1
# Bazaar-NG -- distributed version control
3
# Copyright (C) 2005 by Canonical Ltd
5
# This program is free software; you can redistribute it and/or modify
6
# it under the terms of the GNU General Public License as published by
7
# the Free Software Foundation; either version 2 of the License, or
8
# (at your option) any later version.
10
# This program is distributed in the hope that it will be useful,
11
# but WITHOUT ANY WARRANTY; without even the implied warranty of
12
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13
# GNU General Public License for more details.
15
# You should have received a copy of the GNU General Public License
16
# along with this program; if not, write to the Free Software
17
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
19
from cStringIO import StringIO
22
from os import listdir
26
from shutil import copyfile
28
from stat import (S_ISREG, S_ISDIR, S_ISLNK, ST_MODE, ST_SIZE,
29
S_ISCHR, S_ISBLK, S_ISFIFO, S_ISSOCK)
36
from ntpath import (abspath as _nt_abspath,
38
normpath as _nt_normpath,
39
realpath as _nt_realpath,
43
from bzrlib.errors import (BzrError,
44
BzrBadParameterNotUnicode,
49
from bzrlib.symbol_versioning import *
50
from bzrlib.trace import mutter
51
import bzrlib.win32console
54
def make_readonly(filename):
55
"""Make a filename read-only."""
56
mod = os.stat(filename).st_mode
58
os.chmod(filename, mod)
61
def make_writable(filename):
62
mod = os.stat(filename).st_mode
64
os.chmod(filename, mod)
71
"""Return a quoted filename filename
73
This previously used backslash quoting, but that works poorly on
75
# TODO: I'm not really sure this is the best format either.x
78
_QUOTE_RE = re.compile(r'([^a-zA-Z0-9.,:/\\_~-])')
80
if _QUOTE_RE.search(f):
86
_directory_kind = 'directory'
89
stat.S_IFDIR:_directory_kind,
90
stat.S_IFCHR:'chardev',
94
stat.S_IFLNK:'symlink',
95
stat.S_IFSOCK:'socket',
99
def file_kind_from_stat_mode(stat_mode, _formats=_formats, _unknown='unknown'):
100
"""Generate a file kind from a stat mode. This is used in walkdirs.
102
Its performance is critical: Do not mutate without careful benchmarking.
105
return _formats[stat_mode & 0170000]
110
def file_kind(f, _lstat=os.lstat, _mapper=file_kind_from_stat_mode):
111
return _mapper(_lstat(f).st_mode)
114
def kind_marker(kind):
117
elif kind == _directory_kind:
119
elif kind == 'symlink':
122
raise BzrError('invalid file kind %r' % kind)
124
lexists = getattr(os.path, 'lexists', None)
128
if hasattr(os, 'lstat'):
134
if e.errno == errno.ENOENT:
137
raise BzrError("lstat/stat of (%r): %r" % (f, e))
140
def fancy_rename(old, new, rename_func, unlink_func):
141
"""A fancy rename, when you don't have atomic rename.
143
:param old: The old path, to rename from
144
:param new: The new path, to rename to
145
:param rename_func: The potentially non-atomic rename function
146
:param unlink_func: A way to delete the target file if the full rename succeeds
149
# sftp rename doesn't allow overwriting, so play tricks:
151
base = os.path.basename(new)
152
dirname = os.path.dirname(new)
153
tmp_name = u'tmp.%s.%.9f.%d.%s' % (base, time.time(), os.getpid(), rand_chars(10))
154
tmp_name = pathjoin(dirname, tmp_name)
156
# Rename the file out of the way, but keep track if it didn't exist
157
# We don't want to grab just any exception
158
# something like EACCES should prevent us from continuing
159
# The downside is that the rename_func has to throw an exception
160
# with an errno = ENOENT, or NoSuchFile
163
rename_func(new, tmp_name)
164
except (NoSuchFile,), e:
167
# RBC 20060103 abstraction leakage: the paramiko SFTP clients rename
168
# function raises an IOError with errno == None when a rename fails.
169
# This then gets caught here.
170
if e.errno not in (None, errno.ENOENT, errno.ENOTDIR):
173
if (not hasattr(e, 'errno')
174
or e.errno not in (errno.ENOENT, errno.ENOTDIR)):
181
# This may throw an exception, in which case success will
183
rename_func(old, new)
187
# If the file used to exist, rename it back into place
188
# otherwise just delete it from the tmp location
190
unlink_func(tmp_name)
192
rename_func(tmp_name, new)
195
# In Python 2.4.2 and older, os.path.abspath and os.path.realpath
196
# choke on a Unicode string containing a relative path if
197
# os.getcwd() returns a non-sys.getdefaultencoding()-encoded
199
_fs_enc = sys.getfilesystemencoding()
200
def _posix_abspath(path):
201
return os.path.abspath(path.encode(_fs_enc)).decode(_fs_enc)
202
# jam 20060426 This is another possibility which mimics
203
# os.path.abspath, only uses unicode characters instead
204
# if not os.path.isabs(path):
205
# return os.path.join(os.getcwdu(), path)
209
def _posix_realpath(path):
210
return os.path.realpath(path.encode(_fs_enc)).decode(_fs_enc)
213
def _win32_abspath(path):
214
return _nt_abspath(path.encode(_fs_enc)).decode(_fs_enc).replace('\\', '/')
217
def _win32_realpath(path):
218
return _nt_realpath(path.encode(_fs_enc)).decode(_fs_enc).replace('\\', '/')
221
def _win32_pathjoin(*args):
222
return _nt_join(*args).replace('\\', '/')
225
def _win32_normpath(path):
226
return _nt_normpath(path).replace('\\', '/')
230
return os.getcwdu().replace('\\', '/')
233
def _win32_mkdtemp(*args, **kwargs):
234
return tempfile.mkdtemp(*args, **kwargs).replace('\\', '/')
237
def _win32_rename(old, new):
238
fancy_rename(old, new, rename_func=os.rename, unlink_func=os.unlink)
241
# Default is to just use the python builtins, but these can be rebound on
242
# particular platforms.
243
abspath = _posix_abspath
244
realpath = _posix_realpath
245
pathjoin = os.path.join
246
normpath = os.path.normpath
248
mkdtemp = tempfile.mkdtemp
250
dirname = os.path.dirname
251
basename = os.path.basename
252
rmtree = shutil.rmtree
254
MIN_ABS_PATHLENGTH = 1
257
if sys.platform == 'win32':
258
abspath = _win32_abspath
259
realpath = _win32_realpath
260
pathjoin = _win32_pathjoin
261
normpath = _win32_normpath
262
getcwd = _win32_getcwd
263
mkdtemp = _win32_mkdtemp
264
rename = _win32_rename
266
MIN_ABS_PATHLENGTH = 3
268
def _win32_delete_readonly(function, path, excinfo):
269
"""Error handler for shutil.rmtree function [for win32]
270
Helps to remove files and dirs marked as read-only.
272
type_, value = excinfo[:2]
273
if function in (os.remove, os.rmdir) \
274
and type_ == OSError \
275
and value.errno == errno.EACCES:
276
bzrlib.osutils.make_writable(path)
281
def rmtree(path, ignore_errors=False, onerror=_win32_delete_readonly):
282
"""Replacer for shutil.rmtree: could remove readonly dirs/files"""
283
return shutil.rmtree(path, ignore_errors, onerror)
286
def normalizepath(f):
287
if hasattr(os.path, 'realpath'):
291
[p,e] = os.path.split(f)
292
if e == "" or e == "." or e == "..":
295
return pathjoin(F(p), e)
299
"""Copy a file to a backup.
301
Backups are named in GNU-style, with a ~ suffix.
303
If the file is already a backup, it's not copied.
309
if has_symlinks() and os.path.islink(fn):
310
target = os.readlink(fn)
311
os.symlink(target, bfn)
319
outf = file(bfn, 'wb')
327
"""True if f is an accessible directory."""
329
return S_ISDIR(os.lstat(f)[ST_MODE])
335
"""True if f is a regular file."""
337
return S_ISREG(os.lstat(f)[ST_MODE])
342
"""True if f is a symlink."""
344
return S_ISLNK(os.lstat(f)[ST_MODE])
348
def is_inside(dir, fname):
349
"""True if fname is inside dir.
351
The parameters should typically be passed to osutils.normpath first, so
352
that . and .. and repeated slashes are eliminated, and the separators
353
are canonical for the platform.
355
The empty string as a dir name is taken as top-of-tree and matches
358
>>> is_inside('src', pathjoin('src', 'foo.c'))
360
>>> is_inside('src', 'srccontrol')
362
>>> is_inside('src', pathjoin('src', 'a', 'a', 'a', 'foo.c'))
364
>>> is_inside('foo.c', 'foo.c')
366
>>> is_inside('foo.c', '')
368
>>> is_inside('', 'foo.c')
371
# XXX: Most callers of this can actually do something smarter by
372
# looking at the inventory
382
return fname.startswith(dir)
385
def is_inside_any(dir_list, fname):
386
"""True if fname is inside any of given dirs."""
387
for dirname in dir_list:
388
if is_inside(dirname, fname):
394
def is_inside_or_parent_of_any(dir_list, fname):
395
"""True if fname is a child or a parent of any of the given files."""
396
for dirname in dir_list:
397
if is_inside(dirname, fname) or is_inside(fname, dirname):
403
def pumpfile(fromfile, tofile):
404
"""Copy contents of one file to another."""
407
b = fromfile.read(BUFSIZE)
413
def file_iterator(input_file, readsize=32768):
415
b = input_file.read(readsize)
422
if hasattr(f, 'tell'):
435
def sha_strings(strings):
436
"""Return the sha-1 of concatenation of strings"""
438
map(s.update, strings)
448
def fingerprint_file(f):
453
return {'size': size,
454
'sha1': s.hexdigest()}
457
def compare_files(a, b):
458
"""Returns true if equal in contents"""
469
def local_time_offset(t=None):
470
"""Return offset of local zone from GMT, either at present or at time t."""
471
# python2.3 localtime() can't take None
475
if time.localtime(t).tm_isdst and time.daylight:
478
return -time.timezone
481
def format_date(t, offset=0, timezone='original', date_fmt=None,
483
## TODO: Perhaps a global option to use either universal or local time?
484
## Or perhaps just let people set $TZ?
485
assert isinstance(t, float)
487
if timezone == 'utc':
490
elif timezone == 'original':
493
tt = time.gmtime(t + offset)
494
elif timezone == 'local':
495
tt = time.localtime(t)
496
offset = local_time_offset(t)
498
raise BzrError("unsupported timezone format %r" % timezone,
499
['options are "utc", "original", "local"'])
501
date_fmt = "%a %Y-%m-%d %H:%M:%S"
503
offset_str = ' %+03d%02d' % (offset / 3600, (offset / 60) % 60)
506
return (time.strftime(date_fmt, tt) + offset_str)
509
def compact_date(when):
510
return time.strftime('%Y%m%d%H%M%S', time.gmtime(when))
515
"""Return size of given open file."""
516
return os.fstat(f.fileno())[ST_SIZE]
519
# Define rand_bytes based on platform.
521
# Python 2.4 and later have os.urandom,
522
# but it doesn't work on some arches
524
rand_bytes = os.urandom
525
except (NotImplementedError, AttributeError):
526
# If python doesn't have os.urandom, or it doesn't work,
527
# then try to first pull random data from /dev/urandom
528
if os.path.exists("/dev/urandom"):
529
rand_bytes = file('/dev/urandom', 'rb').read
530
# Otherwise, use this hack as a last resort
532
# not well seeded, but better than nothing
537
s += chr(random.randint(0, 255))
542
ALNUM = '0123456789abcdefghijklmnopqrstuvwxyz'
544
"""Return a random string of num alphanumeric characters
546
The result only contains lowercase chars because it may be used on
547
case-insensitive filesystems.
550
for raw_byte in rand_bytes(num):
551
s += ALNUM[ord(raw_byte) % 36]
555
## TODO: We could later have path objects that remember their list
556
## decomposition (might be too tricksy though.)
559
"""Turn string into list of parts.
565
>>> splitpath('a/./b')
567
>>> splitpath('a/.b')
569
>>> splitpath('a/../b')
570
Traceback (most recent call last):
572
BzrError: sorry, '..' not allowed in path
574
assert isinstance(p, types.StringTypes)
576
# split on either delimiter because people might use either on
578
ps = re.split(r'[\\/]', p)
583
raise BzrError("sorry, %r not allowed in path" % f)
584
elif (f == '.') or (f == ''):
591
assert isinstance(p, list)
593
if (f == '..') or (f == None) or (f == ''):
594
raise BzrError("sorry, %r not allowed in path" % f)
598
@deprecated_function(zero_nine)
599
def appendpath(p1, p2):
603
return pathjoin(p1, p2)
607
"""Split s into lines, but without removing the newline characters."""
608
lines = s.split('\n')
609
result = [line + '\n' for line in lines[:-1]]
611
result.append(lines[-1])
615
def hardlinks_good():
616
return sys.platform not in ('win32', 'cygwin', 'darwin')
619
def link_or_copy(src, dest):
620
"""Hardlink a file, or copy it if it can't be hardlinked."""
621
if not hardlinks_good():
626
except (OSError, IOError), e:
627
if e.errno != errno.EXDEV:
631
def delete_any(full_path):
632
"""Delete a file or directory."""
636
# We may be renaming a dangling inventory id
637
if e.errno not in (errno.EISDIR, errno.EACCES, errno.EPERM):
643
if hasattr(os, 'symlink'):
649
def contains_whitespace(s):
650
"""True if there are any whitespace characters in s."""
651
for ch in string.whitespace:
658
def contains_linebreaks(s):
659
"""True if there is any vertical whitespace in s."""
667
def relpath(base, path):
668
"""Return path relative to base, or raise exception.
670
The path may be either an absolute path or a path relative to the
671
current working directory.
673
os.path.commonprefix (python2.4) has a bad bug that it works just
674
on string prefixes, assuming that '/u' is a prefix of '/u2'. This
678
assert len(base) >= MIN_ABS_PATHLENGTH, ('Length of base must be equal or'
679
' exceed the platform minimum length (which is %d)' %
686
while len(head) >= len(base):
689
head, tail = os.path.split(head)
693
raise PathNotChild(rp, base)
701
def safe_unicode(unicode_or_utf8_string):
702
"""Coerce unicode_or_utf8_string into unicode.
704
If it is unicode, it is returned.
705
Otherwise it is decoded from utf-8. If a decoding error
706
occurs, it is wrapped as a If the decoding fails, the exception is wrapped
707
as a BzrBadParameter exception.
709
if isinstance(unicode_or_utf8_string, unicode):
710
return unicode_or_utf8_string
712
return unicode_or_utf8_string.decode('utf8')
713
except UnicodeDecodeError:
714
raise BzrBadParameterNotUnicode(unicode_or_utf8_string)
717
_platform_normalizes_filenames = False
718
if sys.platform == 'darwin':
719
_platform_normalizes_filenames = True
722
def normalizes_filenames():
723
"""Return True if this platform normalizes unicode filenames.
725
Mac OSX does, Windows/Linux do not.
727
return _platform_normalizes_filenames
730
if _platform_normalizes_filenames:
731
def unicode_filename(path):
732
"""Make sure 'path' is a properly normalized filename.
734
On platforms where the system normalizes filenames (Mac OSX),
735
you can access a file by any path which will normalize
737
Internally, bzr only supports NFC/NFKC normalization, since
738
that is the standard for XML documents.
739
So we return an normalized path, and indicate this has been
742
:return: (path, is_normalized) Return a path which can
743
access the file, and whether or not this path is
746
return unicodedata.normalize('NFKC', path), True
748
def unicode_filename(path):
749
"""Make sure 'path' is a properly normalized filename.
751
On platforms where the system does not normalize filenames
752
(Windows, Linux), you have to access a file by its exact path.
753
Internally, bzr only supports NFC/NFKC normalization, since
754
that is the standard for XML documents.
755
So we return the original path, and indicate if this is
758
:return: (path, is_normalized) Return a path which can
759
access the file, and whether or not this path is
762
return path, unicodedata.normalize('NFKC', path) == path
765
def terminal_width():
766
"""Return estimated terminal width."""
767
if sys.platform == 'win32':
768
import bzrlib.win32console
769
return bzrlib.win32console.get_console_size()[0]
772
import struct, fcntl, termios
773
s = struct.pack('HHHH', 0, 0, 0, 0)
774
x = fcntl.ioctl(1, termios.TIOCGWINSZ, s)
775
width = struct.unpack('HHHH', x)[1]
780
width = int(os.environ['COLUMNS'])
788
def supports_executable():
789
return sys.platform != "win32"
792
_validWin32PathRE = re.compile(r'^([A-Za-z]:[/\\])?[^:<>*"?\|]*$')
795
def check_legal_path(path):
796
"""Check whether the supplied path is legal.
797
This is only required on Windows, so we don't test on other platforms
800
if sys.platform != "win32":
802
if _validWin32PathRE.match(path) is None:
803
raise IllegalPath(path)
807
"""Yield data about all the directories in a tree.
809
This yields all the data about the contents of a directory at a time.
810
After each directory has been yielded, if the caller has mutated the list
811
to exclude some directories, they are then not descended into.
813
The data yielded is of the form:
814
[(relpath, basename, kind, lstat, path_from_top), ...]
816
:return: an iterator over the dirs.
820
_directory = _directory_kind
822
pending = [("", "", _directory, None, top)]
825
currentdir = pending.pop()
826
# 0 - relpath, 1- basename, 2- kind, 3- stat, 4-toppath
829
relroot = currentdir[0] + '/'
832
for name in sorted(_listdir(top)):
833
abspath = top + '/' + name
834
statvalue = lstat(abspath)
835
dirblock.append ((relroot + name, name, file_kind_from_stat_mode(statvalue.st_mode), statvalue, abspath))
837
# push the user specified dirs from dirblock
838
for dir in reversed(dirblock):
839
if dir[2] == _directory:
added
removed
bzrlib/osutils.py
