14
14
# along with this program; if not, write to the Free Software
15
15
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
20
from stat import (S_ISREG, S_ISDIR, S_ISLNK, ST_MODE, ST_SIZE,
21
S_ISCHR, S_ISBLK, S_ISFIFO, S_ISSOCK)
25
26
from bzrlib.lazy_import import lazy_import
26
27
lazy_import(globals(), """
27
29
from datetime import datetime
31
from ntpath import (abspath as _nt_abspath,
33
normpath as _nt_normpath,
34
realpath as _nt_realpath,
35
splitdrive as _nt_splitdrive,
31
# We need to import both shutil and rmtree as we export the later on posix
32
# and need the former on windows
34
from shutil import rmtree
37
# We need to import both tempfile and mkdtemp as we export the later on posix
38
# and need the former on windows
40
from tempfile import mkdtemp
45
from tempfile import (
43
50
from bzrlib import (
51
from bzrlib.symbol_versioning import (
56
57
# sha and md5 modules are deprecated in python2.6 but hashlib is available as
58
59
if sys.version_info < (2, 5):
84
85
# be opened in binary mode, rather than text mode.
85
86
# On other platforms, O_BINARY doesn't exist, because
86
87
# they always open in binary mode, so it is okay to
87
# OR with 0 on those platforms.
88
# O_NOINHERIT and O_TEXT exists only on win32 too.
88
# OR with 0 on those platforms
89
89
O_BINARY = getattr(os, 'O_BINARY', 0)
90
O_TEXT = getattr(os, 'O_TEXT', 0)
91
O_NOINHERIT = getattr(os, 'O_NOINHERIT', 0)
94
92
def get_unicode_argv():
206
202
:param old: The old path, to rename from
207
203
:param new: The new path, to rename to
208
204
:param rename_func: The potentially non-atomic rename function
209
:param unlink_func: A way to delete the target file if the full rename
205
:param unlink_func: A way to delete the target file if the full rename succeeds
212
208
# sftp rename doesn't allow overwriting, so play tricks:
213
209
base = os.path.basename(new)
214
210
dirname = os.path.dirname(new)
215
# callers use different encodings for the paths so the following MUST
216
# respect that. We rely on python upcasting to unicode if new is unicode
217
# and keeping a str if not.
218
tmp_name = 'tmp.%s.%.9f.%d.%s' % (base, time.time(),
219
os.getpid(), rand_chars(10))
211
tmp_name = u'tmp.%s.%.9f.%d.%s' % (base, time.time(), os.getpid(), rand_chars(10))
220
212
tmp_name = pathjoin(dirname, tmp_name)
222
214
# Rename the file out of the way, but keep track if it didn't exist
298
290
running python.exe under cmd.exe return capital C:\\
299
291
running win32 python inside a cygwin shell returns lowercase c:\\
301
drive, path = ntpath.splitdrive(path)
293
drive, path = _nt_splitdrive(path)
302
294
return drive.upper() + path
305
297
def _win32_abspath(path):
306
# Real ntpath.abspath doesn't have a problem with a unicode cwd
307
return _win32_fixdrive(ntpath.abspath(unicode(path)).replace('\\', '/'))
298
# Real _nt_abspath doesn't have a problem with a unicode cwd
299
return _win32_fixdrive(_nt_abspath(unicode(path)).replace('\\', '/'))
310
302
def _win98_abspath(path):
321
313
# /path => C:/path
322
314
path = unicode(path)
323
315
# check for absolute path
324
drive = ntpath.splitdrive(path)[0]
316
drive = _nt_splitdrive(path)[0]
325
317
if drive == '' and path[:2] not in('//','\\\\'):
326
318
cwd = os.getcwdu()
327
319
# we cannot simply os.path.join cwd and path
328
320
# because os.path.join('C:','/path') produce '/path'
329
321
# and this is incorrect
330
322
if path[:1] in ('/','\\'):
331
cwd = ntpath.splitdrive(cwd)[0]
323
cwd = _nt_splitdrive(cwd)[0]
333
325
path = cwd + '\\' + path
334
return _win32_fixdrive(ntpath.normpath(path).replace('\\', '/'))
326
return _win32_fixdrive(_nt_normpath(path).replace('\\', '/'))
337
329
def _win32_realpath(path):
338
# Real ntpath.realpath doesn't have a problem with a unicode cwd
339
return _win32_fixdrive(ntpath.realpath(unicode(path)).replace('\\', '/'))
330
# Real _nt_realpath doesn't have a problem with a unicode cwd
331
return _win32_fixdrive(_nt_realpath(unicode(path)).replace('\\', '/'))
342
334
def _win32_pathjoin(*args):
343
return ntpath.join(*args).replace('\\', '/')
335
return _nt_join(*args).replace('\\', '/')
346
338
def _win32_normpath(path):
347
return _win32_fixdrive(ntpath.normpath(unicode(path)).replace('\\', '/'))
339
return _win32_fixdrive(_nt_normpath(unicode(path)).replace('\\', '/'))
350
342
def _win32_getcwd():
456
446
input_encoding = getattr(sys.stdin, 'encoding', None)
457
447
if not input_encoding:
458
448
output_encoding = get_user_encoding()
460
mutter('encoding stdout as osutils.get_user_encoding() %r',
449
mutter('encoding stdout as osutils.get_user_encoding() %r',
463
452
output_encoding = input_encoding
465
mutter('encoding stdout as sys.stdin encoding %r',
453
mutter('encoding stdout as sys.stdin encoding %r', output_encoding)
469
mutter('encoding stdout as sys.stdout encoding %r', output_encoding)
455
mutter('encoding stdout as sys.stdout encoding %r', output_encoding)
470
456
if output_encoding == 'cp0':
471
457
# invalid encoding (cp0 means 'no codepage' on Windows)
472
458
output_encoding = get_user_encoding()
474
mutter('cp0 is invalid encoding.'
459
mutter('cp0 is invalid encoding.'
475
460
' encoding stdout as osutils.get_user_encoding() %r',
1234
1216
# but for now, we haven't optimized...
1235
1217
return [canonical_relpath(base, p) for p in paths]
1238
def decode_filename(filename):
1239
"""Decode the filename using the filesystem encoding
1241
If it is unicode, it is returned.
1242
Otherwise it is decoded from the the filesystem's encoding. If decoding
1243
fails, a errors.BadFilenameEncoding exception is raised.
1245
if type(filename) is unicode:
1248
return filename.decode(_fs_enc)
1249
except UnicodeDecodeError:
1250
raise errors.BadFilenameEncoding(filename, _fs_enc)
1253
1219
def safe_unicode(unicode_or_utf8_string):
1254
1220
"""Coerce unicode_or_utf8_string into unicode.
1349
1315
On platforms where the system normalizes filenames (Mac OSX),
1350
1316
you can access a file by any path which will normalize correctly.
1351
1317
On platforms where the system does not normalize filenames
1352
(everything else), you have to access a file by its exact path.
1318
(Windows, Linux), you have to access a file by its exact path.
1354
1320
Internally, bzr only supports NFC normalization, since that is
1355
1321
the standard for XML documents.
1374
1340
normalized_filename = _inaccessible_normalized_filename
1377
def set_signal_handler(signum, handler, restart_syscall=True):
1378
"""A wrapper for signal.signal that also calls siginterrupt(signum, False)
1379
on platforms that support that.
1381
:param restart_syscall: if set, allow syscalls interrupted by a signal to
1382
automatically restart (by calling `signal.siginterrupt(signum,
1383
False)`). May be ignored if the feature is not available on this
1384
platform or Python version.
1388
siginterrupt = signal.siginterrupt
1390
# This python implementation doesn't provide signal support, hence no
1393
except AttributeError:
1394
# siginterrupt doesn't exist on this platform, or for this version
1396
siginterrupt = lambda signum, flag: None
1398
def sig_handler(*args):
1399
# Python resets the siginterrupt flag when a signal is
1400
# received. <http://bugs.python.org/issue8354>
1401
# As a workaround for some cases, set it back the way we want it.
1402
siginterrupt(signum, False)
1403
# Now run the handler function passed to set_signal_handler.
1406
sig_handler = handler
1407
old_handler = signal.signal(signum, sig_handler)
1409
siginterrupt(signum, False)
1413
1343
default_terminal_width = 80
1414
1344
"""The default terminal width for ttys.
1433
1357
- if BZR_COLUMNS is set, returns its value
1434
1358
- if there is no controlling terminal, returns None
1435
- query the OS, if the queried size has changed since the last query,
1437
1359
- if COLUMNS is set, returns its value,
1438
- if the OS has a value (even though it's never changed), return its value.
1440
1361
From there, we need to query the OS to get the size of the controlling
1443
On Unices we query the OS by:
1444
1365
- get termios.TIOCGWINSZ
1445
1366
- if an error occurs or a negative value is obtained, returns None
1447
On Windows we query the OS by:
1448
1370
- win32utils.get_console_size() decides,
1449
1371
- returns None on error (provided default value)
1451
# Note to implementors: if changing the rules for determining the width,
1452
# make sure you've considered the behaviour in these cases:
1453
# - M-x shell in emacs, where $COLUMNS is set and TIOCGWINSZ returns 0,0.
1454
# - bzr log | less, in bash, where $COLUMNS not set and TIOCGWINSZ returns
1456
# - (add more interesting cases here, if you find any)
1457
# Some programs implement "Use $COLUMNS (if set) until SIGWINCH occurs",
1458
# but we don't want to register a signal handler because it is impossible
1459
# to do so without risking EINTR errors in Python <= 2.6.5 (see
1460
# <http://bugs.python.org/issue8354>). Instead we check TIOCGWINSZ every
1461
# time so we can notice if the reported size has changed, which should have
1464
1374
# If BZR_COLUMNS is set, take it, user is always right
1470
1380
isatty = getattr(sys.stdout, 'isatty', None)
1471
if isatty is None or not isatty():
1381
if isatty is None or not isatty():
1472
1382
# Don't guess, setting BZR_COLUMNS is the recommended way to override.
1476
width, height = os_size = _terminal_size(None, None)
1477
global _first_terminal_size, _terminal_size_state
1478
if _terminal_size_state == 'no_data':
1479
_first_terminal_size = os_size
1480
_terminal_size_state = 'unchanged'
1481
elif (_terminal_size_state == 'unchanged' and
1482
_first_terminal_size != os_size):
1483
_terminal_size_state = 'changed'
1485
# If the OS claims to know how wide the terminal is, and this value has
1486
# ever changed, use that.
1487
if _terminal_size_state == 'changed':
1488
if width is not None and width > 0:
1491
# If COLUMNS is set, use it.
1385
# If COLUMNS is set, take it, the terminal knows better (even inside a
1386
# given terminal, the application can decide to set COLUMNS to a lower
1387
# value (splitted screen) or a bigger value (scroll bars))
1493
1389
return int(os.environ['COLUMNS'])
1494
1390
except (KeyError, ValueError):
1497
# Finally, use an unchanged size from the OS, if we have one.
1498
if _terminal_size_state == 'unchanged':
1499
if width is not None and width > 0:
1393
width, height = _terminal_size(None, None)
1395
# Consider invalid values as meaning no width
1502
# The width could not be determined.
1506
1401
def _win32_terminal_size(width, height):
1533
1428
_terminal_size = _ioctl_terminal_size
1431
def _terminal_size_changed(signum, frame):
1432
"""Set COLUMNS upon receiving a SIGnal for WINdow size CHange."""
1433
width, height = _terminal_size(None, None)
1434
if width is not None:
1435
os.environ['COLUMNS'] = str(width)
1437
if sys.platform == 'win32':
1438
# Martin (gz) mentioned WINDOW_BUFFER_SIZE_RECORD from ReadConsoleInput but
1439
# I've no idea how to plug that in the current design -- vila 20091216
1442
signal.signal(signal.SIGWINCH, _terminal_size_changed)
1536
1445
def supports_executable():
1537
1446
return sys.platform != "win32"
1856
1765
real_handlers[kind](abspath, relpath)
1859
def copy_ownership_from_path(dst, src=None):
1860
"""Copy usr/grp ownership from src file/dir to dst file/dir.
1862
If src is None, the containing directory is used as source. If chown
1863
fails, the error is ignored and a warning is printed.
1865
chown = getattr(os, 'chown', None)
1870
src = os.path.dirname(dst)
1876
chown(dst, s.st_uid, s.st_gid)
1878
trace.warning("Unable to copy ownership from '%s' to '%s': IOError: %s." % (src, dst, e))
1881
1768
def path_prefix_key(path):
1882
1769
"""Generate a prefix-order path key for path.
1987
1870
return socket.gethostname().decode(get_user_encoding())
1990
# We must not read/write any more than 64k at a time from/to a socket so we
1991
# don't risk "no buffer space available" errors on some platforms. Windows in
1992
# particular is likely to throw WSAECONNABORTED or WSAENOBUFS if given too much
1994
MAX_SOCKET_CHUNK = 64 * 1024
1996
def read_bytes_from_socket(sock, report_activity=None,
1997
max_read_size=MAX_SOCKET_CHUNK):
1998
"""Read up to max_read_size of bytes from sock and notify of progress.
2000
Translates "Connection reset by peer" into file-like EOF (return an
2001
empty string rather than raise an error), and repeats the recv if
2002
interrupted by a signal.
2006
bytes = sock.recv(max_read_size)
2007
except socket.error, e:
2009
if eno == getattr(errno, "WSAECONNRESET", errno.ECONNRESET):
2010
# The connection was closed by the other side. Callers expect
2011
# an empty string to signal end-of-stream.
2013
elif eno == errno.EINTR:
2014
# Retry the interrupted recv.
2018
if report_activity is not None:
2019
report_activity(len(bytes), 'read')
2023
def recv_all(socket, count):
1873
def recv_all(socket, bytes):
2024
1874
"""Receive an exact number of bytes.
2026
1876
Regular Socket.recv() may return less than the requested number of bytes,
2027
depending on what's in the OS buffer. MSG_WAITALL is not available
1877
dependning on what's in the OS buffer. MSG_WAITALL is not available
2028
1878
on all platforms, but this should work everywhere. This will return
2029
1879
less than the requested amount if the remote end closes.
2031
1881
This isn't optimized and is intended mostly for use in testing.
2034
while len(b) < count:
2035
new = read_bytes_from_socket(socket, None, count - len(b))
1884
while len(b) < bytes:
1885
new = until_no_eintr(socket.recv, bytes - len(b))
2042
def send_all(sock, bytes, report_activity=None):
1892
def send_all(socket, bytes, report_activity=None):
2043
1893
"""Send all bytes on a socket.
2045
Breaks large blocks in smaller chunks to avoid buffering limitations on
2046
some platforms, and catches EINTR which may be thrown if the send is
2047
interrupted by a signal.
2049
This is preferred to socket.sendall(), because it avoids portability bugs
2050
and provides activity reporting.
1895
Regular socket.sendall() can give socket error 10053 on Windows. This
1896
implementation sends no more than 64k at a time, which avoids this problem.
2052
1898
:param report_activity: Call this as bytes are read, see
2053
1899
Transport._report_activity
2056
byte_count = len(bytes)
2057
while sent_total < byte_count:
2059
sent = sock.send(buffer(bytes, sent_total, MAX_SOCKET_CHUNK))
2060
except socket.error, e:
2061
if e.args[0] != errno.EINTR:
2065
report_activity(sent, 'write')
1902
for pos in xrange(0, len(bytes), chunk_size):
1903
block = bytes[pos:pos+chunk_size]
1904
if report_activity is not None:
1905
report_activity(len(block), 'write')
1906
until_no_eintr(socket.sendall, block)
2068
1909
def dereference_path(path):
2109
1950
base = dirname(bzrlib.__file__)
2110
1951
if getattr(sys, 'frozen', None): # bzr.exe
2111
1952
base = abspath(pathjoin(base, '..', '..'))
2112
f = file(pathjoin(base, resource_relpath), "rU")
1953
filename = pathjoin(base, resource_relpath)
1954
return open(filename, 'rU').read()
2118
1957
def file_kind_from_stat_mode_thunk(mode):
2119
1958
global file_kind_from_stat_mode
2143
1982
def until_no_eintr(f, *a, **kw):
2144
"""Run f(*a, **kw), retrying if an EINTR error occurs.
2146
WARNING: you must be certain that it is safe to retry the call repeatedly
2147
if EINTR does occur. This is typically only true for low-level operations
2148
like os.read. If in any doubt, don't use this.
2150
Keep in mind that this is not a complete solution to EINTR. There is
2151
probably code in the Python standard library and other dependencies that
2152
may encounter EINTR if a signal arrives (and there is signal handler for
2153
that signal). So this function can reduce the impact for IO that bzrlib
2154
directly controls, but it is not a complete solution.
1983
"""Run f(*a, **kw), retrying if an EINTR error occurs."""
2156
1984
# Borrowed from Twisted's twisted.python.util.untilConcludes function.
2262
2088
_cached_concurrency = concurrency
2263
2089
return concurrency
2266
class UnicodeOrBytesToBytesWriter(codecs.StreamWriter):
2267
"""A stream writer that doesn't decode str arguments."""
2269
def __init__(self, encode, stream, errors='strict'):
2270
codecs.StreamWriter.__init__(self, stream, errors)
2271
self.encode = encode
2273
def write(self, object):
2274
if type(object) is str:
2275
self.stream.write(object)
2277
data, _ = self.encode(object, self.errors)
2278
self.stream.write(data)
2280
if sys.platform == 'win32':
2281
def open_file(filename, mode='r', bufsize=-1):
2282
"""This function is used to override the ``open`` builtin.
2284
But it uses O_NOINHERIT flag so the file handle is not inherited by
2285
child processes. Deleting or renaming a closed file opened with this
2286
function is not blocking child processes.
2288
writing = 'w' in mode
2289
appending = 'a' in mode
2290
updating = '+' in mode
2291
binary = 'b' in mode
2294
# see http://msdn.microsoft.com/en-us/library/yeby3zcb%28VS.71%29.aspx
2295
# for flags for each modes.
2305
flags |= os.O_WRONLY
2306
flags |= os.O_CREAT | os.O_TRUNC
2311
flags |= os.O_WRONLY
2312
flags |= os.O_CREAT | os.O_APPEND
2317
flags |= os.O_RDONLY
2319
return os.fdopen(os.open(filename, flags), mode, bufsize)
2324
def getuser_unicode():
2325
"""Return the username as unicode.
2328
user_encoding = get_user_encoding()
2329
username = getpass.getuser().decode(user_encoding)
2330
except UnicodeDecodeError:
2331
raise errors.BzrError("Can't decode username as %s." % \
added
removed
bzrlib/osutils.py
