_ABS_PATHLENGTH = 3
325
def _win32_delete_readonly(function, path, excinfo):
326
"""Error handler for shutil.rmtree function [for win32]
327
Helps to remove files and dirs marked as read-only.
329
type_, value = excinfo[:2]
330
if function in (os.remove, os.rmdir) \
331
and type_ == OSError \
332
and value.errno == errno.EACCES:
333
bzrlib.osutils.make_writable(path)
338
def rmtree(path, ignore_errors=False, onerror=_win32_delete_readonly):
339
"""Replacer for shutil.rmtree: could remove readonly dirs/files"""
340
return shutil.rmtree(path, ignore_errors, onerror)
341
elif sys.platform == 'darwin':
345
def get_terminal_encoding():
346
"""Find the best encoding for printing to the screen.
348
This attempts to check both sys.stdout and sys.stdin to see
349
what encoding they are in, and if that fails it falls back to
350
bzrlib.user_encoding.
351
The problem is that on Windows, locale.getpreferredencoding()
352
is not the same encoding as that used by the console:
353
http://mail.python.org/pipermail/python-list/2003-May/162357.html
355
On my standard US Windows XP, the preferred encoding is
356
cp1252, but the console is cp437
358
output_encoding = getattr(sys.stdout, 'encoding', None)
359
if not output_encoding:
360
input_encoding = getattr(sys.stdin, 'encoding', None)
361
if not input_encoding:
362
output_encoding = bzrlib.user_encoding
363
mutter('encoding stdout as bzrlib.user_encoding %r', output_encoding)
365
output_encoding = input_encoding
366
mutter('encoding stdout as sys.stdin encoding %r', output_encoding)
368
mutter('encoding stdout as sys.stdout encoding %r', output_encoding)
369
return output_encoding
101
if hasattr(os, 'lstat'):
107
if e.errno == errno.ENOENT:
110
raise BzrError("lstat/stat of (%r): %r" % (f, e))
372
112
def normalizepath(f):
373
if getattr(os.path, 'realpath', None) is not None:
113
if hasattr(os.path, 'realpath'):
377
117
[p,e] = os.path.split(f)
378
118
if e == "" or e == "." or e == "..":
381
return pathjoin(F(p), e)
121
return os.path.join(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
384
139
def backup_file(fn):
385
140
"""Copy a file to a backup.
825
485
s.insert(0, tail)
487
# XXX This should raise a NotChildPath exception, as its not tied
827
489
raise PathNotChild(rp, base)
835
def safe_unicode(unicode_or_utf8_string):
836
"""Coerce unicode_or_utf8_string into unicode.
838
If it is unicode, it is returned.
839
Otherwise it is decoded from utf-8. If a decoding error
840
occurs, it is wrapped as a If the decoding fails, the exception is wrapped
841
as a BzrBadParameter exception.
843
if isinstance(unicode_or_utf8_string, unicode):
844
return unicode_or_utf8_string
846
return unicode_or_utf8_string.decode('utf8')
847
except UnicodeDecodeError:
848
raise BzrBadParameterNotUnicode(unicode_or_utf8_string)
851
_platform_normalizes_filenames = False
852
if sys.platform == 'darwin':
853
_platform_normalizes_filenames = True
856
def normalizes_filenames():
857
"""Return True if this platform normalizes unicode filenames.
859
Mac OSX does, Windows/Linux do not.
861
return _platform_normalizes_filenames
864
def _accessible_normalized_filename(path):
865
"""Get the unicode normalized path, and if you can access the file.
867
On platforms where the system normalizes filenames (Mac OSX),
868
you can access a file by any path which will normalize correctly.
869
On platforms where the system does not normalize filenames
870
(Windows, Linux), you have to access a file by its exact path.
872
Internally, bzr only supports NFC/NFKC normalization, since that is
873
the standard for XML documents.
875
So return the normalized path, and a flag indicating if the file
876
can be accessed by that path.
879
return unicodedata.normalize('NFKC', unicode(path)), True
882
def _inaccessible_normalized_filename(path):
883
__doc__ = _accessible_normalized_filename.__doc__
885
normalized = unicodedata.normalize('NFKC', unicode(path))
886
return normalized, normalized == path
889
if _platform_normalizes_filenames:
890
normalized_filename = _accessible_normalized_filename
892
normalized_filename = _inaccessible_normalized_filename
491
return os.sep.join(s)
895
495
def terminal_width():
896
496
"""Return estimated terminal width."""
897
if sys.platform == 'win32':
898
import bzrlib.win32console
899
return bzrlib.win32console.get_console_size()[0]
902
import struct, fcntl, termios
903
s = struct.pack('HHHH', 0, 0, 0, 0)
904
x = fcntl.ioctl(1, termios.TIOCGWINSZ, s)
905
width = struct.unpack('HHHH', x)[1]
910
width = int(os.environ['COLUMNS'])
919
def supports_executable():
920
return sys.platform != "win32"
923
def set_or_unset_env(env_variable, value):
924
"""Modify the environment, setting or removing the env_variable.
926
:param env_variable: The environment variable in question
927
:param value: The value to set the environment to. If None, then
928
the variable will be removed.
929
:return: The original value of the environment variable.
931
orig_val = os.environ.get(env_variable)
933
if orig_val is not None:
934
del os.environ[env_variable]
936
if isinstance(value, unicode):
937
value = value.encode(bzrlib.user_encoding)
938
os.environ[env_variable] = value
942
_validWin32PathRE = re.compile(r'^([A-Za-z]:[/\\])?[^:<>*"?\|]*$')
945
def check_legal_path(path):
946
"""Check whether the supplied path is legal.
947
This is only required on Windows, so we don't test on other platforms
950
if sys.platform != "win32":
952
if _validWin32PathRE.match(path) is None:
953
raise IllegalPath(path)
956
def walkdirs(top, prefix=""):
957
"""Yield data about all the directories in a tree.
959
This yields all the data about the contents of a directory at a time.
960
After each directory has been yielded, if the caller has mutated the list
961
to exclude some directories, they are then not descended into.
963
The data yielded is of the form:
964
((directory-relpath, directory-path-from-top),
965
[(relpath, basename, kind, lstat), ...]),
966
- directory-relpath is the relative path of the directory being returned
967
with respect to top. prefix is prepended to this.
968
- directory-path-from-root is the path including top for this directory.
969
It is suitable for use with os functions.
970
- relpath is the relative path within the subtree being walked.
971
- basename is the basename of the path
972
- kind is the kind of the file now. If unknown then the file is not
973
present within the tree - but it may be recorded as versioned. See
975
- lstat is the stat data *if* the file was statted.
976
- planned, not implemented:
977
path_from_tree_root is the path from the root of the tree.
979
:param prefix: Prefix the relpaths that are yielded with 'prefix'. This
980
allows one to walk a subtree but get paths that are relative to a tree
982
:return: an iterator over the dirs.
984
#TODO there is a bit of a smell where the results of the directory-
985
# summary in this, and the path from the root, may not agree
986
# depending on top and prefix - i.e. ./foo and foo as a pair leads to
987
# potentially confusing output. We should make this more robust - but
988
# not at a speed cost. RBC 20060731
991
_directory = _directory_kind
993
pending = [(prefix, "", _directory, None, top)]
996
currentdir = pending.pop()
997
# 0 - relpath, 1- basename, 2- kind, 3- stat, 4-toppath
1000
relroot = currentdir[0] + '/'
1003
for name in sorted(_listdir(top)):
1004
abspath = top + '/' + name
1005
statvalue = lstat(abspath)
1006
dirblock.append((relroot + name, name,
1007
file_kind_from_stat_mode(statvalue.st_mode),
1008
statvalue, abspath))
1009
yield (currentdir[0], top), dirblock
1010
# push the user specified dirs from dirblock
1011
for dir in reversed(dirblock):
1012
if dir[2] == _directory:
1016
def copy_tree(from_path, to_path, handlers={}):
1017
"""Copy all of the entries in from_path into to_path.
1019
:param from_path: The base directory to copy.
1020
:param to_path: The target directory. If it does not exist, it will
1022
:param handlers: A dictionary of functions, which takes a source and
1023
destinations for files, directories, etc.
1024
It is keyed on the file kind, such as 'directory', 'symlink', or 'file'
1025
'file', 'directory', and 'symlink' should always exist.
1026
If they are missing, they will be replaced with 'os.mkdir()',
1027
'os.readlink() + os.symlink()', and 'shutil.copy2()', respectively.
1029
# Now, just copy the existing cached tree to the new location
1030
# We use a cheap trick here.
1031
# Absolute paths are prefixed with the first parameter
1032
# relative paths are prefixed with the second.
1033
# So we can get both the source and target returned
1034
# without any extra work.
1036
def copy_dir(source, dest):
1039
def copy_link(source, dest):
1040
"""Copy the contents of a symlink"""
1041
link_to = os.readlink(source)
1042
os.symlink(link_to, dest)
1044
real_handlers = {'file':shutil.copy2,
1045
'symlink':copy_link,
1046
'directory':copy_dir,
1048
real_handlers.update(handlers)
1050
if not os.path.exists(to_path):
1051
real_handlers['directory'](from_path, to_path)
1053
for dir_info, entries in walkdirs(from_path, prefix=to_path):
1054
for relpath, name, kind, st, abspath in entries:
1055
real_handlers[kind](abspath, relpath)
1058
def path_prefix_key(path):
1059
"""Generate a prefix-order path key for path.
1061
This can be used to sort paths in the same way that walkdirs does.
1063
return (dirname(path) , path)
1066
def compare_paths_prefix_order(path_a, path_b):
1067
"""Compare path_a and path_b to generate the same order walkdirs uses."""
1068
key_a = path_prefix_key(path_a)
1069
key_b = path_prefix_key(path_b)
1070
return cmp(key_a, key_b)
1073
_cached_user_encoding = None
1076
def get_user_encoding():
1077
"""Find out what the preferred user encoding is.
1079
This is generally the encoding that is used for command line parameters
1080
and file contents. This may be different from the terminal encoding
1081
or the filesystem encoding.
1083
:return: A string defining the preferred user encoding
1085
global _cached_user_encoding
1086
if _cached_user_encoding is not None:
1087
return _cached_user_encoding
1089
if sys.platform == 'darwin':
1090
# work around egregious python 2.4 bug
1091
sys.platform = 'posix'
1095
sys.platform = 'darwin'
1100
_cached_user_encoding = locale.getpreferredencoding()
1101
except locale.Error, e:
1102
sys.stderr.write('bzr: warning: %s\n'
1103
' Could not determine what text encoding to use.\n'
1104
' This error usually means your Python interpreter\n'
1105
' doesn\'t support the locale set by $LANG (%s)\n'
1106
" Continuing with ascii encoding.\n"
1107
% (e, os.environ.get('LANG')))
1109
if _cached_user_encoding is None:
1110
_cached_user_encoding = 'ascii'
1111
return _cached_user_encoding
498
# TODO: Do something smart on Windows?
500
# TODO: Is there anything that gets a better update when the window
501
# is resized while the program is running? We could use the Python termcap
504
return int(os.environ['COLUMNS'])
505
except (IndexError, KeyError, ValueError):
added
removed
bzrlib/osutils.py
