1
# Bazaar -- distributed version control
1
# Bazaar-NG -- distributed version control
3
3
# Copyright (C) 2005 by Canonical Ltd
5
5
# This program is free software; you can redistribute it and/or modify
6
6
# it under the terms of the GNU General Public License as published by
7
7
# the Free Software Foundation; either version 2 of the License, or
8
8
# (at your option) any later version.
10
10
# This program is distributed in the hope that it will be useful,
11
11
# but WITHOUT ANY WARRANTY; without even the implied warranty of
12
12
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13
13
# GNU General Public License for more details.
15
15
# You should have received a copy of the GNU General Public License
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 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,
53
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)
19
import os, types, re, time, types
20
from stat import S_ISREG, S_ISDIR, S_ISLNK, ST_MODE, ST_SIZE
22
from errors import bailout
64
24
def make_readonly(filename):
65
25
"""Make a filename read-only."""
26
# TODO: probably needs to be fixed for windows
66
27
mod = os.stat(filename).st_mode
67
28
mod = mod & 0777555
68
29
os.chmod(filename, mod)
74
35
os.chmod(filename, mod)
38
_QUOTE_RE = re.compile(r'([^a-zA-Z0-9.,:/_~-])')
81
"""Return a quoted filename filename
83
This previously used backslash quoting, but that works poorly on
85
# TODO: I'm not really sure this is the best format either.x
88
_QUOTE_RE = re.compile(r'([^a-zA-Z0-9.,:/\\_~-])')
90
if _QUOTE_RE.search(f):
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
139
def kind_marker(kind):
142
elif kind == _directory_kind:
144
elif kind == 'symlink':
147
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
372
def normalizepath(f):
373
if getattr(os.path, 'realpath', None) is not None:
377
[p,e] = os.path.split(f)
378
if e == "" or e == "." or e == "..":
381
return pathjoin(F(p), e)
385
"""Copy a file to a backup.
387
Backups are named in GNU-style, with a ~ suffix.
389
If the file is already a backup, it's not copied.
395
if has_symlinks() and os.path.islink(fn):
396
target = os.readlink(fn)
397
os.symlink(target, bfn)
405
outf = file(bfn, 'wb')
40
"""Return shell-quoted filename"""
41
## We could be a bit more terse by using double-quotes etc
42
f = _QUOTE_RE.sub(r'\\\1', f)
49
mode = os.lstat(f)[ST_MODE]
57
bailout("can't handle file kind with mode %o of %r" % (mode, f))
428
"""True if f is a symlink."""
430
return S_ISLNK(os.lstat(f)[ST_MODE])
434
def is_inside(dir, fname):
435
"""True if fname is inside dir.
437
The parameters should typically be passed to osutils.normpath first, so
438
that . and .. and repeated slashes are eliminated, and the separators
439
are canonical for the platform.
441
The empty string as a dir name is taken as top-of-tree and matches
444
>>> is_inside('src', pathjoin('src', 'foo.c'))
446
>>> is_inside('src', 'srccontrol')
448
>>> is_inside('src', pathjoin('src', 'a', 'a', 'a', 'foo.c'))
450
>>> is_inside('foo.c', 'foo.c')
452
>>> is_inside('foo.c', '')
454
>>> is_inside('', 'foo.c')
457
# XXX: Most callers of this can actually do something smarter by
458
# looking at the inventory
468
return fname.startswith(dir)
471
def is_inside_any(dir_list, fname):
472
"""True if fname is inside any of given dirs."""
473
for dirname in dir_list:
474
if is_inside(dirname, fname):
480
def is_inside_or_parent_of_any(dir_list, fname):
481
"""True if fname is a child or a parent of any of the given files."""
482
for dirname in dir_list:
483
if is_inside(dirname, fname) or is_inside(fname, dirname):
489
78
def pumpfile(fromfile, tofile):
490
79
"""Copy contents of one file to another."""
493
b = fromfile.read(BUFSIZE)
499
def file_iterator(input_file, readsize=32768):
501
b = input_file.read(readsize)
80
tofile.write(fromfile.read())
84
"""Return a new UUID"""
86
## XXX: Could alternatively read /proc/sys/kernel/random/uuid on
87
## Linux, but we need something portable for other systems;
88
## preferably an implementation in Python.
90
return chomp(file('/proc/sys/kernel/random/uuid').readline())
92
return chomp(os.popen('uuidgen').readline())
96
if s and (s[-1] == '\n'):
508
if getattr(f, 'tell', None) is not None:
104
## TODO: Maybe read in chunks to handle big files
105
if hasattr(f, 'tell'):
509
106
assert f.tell() == 0
521
def sha_strings(strings):
522
"""Return the sha-1 of concatenation of strings"""
524
map(s.update, strings)
525
109
return s.hexdigest()
528
112
def sha_string(f):
531
116
return s.hexdigest()
534
120
def fingerprint_file(f):
655
266
>>> splitpath('a/../b')
656
267
Traceback (most recent call last):
658
BzrError: sorry, '..' not allowed in path
269
BzrError: ("sorry, '..' not allowed in path", [])
660
271
assert isinstance(p, types.StringTypes)
662
# split on either delimiter because people might use either on
664
ps = re.split(r'[\\/]', p)
272
ps = [f for f in p.split('/') if f != '.']
669
raise BzrError("sorry, %r not allowed in path" % f)
670
elif (f == '.') or (f == ''):
275
bailout("sorry, %r not allowed in path" % f)
677
279
assert isinstance(p, list)
679
281
if (f == '..') or (f is None) or (f == ''):
680
raise BzrError("sorry, %r not allowed in path" % f)
684
@deprecated_function(zero_nine)
282
bailout("sorry, %r not allowed in path" % f)
685
286
def appendpath(p1, p2):
689
return pathjoin(p1, p2)
693
"""Split s into lines, but without removing the newline characters."""
694
lines = s.split('\n')
695
result = [line + '\n' for line in lines[:-1]]
697
result.append(lines[-1])
701
def hardlinks_good():
702
return sys.platform not in ('win32', 'cygwin', 'darwin')
705
def link_or_copy(src, dest):
706
"""Hardlink a file, or copy it if it can't be hardlinked."""
707
if not hardlinks_good():
712
except (OSError, IOError), e:
713
if e.errno != errno.EXDEV:
717
def delete_any(full_path):
718
"""Delete a file or directory."""
722
# We may be renaming a dangling inventory id
723
if e.errno not in (errno.EISDIR, errno.EACCES, errno.EPERM):
729
if getattr(os, 'symlink', None) is not None:
735
def contains_whitespace(s):
736
"""True if there are any whitespace characters in s."""
737
for ch in string.whitespace:
744
def contains_linebreaks(s):
745
"""True if there is any vertical whitespace in s."""
753
def relpath(base, path):
754
"""Return path relative to base, or raise exception.
756
The path may be either an absolute path or a path relative to the
757
current working directory.
759
os.path.commonprefix (python2.4) has a bad bug that it works just
760
on string prefixes, assuming that '/u' is a prefix of '/u2'. This
764
assert len(base) >= MIN_ABS_PATHLENGTH, ('Length of base must be equal or'
765
' exceed the platform minimum length (which is %d)' %
772
while len(head) >= len(base):
775
head, tail = os.path.split(head)
779
raise PathNotChild(rp, base)
787
def safe_unicode(unicode_or_utf8_string):
788
"""Coerce unicode_or_utf8_string into unicode.
790
If it is unicode, it is returned.
791
Otherwise it is decoded from utf-8. If a decoding error
792
occurs, it is wrapped as a If the decoding fails, the exception is wrapped
793
as a BzrBadParameter exception.
795
if isinstance(unicode_or_utf8_string, unicode):
796
return unicode_or_utf8_string
798
return unicode_or_utf8_string.decode('utf8')
799
except UnicodeDecodeError:
800
raise BzrBadParameterNotUnicode(unicode_or_utf8_string)
803
_platform_normalizes_filenames = False
804
if sys.platform == 'darwin':
805
_platform_normalizes_filenames = True
808
def normalizes_filenames():
809
"""Return True if this platform normalizes unicode filenames.
811
Mac OSX does, Windows/Linux do not.
813
return _platform_normalizes_filenames
816
def _accessible_normalized_filename(path):
817
"""Get the unicode normalized path, and if you can access the file.
819
On platforms where the system normalizes filenames (Mac OSX),
820
you can access a file by any path which will normalize correctly.
821
On platforms where the system does not normalize filenames
822
(Windows, Linux), you have to access a file by its exact path.
824
Internally, bzr only supports NFC/NFKC normalization, since that is
825
the standard for XML documents.
827
So return the normalized path, and a flag indicating if the file
828
can be accessed by that path.
831
return unicodedata.normalize('NFKC', unicode(path)), True
834
def _inaccessible_normalized_filename(path):
835
__doc__ = _accessible_normalized_filename.__doc__
837
normalized = unicodedata.normalize('NFKC', unicode(path))
838
return normalized, normalized == path
841
if _platform_normalizes_filenames:
842
normalized_filename = _accessible_normalized_filename
844
normalized_filename = _inaccessible_normalized_filename
847
def terminal_width():
848
"""Return estimated terminal width."""
849
if sys.platform == 'win32':
850
import bzrlib.win32console
851
return bzrlib.win32console.get_console_size()[0]
854
import struct, fcntl, termios
855
s = struct.pack('HHHH', 0, 0, 0, 0)
856
x = fcntl.ioctl(1, termios.TIOCGWINSZ, s)
857
width = struct.unpack('HHHH', x)[1]
862
width = int(os.environ['COLUMNS'])
871
def supports_executable():
872
return sys.platform != "win32"
875
def set_or_unset_env(env_variable, value):
876
"""Modify the environment, setting or removing the env_variable.
878
:param env_variable: The environment variable in question
879
:param value: The value to set the environment to. If None, then
880
the variable will be removed.
881
:return: The original value of the environment variable.
883
orig_val = os.environ.get(env_variable)
885
if orig_val is not None:
886
del os.environ[env_variable]
888
if isinstance(value, unicode):
889
value = value.encode(bzrlib.user_encoding)
890
os.environ[env_variable] = value
894
_validWin32PathRE = re.compile(r'^([A-Za-z]:[/\\])?[^:<>*"?\|]*$')
897
def check_legal_path(path):
898
"""Check whether the supplied path is legal.
899
This is only required on Windows, so we don't test on other platforms
902
if sys.platform != "win32":
904
if _validWin32PathRE.match(path) is None:
905
raise IllegalPath(path)
908
def walkdirs(top, prefix=""):
909
"""Yield data about all the directories in a tree.
911
This yields all the data about the contents of a directory at a time.
912
After each directory has been yielded, if the caller has mutated the list
913
to exclude some directories, they are then not descended into.
915
The data yielded is of the form:
916
((directory-relpath, directory-path-from-top),
917
[(relpath, basename, kind, lstat), ...]),
918
- directory-relpath is the relative path of the directory being returned
919
with respect to top. prefix is prepended to this.
920
- directory-path-from-root is the path including top for this directory.
921
It is suitable for use with os functions.
922
- relpath is the relative path within the subtree being walked.
923
- basename is the basename of the path
924
- kind is the kind of the file now. If unknown then the file is not
925
present within the tree - but it may be recorded as versioned. See
927
- lstat is the stat data *if* the file was statted.
928
- planned, not implemented:
929
path_from_tree_root is the path from the root of the tree.
931
:param prefix: Prefix the relpaths that are yielded with 'prefix'. This
932
allows one to walk a subtree but get paths that are relative to a tree
934
:return: an iterator over the dirs.
936
#TODO there is a bit of a smell where the results of the directory-
937
# summary in this, and the path from the root, may not agree
938
# depending on top and prefix - i.e. ./foo and foo as a pair leads to
939
# potentially confusing output. We should make this more robust - but
940
# not at a speed cost. RBC 20060731
943
_directory = _directory_kind
945
pending = [(prefix, "", _directory, None, top)]
948
currentdir = pending.pop()
949
# 0 - relpath, 1- basename, 2- kind, 3- stat, 4-toppath
952
relroot = currentdir[0] + '/'
955
for name in sorted(_listdir(top)):
956
abspath = top + '/' + name
957
statvalue = lstat(abspath)
958
dirblock.append((relroot + name, name,
959
file_kind_from_stat_mode(statvalue.st_mode),
961
yield (currentdir[0], top), dirblock
962
# push the user specified dirs from dirblock
963
for dir in reversed(dirblock):
964
if dir[2] == _directory:
968
def copy_tree(from_path, to_path, handlers={}):
969
"""Copy all of the entries in from_path into to_path.
971
:param from_path: The base directory to copy.
972
:param to_path: The target directory. If it does not exist, it will
974
:param handlers: A dictionary of functions, which takes a source and
975
destinations for files, directories, etc.
976
It is keyed on the file kind, such as 'directory', 'symlink', or 'file'
977
'file', 'directory', and 'symlink' should always exist.
978
If they are missing, they will be replaced with 'os.mkdir()',
979
'os.readlink() + os.symlink()', and 'shutil.copy2()', respectively.
981
# Now, just copy the existing cached tree to the new location
982
# We use a cheap trick here.
983
# Absolute paths are prefixed with the first parameter
984
# relative paths are prefixed with the second.
985
# So we can get both the source and target returned
986
# without any extra work.
988
def copy_dir(source, dest):
991
def copy_link(source, dest):
992
"""Copy the contents of a symlink"""
993
link_to = os.readlink(source)
994
os.symlink(link_to, dest)
996
real_handlers = {'file':shutil.copy2,
998
'directory':copy_dir,
1000
real_handlers.update(handlers)
1002
if not os.path.exists(to_path):
1003
real_handlers['directory'](from_path, to_path)
1005
for dir_info, entries in walkdirs(from_path, prefix=to_path):
1006
for relpath, name, kind, st, abspath in entries:
1007
real_handlers[kind](abspath, relpath)
1010
def path_prefix_key(path):
1011
"""Generate a prefix-order path key for path.
1013
This can be used to sort paths in the same way that walkdirs does.
1015
return (dirname(path) , path)
1018
def compare_paths_prefix_order(path_a, path_b):
1019
"""Compare path_a and path_b to generate the same order walkdirs uses."""
1020
key_a = path_prefix_key(path_a)
1021
key_b = path_prefix_key(path_b)
1022
return cmp(key_a, key_b)
1025
_cached_user_encoding = None
1028
def get_user_encoding():
1029
"""Find out what the preferred user encoding is.
1031
This is generally the encoding that is used for command line parameters
1032
and file contents. This may be different from the terminal encoding
1033
or the filesystem encoding.
1035
:return: A string defining the preferred user encoding
1037
global _cached_user_encoding
1038
if _cached_user_encoding is not None:
1039
return _cached_user_encoding
1041
if sys.platform == 'darwin':
1042
# work around egregious python 2.4 bug
1043
sys.platform = 'posix'
1047
sys.platform = 'darwin'
1052
_cached_user_encoding = locale.getpreferredencoding()
1053
except locale.Error, e:
1054
sys.stderr.write('bzr: warning: %s\n'
1055
' Could not determine what text encoding to use.\n'
1056
' This error usually means your Python interpreter\n'
1057
' doesn\'t support the locale set by $LANG (%s)\n'
1058
" Continuing with ascii encoding.\n"
1059
% (e, os.environ.get('LANG')))
1061
if _cached_user_encoding is None:
1062
_cached_user_encoding = 'ascii'
1063
return _cached_user_encoding
293
def extern_command(cmd, ignore_errors = False):
294
mutter('external command: %s' % `cmd`)
296
if not ignore_errors:
297
bailout('command failed')
added
removed
bzrlib/osutils.py
