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):
112
return _mapper(_lstat(f).st_mode)
114
if getattr(e, 'errno', None) == errno.ENOENT:
115
raise bzrlib.errors.NoSuchFile(f)
118
return _mapper(_lstat(f).st_mode)
121
def kind_marker(kind):
124
elif kind == _directory_kind:
126
elif kind == 'symlink':
129
raise BzrError('invalid file kind %r' % kind)
131
lexists = getattr(os.path, 'lexists', None)
135
if hasattr(os, 'lstat'):
141
if e.errno == errno.ENOENT:
144
raise BzrError("lstat/stat of (%r): %r" % (f, e))
147
def fancy_rename(old, new, rename_func, unlink_func):
148
"""A fancy rename, when you don't have atomic rename.
150
:param old: The old path, to rename from
151
:param new: The new path, to rename to
152
:param rename_func: The potentially non-atomic rename function
153
:param unlink_func: A way to delete the target file if the full rename succeeds
156
# sftp rename doesn't allow overwriting, so play tricks:
158
base = os.path.basename(new)
159
dirname = os.path.dirname(new)
160
tmp_name = u'tmp.%s.%.9f.%d.%s' % (base, time.time(), os.getpid(), rand_chars(10))
161
tmp_name = pathjoin(dirname, tmp_name)
163
# Rename the file out of the way, but keep track if it didn't exist
164
# We don't want to grab just any exception
165
# something like EACCES should prevent us from continuing
166
# The downside is that the rename_func has to throw an exception
167
# with an errno = ENOENT, or NoSuchFile
170
rename_func(new, tmp_name)
171
except (NoSuchFile,), e:
174
# RBC 20060103 abstraction leakage: the paramiko SFTP clients rename
175
# function raises an IOError with errno == None when a rename fails.
176
# This then gets caught here.
177
if e.errno not in (None, errno.ENOENT, errno.ENOTDIR):
180
if (not hasattr(e, 'errno')
181
or e.errno not in (errno.ENOENT, errno.ENOTDIR)):
188
# This may throw an exception, in which case success will
190
rename_func(old, new)
194
# If the file used to exist, rename it back into place
195
# otherwise just delete it from the tmp location
197
unlink_func(tmp_name)
199
rename_func(tmp_name, new)
202
# In Python 2.4.2 and older, os.path.abspath and os.path.realpath
203
# choke on a Unicode string containing a relative path if
204
# os.getcwd() returns a non-sys.getdefaultencoding()-encoded
206
_fs_enc = sys.getfilesystemencoding()
207
def _posix_abspath(path):
208
return os.path.abspath(path.encode(_fs_enc)).decode(_fs_enc)
209
# jam 20060426 This is another possibility which mimics
210
# os.path.abspath, only uses unicode characters instead
211
# if not os.path.isabs(path):
212
# return os.path.join(os.getcwdu(), path)
216
def _posix_realpath(path):
217
return os.path.realpath(path.encode(_fs_enc)).decode(_fs_enc)
220
def _win32_abspath(path):
221
return _nt_abspath(path.encode(_fs_enc)).decode(_fs_enc).replace('\\', '/')
224
def _win32_realpath(path):
225
return _nt_realpath(path.encode(_fs_enc)).decode(_fs_enc).replace('\\', '/')
228
def _win32_pathjoin(*args):
229
return _nt_join(*args).replace('\\', '/')
232
def _win32_normpath(path):
233
return _nt_normpath(path).replace('\\', '/')
237
return os.getcwdu().replace('\\', '/')
240
def _win32_mkdtemp(*args, **kwargs):
241
return tempfile.mkdtemp(*args, **kwargs).replace('\\', '/')
244
def _win32_rename(old, new):
245
fancy_rename(old, new, rename_func=os.rename, unlink_func=os.unlink)
248
# Default is to just use the python builtins, but these can be rebound on
249
# particular platforms.
250
abspath = _posix_abspath
251
realpath = _posix_realpath
252
pathjoin = os.path.join
253
normpath = os.path.normpath
255
mkdtemp = tempfile.mkdtemp
257
dirname = os.path.dirname
258
basename = os.path.basename
259
rmtree = shutil.rmtree
261
MIN_ABS_PATHLENGTH = 1
264
if sys.platform == 'win32':
265
abspath = _win32_abspath
266
realpath = _win32_realpath
267
pathjoin = _win32_pathjoin
268
normpath = _win32_normpath
269
getcwd = _win32_getcwd
270
mkdtemp = _win32_mkdtemp
271
rename = _win32_rename
273
MIN_ABS_PATHLENGTH = 3
275
def _win32_delete_readonly(function, path, excinfo):
276
"""Error handler for shutil.rmtree function [for win32]
277
Helps to remove files and dirs marked as read-only.
279
type_, value = excinfo[:2]
280
if function in (os.remove, os.rmdir) \
281
and type_ == OSError \
282
and value.errno == errno.EACCES:
283
bzrlib.osutils.make_writable(path)
288
def rmtree(path, ignore_errors=False, onerror=_win32_delete_readonly):
289
"""Replacer for shutil.rmtree: could remove readonly dirs/files"""
290
return shutil.rmtree(path, ignore_errors, onerror)
293
def normalizepath(f):
294
if hasattr(os.path, 'realpath'):
298
[p,e] = os.path.split(f)
299
if e == "" or e == "." or e == "..":
302
return pathjoin(F(p), e)
306
"""Copy a file to a backup.
308
Backups are named in GNU-style, with a ~ suffix.
310
If the file is already a backup, it's not copied.
316
if has_symlinks() and os.path.islink(fn):
317
target = os.readlink(fn)
318
os.symlink(target, bfn)
326
outf = file(bfn, 'wb')
334
"""True if f is an accessible directory."""
336
return S_ISDIR(os.lstat(f)[ST_MODE])
342
"""True if f is a regular file."""
344
return S_ISREG(os.lstat(f)[ST_MODE])
349
"""True if f is a symlink."""
351
return S_ISLNK(os.lstat(f)[ST_MODE])
355
def is_inside(dir, fname):
356
"""True if fname is inside dir.
358
The parameters should typically be passed to osutils.normpath first, so
359
that . and .. and repeated slashes are eliminated, and the separators
360
are canonical for the platform.
362
The empty string as a dir name is taken as top-of-tree and matches
365
>>> is_inside('src', pathjoin('src', 'foo.c'))
367
>>> is_inside('src', 'srccontrol')
369
>>> is_inside('src', pathjoin('src', 'a', 'a', 'a', 'foo.c'))
371
>>> is_inside('foo.c', 'foo.c')
373
>>> is_inside('foo.c', '')
375
>>> is_inside('', 'foo.c')
378
# XXX: Most callers of this can actually do something smarter by
379
# looking at the inventory
389
return fname.startswith(dir)
392
def is_inside_any(dir_list, fname):
393
"""True if fname is inside any of given dirs."""
394
for dirname in dir_list:
395
if is_inside(dirname, fname):
401
def is_inside_or_parent_of_any(dir_list, fname):
402
"""True if fname is a child or a parent of any of the given files."""
403
for dirname in dir_list:
404
if is_inside(dirname, fname) or is_inside(fname, dirname):
410
def pumpfile(fromfile, tofile):
411
"""Copy contents of one file to another."""
414
b = fromfile.read(BUFSIZE)
420
def file_iterator(input_file, readsize=32768):
422
b = input_file.read(readsize)
429
if hasattr(f, 'tell'):
442
def sha_strings(strings):
443
"""Return the sha-1 of concatenation of strings"""
445
map(s.update, strings)
455
def fingerprint_file(f):
460
return {'size': size,
461
'sha1': s.hexdigest()}
464
def compare_files(a, b):
465
"""Returns true if equal in contents"""
476
def local_time_offset(t=None):
477
"""Return offset of local zone from GMT, either at present or at time t."""
478
# python2.3 localtime() can't take None
482
if time.localtime(t).tm_isdst and time.daylight:
485
return -time.timezone
488
def format_date(t, offset=0, timezone='original', date_fmt=None,
490
## TODO: Perhaps a global option to use either universal or local time?
491
## Or perhaps just let people set $TZ?
492
assert isinstance(t, float)
494
if timezone == 'utc':
497
elif timezone == 'original':
500
tt = time.gmtime(t + offset)
501
elif timezone == 'local':
502
tt = time.localtime(t)
503
offset = local_time_offset(t)
505
raise BzrError("unsupported timezone format %r" % timezone,
506
['options are "utc", "original", "local"'])
508
date_fmt = "%a %Y-%m-%d %H:%M:%S"
510
offset_str = ' %+03d%02d' % (offset / 3600, (offset / 60) % 60)
513
return (time.strftime(date_fmt, tt) + offset_str)
516
def compact_date(when):
517
return time.strftime('%Y%m%d%H%M%S', time.gmtime(when))
522
"""Return size of given open file."""
523
return os.fstat(f.fileno())[ST_SIZE]
526
# Define rand_bytes based on platform.
528
# Python 2.4 and later have os.urandom,
529
# but it doesn't work on some arches
531
rand_bytes = os.urandom
532
except (NotImplementedError, AttributeError):
533
# If python doesn't have os.urandom, or it doesn't work,
534
# then try to first pull random data from /dev/urandom
535
if os.path.exists("/dev/urandom"):
536
rand_bytes = file('/dev/urandom', 'rb').read
537
# Otherwise, use this hack as a last resort
539
# not well seeded, but better than nothing
544
s += chr(random.randint(0, 255))
549
ALNUM = '0123456789abcdefghijklmnopqrstuvwxyz'
551
"""Return a random string of num alphanumeric characters
553
The result only contains lowercase chars because it may be used on
554
case-insensitive filesystems.
557
for raw_byte in rand_bytes(num):
558
s += ALNUM[ord(raw_byte) % 36]
562
## TODO: We could later have path objects that remember their list
563
## decomposition (might be too tricksy though.)
566
"""Turn string into list of parts.
572
>>> splitpath('a/./b')
574
>>> splitpath('a/.b')
576
>>> splitpath('a/../b')
577
Traceback (most recent call last):
579
BzrError: sorry, '..' not allowed in path
581
assert isinstance(p, types.StringTypes)
583
# split on either delimiter because people might use either on
585
ps = re.split(r'[\\/]', p)
590
raise BzrError("sorry, %r not allowed in path" % f)
591
elif (f == '.') or (f == ''):
598
assert isinstance(p, list)
600
if (f == '..') or (f == None) or (f == ''):
601
raise BzrError("sorry, %r not allowed in path" % f)
605
@deprecated_function(zero_nine)
606
def appendpath(p1, p2):
610
return pathjoin(p1, p2)
614
"""Split s into lines, but without removing the newline characters."""
615
lines = s.split('\n')
616
result = [line + '\n' for line in lines[:-1]]
618
result.append(lines[-1])
622
def hardlinks_good():
623
return sys.platform not in ('win32', 'cygwin', 'darwin')
626
def link_or_copy(src, dest):
627
"""Hardlink a file, or copy it if it can't be hardlinked."""
628
if not hardlinks_good():
633
except (OSError, IOError), e:
634
if e.errno != errno.EXDEV:
638
def delete_any(full_path):
639
"""Delete a file or directory."""
643
# We may be renaming a dangling inventory id
644
if e.errno not in (errno.EISDIR, errno.EACCES, errno.EPERM):
650
if hasattr(os, 'symlink'):
656
def contains_whitespace(s):
657
"""True if there are any whitespace characters in s."""
658
for ch in string.whitespace:
665
def contains_linebreaks(s):
666
"""True if there is any vertical whitespace in s."""
674
def relpath(base, path):
675
"""Return path relative to base, or raise exception.
677
The path may be either an absolute path or a path relative to the
678
current working directory.
680
os.path.commonprefix (python2.4) has a bad bug that it works just
681
on string prefixes, assuming that '/u' is a prefix of '/u2'. This
685
assert len(base) >= MIN_ABS_PATHLENGTH, ('Length of base must be equal or'
686
' exceed the platform minimum length (which is %d)' %
693
while len(head) >= len(base):
696
head, tail = os.path.split(head)
700
raise PathNotChild(rp, base)
708
def safe_unicode(unicode_or_utf8_string):
709
"""Coerce unicode_or_utf8_string into unicode.
711
If it is unicode, it is returned.
712
Otherwise it is decoded from utf-8. If a decoding error
713
occurs, it is wrapped as a If the decoding fails, the exception is wrapped
714
as a BzrBadParameter exception.
716
if isinstance(unicode_or_utf8_string, unicode):
717
return unicode_or_utf8_string
719
return unicode_or_utf8_string.decode('utf8')
720
except UnicodeDecodeError:
721
raise BzrBadParameterNotUnicode(unicode_or_utf8_string)
724
_platform_normalizes_filenames = False
725
if sys.platform == 'darwin':
726
_platform_normalizes_filenames = True
729
def normalizes_filenames():
730
"""Return True if this platform normalizes unicode filenames.
732
Mac OSX does, Windows/Linux do not.
734
return _platform_normalizes_filenames
737
if _platform_normalizes_filenames:
738
def unicode_filename(path):
739
"""Make sure 'path' is a properly normalized filename.
741
On platforms where the system normalizes filenames (Mac OSX),
742
you can access a file by any path which will normalize
744
Internally, bzr only supports NFC/NFKC normalization, since
745
that is the standard for XML documents.
746
So we return an normalized path, and indicate this has been
749
:return: (path, is_normalized) Return a path which can
750
access the file, and whether or not this path is
753
return unicodedata.normalize('NFKC', path), True
755
def unicode_filename(path):
756
"""Make sure 'path' is a properly normalized filename.
758
On platforms where the system does not normalize filenames
759
(Windows, Linux), you have to access a file by its exact path.
760
Internally, bzr only supports NFC/NFKC normalization, since
761
that is the standard for XML documents.
762
So we return the original path, and indicate if this is
765
:return: (path, is_normalized) Return a path which can
766
access the file, and whether or not this path is
769
return path, unicodedata.normalize('NFKC', path) == path
772
def terminal_width():
773
"""Return estimated terminal width."""
774
if sys.platform == 'win32':
775
import bzrlib.win32console
776
return bzrlib.win32console.get_console_size()[0]
779
import struct, fcntl, termios
780
s = struct.pack('HHHH', 0, 0, 0, 0)
781
x = fcntl.ioctl(1, termios.TIOCGWINSZ, s)
782
width = struct.unpack('HHHH', x)[1]
787
width = int(os.environ['COLUMNS'])
795
def supports_executable():
796
return sys.platform != "win32"
799
_validWin32PathRE = re.compile(r'^([A-Za-z]:[/\\])?[^:<>*"?\|]*$')
802
def check_legal_path(path):
803
"""Check whether the supplied path is legal.
804
This is only required on Windows, so we don't test on other platforms
807
if sys.platform != "win32":
809
if _validWin32PathRE.match(path) is None:
810
raise IllegalPath(path)
813
def walkdirs(top, prefix=""):
814
"""Yield data about all the directories in a tree.
816
This yields all the data about the contents of a directory at a time.
817
After each directory has been yielded, if the caller has mutated the list
818
to exclude some directories, they are then not descended into.
820
The data yielded is of the form:
821
[(relpath, basename, kind, lstat, path_from_top), ...]
823
:param prefix: Treat prefix as the initial root of the tree, rather than ""
824
:return: an iterator over the dirs.
828
_directory = _directory_kind
830
pending = [(prefix, "", _directory, None, top)]
833
currentdir = pending.pop()
834
# 0 - relpath, 1- basename, 2- kind, 3- stat, 4-toppath
837
relroot = currentdir[0] + '/'
840
for name in sorted(_listdir(top)):
841
abspath = top + '/' + name
842
statvalue = lstat(abspath)
843
dirblock.append ((relroot + name, name, file_kind_from_stat_mode(statvalue.st_mode), statvalue, abspath))
845
# push the user specified dirs from dirblock
846
for dir in reversed(dirblock):
847
if dir[2] == _directory:
added
removed
bzrlib/osutils.py
