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 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)
22
from cStringIO import StringIO
33
from ntpath import (abspath as _nt_abspath,
35
normpath as _nt_normpath,
36
realpath as _nt_realpath,
40
from bzrlib.errors import (BzrError,
41
BzrBadParameterNotUnicode,
46
from bzrlib.trace import mutter
49
def make_readonly(filename):
50
"""Make a filename read-only."""
51
mod = os.stat(filename).st_mode
53
os.chmod(filename, mod)
56
def make_writable(filename):
57
mod = os.stat(filename).st_mode
59
os.chmod(filename, mod)
66
"""Return a quoted filename filename
68
This previously used backslash quoting, but that works poorly on
70
# TODO: I'm not really sure this is the best format either.x
73
_QUOTE_RE = re.compile(r'([^a-zA-Z0-9.,:/\\_~-])')
75
if _QUOTE_RE.search(f):
82
mode = os.lstat(f)[ST_MODE]
101
def kind_marker(kind):
104
elif kind == 'directory':
106
elif kind == 'symlink':
109
raise BzrError('invalid file kind %r' % kind)
112
if hasattr(os.path, 'lexists'):
113
return os.path.lexists(f)
115
if hasattr(os, 'lstat'):
121
if e.errno == errno.ENOENT:
124
raise BzrError("lstat/stat of (%r): %r" % (f, e))
126
def fancy_rename(old, new, rename_func, unlink_func):
127
"""A fancy rename, when you don't have atomic rename.
129
:param old: The old path, to rename from
130
:param new: The new path, to rename to
131
:param rename_func: The potentially non-atomic rename function
132
:param unlink_func: A way to delete the target file if the full rename succeeds
135
# sftp rename doesn't allow overwriting, so play tricks:
137
base = os.path.basename(new)
138
dirname = os.path.dirname(new)
139
tmp_name = u'tmp.%s.%.9f.%d.%s' % (base, time.time(), os.getpid(), rand_chars(10))
140
tmp_name = pathjoin(dirname, tmp_name)
142
# Rename the file out of the way, but keep track if it didn't exist
143
# We don't want to grab just any exception
144
# something like EACCES should prevent us from continuing
145
# The downside is that the rename_func has to throw an exception
146
# with an errno = ENOENT, or NoSuchFile
149
rename_func(new, tmp_name)
150
except (NoSuchFile,), e:
153
# RBC 20060103 abstraction leakage: the paramiko SFTP clients rename
154
# function raises an IOError with errno == None when a rename fails.
155
# This then gets caught here.
156
if e.errno not in (None, errno.ENOENT, errno.ENOTDIR):
159
if (not hasattr(e, 'errno')
160
or e.errno not in (errno.ENOENT, errno.ENOTDIR)):
167
# This may throw an exception, in which case success will
169
rename_func(old, new)
173
# If the file used to exist, rename it back into place
174
# otherwise just delete it from the tmp location
176
unlink_func(tmp_name)
178
rename_func(tmp_name, new)
181
# In Python 2.4.2 and older, os.path.abspath and os.path.realpath
182
# choke on a Unicode string containing a relative path if
183
# os.getcwd() returns a non-sys.getdefaultencoding()-encoded
185
_fs_enc = sys.getfilesystemencoding()
186
def _posix_abspath(path):
187
return os.path.abspath(path.encode(_fs_enc)).decode(_fs_enc)
188
# jam 20060426 This is another possibility which mimics
189
# os.path.abspath, only uses unicode characters instead
190
# if not os.path.isabs(path):
191
# return os.path.join(os.getcwdu(), path)
195
def _posix_realpath(path):
196
return os.path.realpath(path.encode(_fs_enc)).decode(_fs_enc)
199
def _win32_abspath(path):
200
return _nt_abspath(path.encode(_fs_enc)).decode(_fs_enc).replace('\\', '/')
203
def _win32_realpath(path):
204
return _nt_realpath(path.encode(_fs_enc)).decode(_fs_enc).replace('\\', '/')
207
def _win32_pathjoin(*args):
208
return _nt_join(*args).replace('\\', '/')
211
def _win32_normpath(path):
212
return _nt_normpath(path).replace('\\', '/')
216
return os.getcwdu().replace('\\', '/')
219
def _win32_mkdtemp(*args, **kwargs):
220
return tempfile.mkdtemp(*args, **kwargs).replace('\\', '/')
223
def _win32_rename(old, new):
224
fancy_rename(old, new, rename_func=os.rename, unlink_func=os.unlink)
227
# Default is to just use the python builtins
228
abspath = _posix_abspath
229
realpath = _posix_realpath
230
pathjoin = os.path.join
231
normpath = os.path.normpath
233
mkdtemp = tempfile.mkdtemp
235
dirname = os.path.dirname
236
basename = os.path.basename
238
MIN_ABS_PATHLENGTH = 1
241
if sys.platform == 'win32':
242
abspath = _win32_abspath
243
realpath = _win32_realpath
244
pathjoin = _win32_pathjoin
245
normpath = _win32_normpath
246
getcwd = _win32_getcwd
247
mkdtemp = _win32_mkdtemp
248
rename = _win32_rename
250
MIN_ABS_PATHLENGTH = 3
253
def normalizepath(f):
254
if hasattr(os.path, 'realpath'):
258
[p,e] = os.path.split(f)
259
if e == "" or e == "." or e == "..":
262
return pathjoin(F(p), e)
266
"""Copy a file to a backup.
268
Backups are named in GNU-style, with a ~ suffix.
270
If the file is already a backup, it's not copied.
276
if has_symlinks() and os.path.islink(fn):
277
target = os.readlink(fn)
278
os.symlink(target, bfn)
286
outf = file(bfn, 'wb')
294
"""True if f is an accessible directory."""
296
return S_ISDIR(os.lstat(f)[ST_MODE])
302
"""True if f is a regular file."""
304
return S_ISREG(os.lstat(f)[ST_MODE])
309
"""True if f is a symlink."""
311
return S_ISLNK(os.lstat(f)[ST_MODE])
315
def is_inside(dir, fname):
316
"""True if fname is inside dir.
318
The parameters should typically be passed to osutils.normpath first, so
319
that . and .. and repeated slashes are eliminated, and the separators
320
are canonical for the platform.
322
The empty string as a dir name is taken as top-of-tree and matches
325
>>> is_inside('src', pathjoin('src', 'foo.c'))
327
>>> is_inside('src', 'srccontrol')
329
>>> is_inside('src', pathjoin('src', 'a', 'a', 'a', 'foo.c'))
331
>>> is_inside('foo.c', 'foo.c')
333
>>> is_inside('foo.c', '')
335
>>> is_inside('', 'foo.c')
338
# XXX: Most callers of this can actually do something smarter by
339
# looking at the inventory
349
return fname.startswith(dir)
352
def is_inside_any(dir_list, fname):
353
"""True if fname is inside any of given dirs."""
354
for dirname in dir_list:
355
if is_inside(dirname, fname):
361
def pumpfile(fromfile, tofile):
362
"""Copy contents of one file to another."""
365
b = fromfile.read(BUFSIZE)
371
def file_iterator(input_file, readsize=32768):
373
b = input_file.read(readsize)
380
if hasattr(f, 'tell'):
393
def sha_strings(strings):
394
"""Return the sha-1 of concatenation of strings"""
396
map(s.update, strings)
406
def fingerprint_file(f):
411
return {'size': size,
412
'sha1': s.hexdigest()}
415
def compare_files(a, b):
416
"""Returns true if equal in contents"""
427
def local_time_offset(t=None):
428
"""Return offset of local zone from GMT, either at present or at time t."""
429
# python2.3 localtime() can't take None
433
if time.localtime(t).tm_isdst and time.daylight:
436
return -time.timezone
439
def format_date(t, offset=0, timezone='original', date_fmt=None,
441
## TODO: Perhaps a global option to use either universal or local time?
442
## Or perhaps just let people set $TZ?
443
assert isinstance(t, float)
445
if timezone == 'utc':
448
elif timezone == 'original':
451
tt = time.gmtime(t + offset)
452
elif timezone == 'local':
453
tt = time.localtime(t)
454
offset = local_time_offset(t)
456
raise BzrError("unsupported timezone format %r" % timezone,
457
['options are "utc", "original", "local"'])
459
date_fmt = "%a %Y-%m-%d %H:%M:%S"
461
offset_str = ' %+03d%02d' % (offset / 3600, (offset / 60) % 60)
464
return (time.strftime(date_fmt, tt) + offset_str)
467
def compact_date(when):
468
return time.strftime('%Y%m%d%H%M%S', time.gmtime(when))
473
"""Return size of given open file."""
474
return os.fstat(f.fileno())[ST_SIZE]
477
# Define rand_bytes based on platform.
479
# Python 2.4 and later have os.urandom,
480
# but it doesn't work on some arches
482
rand_bytes = os.urandom
483
except (NotImplementedError, AttributeError):
484
# If python doesn't have os.urandom, or it doesn't work,
485
# then try to first pull random data from /dev/urandom
486
if os.path.exists("/dev/urandom"):
487
rand_bytes = file('/dev/urandom', 'rb').read
488
# Otherwise, use this hack as a last resort
490
# not well seeded, but better than nothing
495
s += chr(random.randint(0, 255))
500
ALNUM = '0123456789abcdefghijklmnopqrstuvwxyz'
502
"""Return a random string of num alphanumeric characters
504
The result only contains lowercase chars because it may be used on
505
case-insensitive filesystems.
508
for raw_byte in rand_bytes(num):
509
s += ALNUM[ord(raw_byte) % 36]
513
## TODO: We could later have path objects that remember their list
514
## decomposition (might be too tricksy though.)
517
"""Turn string into list of parts.
523
>>> splitpath('a/./b')
525
>>> splitpath('a/.b')
527
>>> splitpath('a/../b')
528
Traceback (most recent call last):
530
BzrError: sorry, '..' not allowed in path
532
assert isinstance(p, types.StringTypes)
534
# split on either delimiter because people might use either on
536
ps = re.split(r'[\\/]', p)
541
raise BzrError("sorry, %r not allowed in path" % f)
542
elif (f == '.') or (f == ''):
549
assert isinstance(p, list)
551
if (f == '..') or (f == None) or (f == ''):
552
raise BzrError("sorry, %r not allowed in path" % f)
556
def appendpath(p1, p2):
560
return pathjoin(p1, p2)
564
"""Split s into lines, but without removing the newline characters."""
565
lines = s.split('\n')
566
result = [line + '\n' for line in lines[:-1]]
568
result.append(lines[-1])
572
def hardlinks_good():
573
return sys.platform not in ('win32', 'cygwin', 'darwin')
576
def link_or_copy(src, dest):
577
"""Hardlink a file, or copy it if it can't be hardlinked."""
578
if not hardlinks_good():
583
except (OSError, IOError), e:
584
if e.errno != errno.EXDEV:
588
def delete_any(full_path):
589
"""Delete a file or directory."""
593
# We may be renaming a dangling inventory id
594
if e.errno not in (errno.EISDIR, errno.EACCES, errno.EPERM):
600
if hasattr(os, 'symlink'):
606
def contains_whitespace(s):
607
"""True if there are any whitespace characters in s."""
608
for ch in string.whitespace:
615
def contains_linebreaks(s):
616
"""True if there is any vertical whitespace in s."""
624
def relpath(base, path):
625
"""Return path relative to base, or raise exception.
627
The path may be either an absolute path or a path relative to the
628
current working directory.
630
os.path.commonprefix (python2.4) has a bad bug that it works just
631
on string prefixes, assuming that '/u' is a prefix of '/u2'. This
635
assert len(base) >= MIN_ABS_PATHLENGTH, ('Length of base must be equal or'
636
' exceed the platform minimum length (which is %d)' %
643
while len(head) >= len(base):
646
head, tail = os.path.split(head)
650
raise PathNotChild(rp, base)
658
def safe_unicode(unicode_or_utf8_string):
659
"""Coerce unicode_or_utf8_string into unicode.
661
If it is unicode, it is returned.
662
Otherwise it is decoded from utf-8. If a decoding error
663
occurs, it is wrapped as a If the decoding fails, the exception is wrapped
664
as a BzrBadParameter exception.
666
if isinstance(unicode_or_utf8_string, unicode):
667
return unicode_or_utf8_string
669
return unicode_or_utf8_string.decode('utf8')
670
except UnicodeDecodeError:
671
raise BzrBadParameterNotUnicode(unicode_or_utf8_string)
674
_platform_normalizes_filenames = False
675
if sys.platform == 'darwin':
676
_platform_normalizes_filenames = True
679
def normalizes_filenames():
680
"""Return True if this platform normalizes unicode filenames.
682
Mac OSX does, Windows/Linux do not.
684
return _platform_normalizes_filenames
687
if _platform_normalizes_filenames:
688
def unicode_filename(path):
689
"""Make sure 'path' is a properly normalized filename.
691
On platforms where the system normalizes filenames (Mac OSX),
692
you can access a file by any path which will normalize
694
Internally, bzr only supports NFC/NFKC normalization, since
695
that is the standard for XML documents.
696
So we return an normalized path, and indicate this has been
699
:return: (path, is_normalized) Return a path which can
700
access the file, and whether or not this path is
703
return unicodedata.normalize('NFKC', path), True
705
def unicode_filename(path):
706
"""Make sure 'path' is a properly normalized filename.
708
On platforms where the system does not normalize filenames
709
(Windows, Linux), you have to access a file by its exact path.
710
Internally, bzr only supports NFC/NFKC normalization, since
711
that is the standard for XML documents.
712
So we return the original path, and indicate if this is
715
:return: (path, is_normalized) Return a path which can
716
access the file, and whether or not this path is
719
return path, unicodedata.normalize('NFKC', path) == path
722
def terminal_width():
723
"""Return estimated terminal width."""
725
# TODO: Do something smart on Windows?
727
# TODO: Is there anything that gets a better update when the window
728
# is resized while the program is running? We could use the Python termcap
731
return int(os.environ['COLUMNS'])
732
except (IndexError, KeyError, ValueError):
735
def supports_executable():
736
return sys.platform != "win32"
739
_validWin32PathRE = re.compile(r'^([A-Za-z]:[/\\])?[^:<>*"?\|]*$')
742
def check_legal_path(path):
743
"""Check whether the supplied path is legal.
744
This is only required on Windows, so we don't test on other platforms
747
if sys.platform != "win32":
749
if _validWin32PathRE.match(path) is None:
750
raise IllegalPath(path)
added
removed
bzrlib/osutils.py
