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,
32
from bzrlib.errors import BzrError
53
33
from bzrlib.trace import mutter
88
_directory_kind = 'directory'
91
stat.S_IFDIR:_directory_kind,
92
stat.S_IFCHR:'chardev',
96
stat.S_IFLNK:'symlink',
97
stat.S_IFSOCK:'socket',
101
def file_kind_from_stat_mode(stat_mode, _formats=_formats, _unknown='unknown'):
102
"""Generate a file kind from a stat mode. This is used in walkdirs.
104
Its performance is critical: Do not mutate without careful benchmarking.
107
return _formats[stat_mode & 0170000]
112
def file_kind(f, _lstat=os.lstat, _mapper=file_kind_from_stat_mode):
114
return _mapper(_lstat(f).st_mode)
116
if getattr(e, 'errno', None) == errno.ENOENT:
117
raise bzrlib.errors.NoSuchFile(f)
69
mode = os.lstat(f)[ST_MODE]
121
88
def kind_marker(kind):
122
89
if kind == 'file':
124
elif kind == _directory_kind:
91
elif kind == 'directory':
126
93
elif kind == 'symlink':
129
96
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
# jam 20060426 rather than encoding to fsencoding
209
# copy posixpath.abspath, but use os.getcwdu instead
210
if not posixpath.isabs(path):
211
path = posixpath.join(getcwd(), path)
212
return posixpath.normpath(path)
215
def _posix_realpath(path):
216
return posixpath.realpath(path.encode(_fs_enc)).decode(_fs_enc)
219
def _win32_fixdrive(path):
220
"""Force drive letters to be consistent.
222
win32 is inconsistent whether it returns lower or upper case
223
and even if it was consistent the user might type the other
224
so we force it to uppercase
225
running python.exe under cmd.exe return capital C:\\
226
running win32 python inside a cygwin shell returns lowercase c:\\
228
drive, path = _nt_splitdrive(path)
229
return drive.upper() + path
232
def _win32_abspath(path):
233
# Real _nt_abspath doesn't have a problem with a unicode cwd
234
return _win32_fixdrive(_nt_abspath(unicode(path)).replace('\\', '/'))
237
def _win32_realpath(path):
238
# Real _nt_realpath doesn't have a problem with a unicode cwd
239
return _win32_fixdrive(_nt_realpath(unicode(path)).replace('\\', '/'))
242
def _win32_pathjoin(*args):
243
return _nt_join(*args).replace('\\', '/')
246
def _win32_normpath(path):
247
return _win32_fixdrive(_nt_normpath(unicode(path)).replace('\\', '/'))
251
return _win32_fixdrive(os.getcwdu().replace('\\', '/'))
254
def _win32_mkdtemp(*args, **kwargs):
255
return _win32_fixdrive(tempfile.mkdtemp(*args, **kwargs).replace('\\', '/'))
258
def _win32_rename(old, new):
259
fancy_rename(old, new, rename_func=os.rename, unlink_func=os.unlink)
262
# Default is to just use the python builtins, but these can be rebound on
263
# particular platforms.
264
abspath = _posix_abspath
265
realpath = _posix_realpath
266
pathjoin = os.path.join
267
normpath = os.path.normpath
269
mkdtemp = tempfile.mkdtemp
271
dirname = os.path.dirname
272
basename = os.path.basename
273
rmtree = shutil.rmtree
275
MIN_ABS_PATHLENGTH = 1
278
if sys.platform == 'win32':
279
abspath = _win32_abspath
280
realpath = _win32_realpath
281
pathjoin = _win32_pathjoin
282
normpath = _win32_normpath
283
getcwd = _win32_getcwd
284
mkdtemp = _win32_mkdtemp
285
rename = _win32_rename
287
MIN_ABS_PATHLENGTH = 3
289
def _win32_delete_readonly(function, path, excinfo):
290
"""Error handler for shutil.rmtree function [for win32]
291
Helps to remove files and dirs marked as read-only.
293
type_, value = excinfo[:2]
294
if function in (os.remove, os.rmdir) \
295
and type_ == OSError \
296
and value.errno == errno.EACCES:
297
bzrlib.osutils.make_writable(path)
302
def rmtree(path, ignore_errors=False, onerror=_win32_delete_readonly):
303
"""Replacer for shutil.rmtree: could remove readonly dirs/files"""
304
return shutil.rmtree(path, ignore_errors, onerror)
307
def get_terminal_encoding():
308
"""Find the best encoding for printing to the screen.
310
This attempts to check both sys.stdout and sys.stdin to see
311
what encoding they are in, and if that fails it falls back to
312
bzrlib.user_encoding.
313
The problem is that on Windows, locale.getpreferredencoding()
314
is not the same encoding as that used by the console:
315
http://mail.python.org/pipermail/python-list/2003-May/162357.html
317
On my standard US Windows XP, the preferred encoding is
318
cp1252, but the console is cp437
320
output_encoding = getattr(sys.stdout, 'encoding', None)
321
if not output_encoding:
322
input_encoding = getattr(sys.stdin, 'encoding', None)
323
if not input_encoding:
324
output_encoding = bzrlib.user_encoding
325
mutter('encoding stdout as bzrlib.user_encoding %r', output_encoding)
327
output_encoding = input_encoding
328
mutter('encoding stdout as sys.stdin encoding %r', output_encoding)
330
mutter('encoding stdout as sys.stdout encoding %r', output_encoding)
331
return output_encoding
100
if hasattr(os, 'lstat'):
106
if e.errno == errno.ENOENT:
109
raise BzrError("lstat/stat of (%r): %r" % (f, e))
334
111
def normalizepath(f):
335
112
if hasattr(os.path, 'realpath'):
339
116
[p,e] = os.path.split(f)
340
117
if e == "" or e == "." or e == "..":
343
return pathjoin(F(p), e)
120
return os.path.join(F(p), e)
346
123
def backup_file(fn):
347
124
"""Copy a file to a backup.
502
259
'sha1': s.hexdigest()}
263
"""Return per-user configuration directory.
265
By default this is ~/.bzr.conf/
267
TODO: Global option --config-dir to override this.
269
return os.path.join(os.path.expanduser("~"), ".bzr.conf")
273
"""Calculate automatic user identification.
275
Returns (realname, email).
277
Only used when none is set in the environment or the id file.
279
This previously used the FQDN as the default domain, but that can
280
be very slow on machines where DNS is broken. So now we simply
285
# XXX: Any good way to get real user name on win32?
290
w = pwd.getpwuid(uid)
291
gecos = w.pw_gecos.decode(bzrlib.user_encoding)
292
username = w.pw_name.decode(bzrlib.user_encoding)
293
comma = gecos.find(',')
297
realname = gecos[:comma]
303
realname = username = getpass.getuser().decode(bzrlib.user_encoding)
305
return realname, (username + '@' + socket.gethostname())
308
def _get_user_id(branch):
309
"""Return the full user id from a file or environment variable.
311
e.g. "John Hacker <jhacker@foo.org>"
314
A branch to use for a per-branch configuration, or None.
316
The following are searched in order:
319
2. .bzr/email for this branch.
323
v = os.environ.get('BZREMAIL')
325
return v.decode(bzrlib.user_encoding)
329
return (branch.controlfile("email", "r")
331
.decode(bzrlib.user_encoding)
334
if e.errno != errno.ENOENT:
340
return (open(os.path.join(config_dir(), "email"))
342
.decode(bzrlib.user_encoding)
345
if e.errno != errno.ENOENT:
348
v = os.environ.get('EMAIL')
350
return v.decode(bzrlib.user_encoding)
355
def username(branch):
356
"""Return email-style username.
358
Something similar to 'Martin Pool <mbp@sourcefrog.net>'
360
TODO: Check it's reasonably well-formed.
362
v = _get_user_id(branch)
366
name, email = _auto_user_id()
368
return '%s <%s>' % (name, email)
373
def user_email(branch):
374
"""Return just the email component of a username."""
375
e = _get_user_id(branch)
377
m = re.search(r'[\w+.-]+@[\w+.-]+', e)
379
raise BzrError("%r doesn't seem to contain "
380
"a reasonable email address" % e)
383
return _auto_user_id()[1]
505
386
def compare_files(a, b):
506
387
"""Returns true if equal in contents"""
677
545
copyfile(src, dest)
679
def delete_any(full_path):
680
"""Delete a file or directory."""
684
# We may be renaming a dangling inventory id
685
if e.errno not in (errno.EISDIR, errno.EACCES, errno.EPERM):
690
548
def has_symlinks():
691
549
if hasattr(os, 'symlink'):
697
def contains_whitespace(s):
698
"""True if there are any whitespace characters in s."""
699
for ch in string.whitespace:
706
def contains_linebreaks(s):
707
"""True if there is any vertical whitespace in s."""
715
def relpath(base, path):
716
"""Return path relative to base, or raise exception.
718
The path may be either an absolute path or a path relative to the
719
current working directory.
721
os.path.commonprefix (python2.4) has a bad bug that it works just
722
on string prefixes, assuming that '/u' is a prefix of '/u2'. This
726
assert len(base) >= MIN_ABS_PATHLENGTH, ('Length of base must be equal or'
727
' exceed the platform minimum length (which is %d)' %
734
while len(head) >= len(base):
737
head, tail = os.path.split(head)
741
raise PathNotChild(rp, base)
749
def safe_unicode(unicode_or_utf8_string):
750
"""Coerce unicode_or_utf8_string into unicode.
752
If it is unicode, it is returned.
753
Otherwise it is decoded from utf-8. If a decoding error
754
occurs, it is wrapped as a If the decoding fails, the exception is wrapped
755
as a BzrBadParameter exception.
757
if isinstance(unicode_or_utf8_string, unicode):
758
return unicode_or_utf8_string
760
return unicode_or_utf8_string.decode('utf8')
761
except UnicodeDecodeError:
762
raise BzrBadParameterNotUnicode(unicode_or_utf8_string)
765
_platform_normalizes_filenames = False
766
if sys.platform == 'darwin':
767
_platform_normalizes_filenames = True
770
def normalizes_filenames():
771
"""Return True if this platform normalizes unicode filenames.
773
Mac OSX does, Windows/Linux do not.
775
return _platform_normalizes_filenames
778
def _accessible_normalized_filename(path):
779
"""Get the unicode normalized path, and if you can access the file.
781
On platforms where the system normalizes filenames (Mac OSX),
782
you can access a file by any path which will normalize correctly.
783
On platforms where the system does not normalize filenames
784
(Windows, Linux), you have to access a file by its exact path.
786
Internally, bzr only supports NFC/NFKC normalization, since that is
787
the standard for XML documents.
789
So return the normalized path, and a flag indicating if the file
790
can be accessed by that path.
793
return unicodedata.normalize('NFKC', path), True
796
def _inaccessible_normalized_filename(path):
797
__doc__ = _accessible_normalized_filename.__doc__
799
normalized = unicodedata.normalize('NFKC', path)
800
return normalized, normalized == path
803
if _platform_normalizes_filenames:
804
normalized_filename = _accessible_normalized_filename
806
normalized_filename = _inaccessible_normalized_filename
809
def terminal_width():
810
"""Return estimated terminal width."""
811
if sys.platform == 'win32':
812
import bzrlib.win32console
813
return bzrlib.win32console.get_console_size()[0]
816
import struct, fcntl, termios
817
s = struct.pack('HHHH', 0, 0, 0, 0)
818
x = fcntl.ioctl(1, termios.TIOCGWINSZ, s)
819
width = struct.unpack('HHHH', x)[1]
824
width = int(os.environ['COLUMNS'])
832
def supports_executable():
833
return sys.platform != "win32"
836
_validWin32PathRE = re.compile(r'^([A-Za-z]:[/\\])?[^:<>*"?\|]*$')
839
def check_legal_path(path):
840
"""Check whether the supplied path is legal.
841
This is only required on Windows, so we don't test on other platforms
844
if sys.platform != "win32":
846
if _validWin32PathRE.match(path) is None:
847
raise IllegalPath(path)
850
def walkdirs(top, prefix=""):
851
"""Yield data about all the directories in a tree.
853
This yields all the data about the contents of a directory at a time.
854
After each directory has been yielded, if the caller has mutated the list
855
to exclude some directories, they are then not descended into.
857
The data yielded is of the form:
858
[(relpath, basename, kind, lstat, path_from_top), ...]
860
:param prefix: Prefix the relpaths that are yielded with 'prefix'. This
861
allows one to walk a subtree but get paths that are relative to a tree
863
:return: an iterator over the dirs.
867
_directory = _directory_kind
869
pending = [(prefix, "", _directory, None, top)]
872
currentdir = pending.pop()
873
# 0 - relpath, 1- basename, 2- kind, 3- stat, 4-toppath
876
relroot = currentdir[0] + '/'
879
for name in sorted(_listdir(top)):
880
abspath = top + '/' + name
881
statvalue = lstat(abspath)
882
dirblock.append ((relroot + name, name, file_kind_from_stat_mode(statvalue.st_mode), statvalue, abspath))
884
# push the user specified dirs from dirblock
885
for dir in reversed(dirblock):
886
if dir[2] == _directory:
890
def path_prefix_key(path):
891
"""Generate a prefix-order path key for path.
893
This can be used to sort paths in the same way that walkdirs does.
895
return (dirname(path) , path)
898
def compare_paths_prefix_order(path_a, path_b):
899
"""Compare path_a and path_b to generate the same order walkdirs uses."""
900
key_a = path_prefix_key(path_a)
901
key_b = path_prefix_key(path_b)
902
return cmp(key_a, key_b)
added
removed
bzrlib/osutils.py
