~bzr/bzr/readdir

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
# Copyright (C) 2005, 2006, 2007 Canonical Ltd
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA

from cStringIO import StringIO
import os
import re
import stat
from stat import (S_ISREG, S_ISDIR, S_ISLNK, ST_MODE, ST_SIZE,
                  S_ISCHR, S_ISBLK, S_ISFIFO, S_ISSOCK)
import sys
import time

from bzrlib.lazy_import import lazy_import
lazy_import(globals(), """
import codecs
from datetime import datetime
import errno
from ntpath import (abspath as _nt_abspath,
                    join as _nt_join,
                    normpath as _nt_normpath,
                    realpath as _nt_realpath,
                    splitdrive as _nt_splitdrive,
                    )
import posixpath
import sha
import shutil
from shutil import (
    rmtree,
    )
import tempfile
from tempfile import (
    mkdtemp,
    )
import unicodedata

from bzrlib import (
    cache_utf8,
    errors,
    win32utils,
    )
""")

import bzrlib
from bzrlib.readdir import read_dir
from bzrlib import symbol_versioning
from bzrlib.symbol_versioning import (
    deprecated_function,
    zero_nine,
    )
from bzrlib.trace import mutter


# On win32, O_BINARY is used to indicate the file should
# be opened in binary mode, rather than text mode.
# On other platforms, O_BINARY doesn't exist, because
# they always open in binary mode, so it is okay to
# OR with 0 on those platforms
O_BINARY = getattr(os, 'O_BINARY', 0)


def make_readonly(filename):
    """Make a filename read-only."""
    mod = os.stat(filename).st_mode
    mod = mod & 0777555
    os.chmod(filename, mod)


def make_writable(filename):
    mod = os.stat(filename).st_mode
    mod = mod | 0200
    os.chmod(filename, mod)


_QUOTE_RE = None


def quotefn(f):
    """Return a quoted filename filename

    This previously used backslash quoting, but that works poorly on
    Windows."""
    # TODO: I'm not really sure this is the best format either.x
    global _QUOTE_RE
    if _QUOTE_RE is None:
        _QUOTE_RE = re.compile(r'([^a-zA-Z0-9.,:/\\_~-])')
        
    if _QUOTE_RE.search(f):
        return '"' + f + '"'
    else:
        return f


_directory_kind = 'directory'

_formats = {
    stat.S_IFDIR:_directory_kind,
    stat.S_IFCHR:'chardev',
    stat.S_IFBLK:'block',
    stat.S_IFREG:'file',
    stat.S_IFIFO:'fifo',
    stat.S_IFLNK:'symlink',
    stat.S_IFSOCK:'socket',
}


def file_kind_from_stat_mode(stat_mode, _formats=_formats, _unknown='unknown'):
    """Generate a file kind from a stat mode. This is used in walkdirs.

    Its performance is critical: Do not mutate without careful benchmarking.
    """
    try:
        return _formats[stat_mode & 0170000]
    except KeyError:
        return _unknown


def file_kind(f, _lstat=os.lstat, _mapper=file_kind_from_stat_mode):
    try:
        return _mapper(_lstat(f).st_mode)
    except OSError, e:
        if getattr(e, 'errno', None) == errno.ENOENT:
            raise errors.NoSuchFile(f)
        raise


def get_umask():
    """Return the current umask"""
    # Assume that people aren't messing with the umask while running
    # XXX: This is not thread safe, but there is no way to get the
    #      umask without setting it
    umask = os.umask(0)
    os.umask(umask)
    return umask


def kind_marker(kind):
    if kind == 'file':
        return ''
    elif kind == _directory_kind:
        return '/'
    elif kind == 'symlink':
        return '@'
    else:
        raise errors.BzrError('invalid file kind %r' % kind)

lexists = getattr(os.path, 'lexists', None)
if lexists is None:
    def lexists(f):
        try:
            if getattr(os, 'lstat') is not None:
                os.lstat(f)
            else:
                os.stat(f)
            return True
        except OSError,e:
            if e.errno == errno.ENOENT:
                return False;
            else:
                raise errors.BzrError("lstat/stat of (%r): %r" % (f, e))


def fancy_rename(old, new, rename_func, unlink_func):
    """A fancy rename, when you don't have atomic rename.
    
    :param old: The old path, to rename from
    :param new: The new path, to rename to
    :param rename_func: The potentially non-atomic rename function
    :param unlink_func: A way to delete the target file if the full rename succeeds
    """

    # sftp rename doesn't allow overwriting, so play tricks:
    import random
    base = os.path.basename(new)
    dirname = os.path.dirname(new)
    tmp_name = u'tmp.%s.%.9f.%d.%s' % (base, time.time(), os.getpid(), rand_chars(10))
    tmp_name = pathjoin(dirname, tmp_name)

    # Rename the file out of the way, but keep track if it didn't exist
    # We don't want to grab just any exception
    # something like EACCES should prevent us from continuing
    # The downside is that the rename_func has to throw an exception
    # with an errno = ENOENT, or NoSuchFile
    file_existed = False
    try:
        rename_func(new, tmp_name)
    except (errors.NoSuchFile,), e:
        pass
    except IOError, e:
        # RBC 20060103 abstraction leakage: the paramiko SFTP clients rename
        # function raises an IOError with errno is None when a rename fails.
        # This then gets caught here.
        if e.errno not in (None, errno.ENOENT, errno.ENOTDIR):
            raise
    except Exception, e:
        if (getattr(e, 'errno', None) is None
            or e.errno not in (errno.ENOENT, errno.ENOTDIR)):
            raise
    else:
        file_existed = True

    success = False
    try:
        # This may throw an exception, in which case success will
        # not be set.
        rename_func(old, new)
        success = True
    finally:
        if file_existed:
            # If the file used to exist, rename it back into place
            # otherwise just delete it from the tmp location
            if success:
                unlink_func(tmp_name)
            else:
                rename_func(tmp_name, new)


# In Python 2.4.2 and older, os.path.abspath and os.path.realpath
# choke on a Unicode string containing a relative path if
# os.getcwd() returns a non-sys.getdefaultencoding()-encoded
# string.
_fs_enc = sys.getfilesystemencoding() or 'utf-8'
def _posix_abspath(path):
    # jam 20060426 rather than encoding to fsencoding
    # copy posixpath.abspath, but use os.getcwdu instead
    if not posixpath.isabs(path):
        path = posixpath.join(getcwd(), path)
    return posixpath.normpath(path)


def _posix_realpath(path):
    return posixpath.realpath(path.encode(_fs_enc)).decode(_fs_enc)


def _win32_fixdrive(path):
    """Force drive letters to be consistent.

    win32 is inconsistent whether it returns lower or upper case
    and even if it was consistent the user might type the other
    so we force it to uppercase
    running python.exe under cmd.exe return capital C:\\
    running win32 python inside a cygwin shell returns lowercase c:\\
    """
    drive, path = _nt_splitdrive(path)
    return drive.upper() + path


def _win32_abspath(path):
    # Real _nt_abspath doesn't have a problem with a unicode cwd
    return _win32_fixdrive(_nt_abspath(unicode(path)).replace('\\', '/'))


def _win98_abspath(path):
    """Return the absolute version of a path.
    Windows 98 safe implementation (python reimplementation
    of Win32 API function GetFullPathNameW)
    """
    # Corner cases:
    #   C:\path     => C:/path
    #   C:/path     => C:/path
    #   \\HOST\path => //HOST/path
    #   //HOST/path => //HOST/path
    #   path        => C:/cwd/path
    #   /path       => C:/path
    path = unicode(path)
    # check for absolute path
    drive = _nt_splitdrive(path)[0]
    if drive == '' and path[:2] not in('//','\\\\'):
        cwd = os.getcwdu()
        # we cannot simply os.path.join cwd and path
        # because os.path.join('C:','/path') produce '/path'
        # and this is incorrect
        if path[:1] in ('/','\\'):
            cwd = _nt_splitdrive(cwd)[0]
            path = path[1:]
        path = cwd + '\\' + path
    return _win32_fixdrive(_nt_normpath(path).replace('\\', '/'))

if win32utils.winver == 'Windows 98':
    _win32_abspath = _win98_abspath


def _win32_realpath(path):
    # Real _nt_realpath doesn't have a problem with a unicode cwd
    return _win32_fixdrive(_nt_realpath(unicode(path)).replace('\\', '/'))


def _win32_pathjoin(*args):
    return _nt_join(*args).replace('\\', '/')


def _win32_normpath(path):
    return _win32_fixdrive(_nt_normpath(unicode(path)).replace('\\', '/'))


def _win32_getcwd():
    return _win32_fixdrive(os.getcwdu().replace('\\', '/'))


def _win32_mkdtemp(*args, **kwargs):
    return _win32_fixdrive(tempfile.mkdtemp(*args, **kwargs).replace('\\', '/'))


def _win32_rename(old, new):
    """We expect to be able to atomically replace 'new' with old.

    On win32, if new exists, it must be moved out of the way first,
    and then deleted. 
    """
    try:
        fancy_rename(old, new, rename_func=os.rename, unlink_func=os.unlink)
    except OSError, e:
        if e.errno in (errno.EPERM, errno.EACCES, errno.EBUSY, errno.EINVAL):
            # If we try to rename a non-existant file onto cwd, we get 
            # EPERM or EACCES instead of ENOENT, this will raise ENOENT 
            # if the old path doesn't exist, sometimes we get EACCES
            # On Linux, we seem to get EBUSY, on Mac we get EINVAL
            os.lstat(old)
        raise


def _mac_getcwd():
    return unicodedata.normalize('NFKC', os.getcwdu())


# Default is to just use the python builtins, but these can be rebound on
# particular platforms.
abspath = _posix_abspath
realpath = _posix_realpath
pathjoin = os.path.join
normpath = os.path.normpath
getcwd = os.getcwdu
rename = os.rename
dirname = os.path.dirname
basename = os.path.basename
split = os.path.split
splitext = os.path.splitext
# These were already imported into local scope
# mkdtemp = tempfile.mkdtemp
# rmtree = shutil.rmtree

MIN_ABS_PATHLENGTH = 1


if sys.platform == 'win32':
    abspath = _win32_abspath
    realpath = _win32_realpath
    pathjoin = _win32_pathjoin
    normpath = _win32_normpath
    getcwd = _win32_getcwd
    mkdtemp = _win32_mkdtemp
    rename = _win32_rename

    MIN_ABS_PATHLENGTH = 3

    def _win32_delete_readonly(function, path, excinfo):
        """Error handler for shutil.rmtree function [for win32]
        Helps to remove files and dirs marked as read-only.
        """
        exception = excinfo[1]
        if function in (os.remove, os.rmdir) \
            and isinstance(exception, OSError) \
            and exception.errno == errno.EACCES:
            make_writable(path)
            function(path)
        else:
            raise

    def rmtree(path, ignore_errors=False, onerror=_win32_delete_readonly):
        """Replacer for shutil.rmtree: could remove readonly dirs/files"""
        return shutil.rmtree(path, ignore_errors, onerror)
elif sys.platform == 'darwin':
    getcwd = _mac_getcwd


def get_terminal_encoding():
    """Find the best encoding for printing to the screen.

    This attempts to check both sys.stdout and sys.stdin to see
    what encoding they are in, and if that fails it falls back to
    bzrlib.user_encoding.
    The problem is that on Windows, locale.getpreferredencoding()
    is not the same encoding as that used by the console:
    http://mail.python.org/pipermail/python-list/2003-May/162357.html

    On my standard US Windows XP, the preferred encoding is
    cp1252, but the console is cp437
    """
    output_encoding = getattr(sys.stdout, 'encoding', None)
    if not output_encoding:
        input_encoding = getattr(sys.stdin, 'encoding', None)
        if not input_encoding:
            output_encoding = bzrlib.user_encoding
            mutter('encoding stdout as bzrlib.user_encoding %r', output_encoding)
        else:
            output_encoding = input_encoding
            mutter('encoding stdout as sys.stdin encoding %r', output_encoding)
    else:
        mutter('encoding stdout as sys.stdout encoding %r', output_encoding)
    if output_encoding == 'cp0':
        # invalid encoding (cp0 means 'no codepage' on Windows)
        output_encoding = bzrlib.user_encoding
        mutter('cp0 is invalid encoding.'
               ' encoding stdout as bzrlib.user_encoding %r', output_encoding)
    # check encoding
    try:
        codecs.lookup(output_encoding)
    except LookupError:
        sys.stderr.write('bzr: warning:'
                         ' unknown terminal encoding %s.\n'
                         '  Using encoding %s instead.\n'
                         % (output_encoding, bzrlib.user_encoding)
                        )
        output_encoding = bzrlib.user_encoding

    return output_encoding


def normalizepath(f):
    if getattr(os.path, 'realpath', None) is not None:
        F = realpath
    else:
        F = abspath
    [p,e] = os.path.split(f)
    if e == "" or e == "." or e == "..":
        return F(f)
    else:
        return pathjoin(F(p), e)


def backup_file(fn):
    """Copy a file to a backup.

    Backups are named in GNU-style, with a ~ suffix.

    If the file is already a backup, it's not copied.
    """
    if fn[-1] == '~':
        return
    bfn = fn + '~'

    if has_symlinks() and os.path.islink(fn):
        target = os.readlink(fn)
        os.symlink(target, bfn)
        return
    inf = file(fn, 'rb')
    try:
        content = inf.read()
    finally:
        inf.close()
    
    outf = file(bfn, 'wb')
    try:
        outf.write(content)
    finally:
        outf.close()


def isdir(f):
    """True if f is an accessible directory."""
    try:
        return S_ISDIR(os.lstat(f)[ST_MODE])
    except OSError:
        return False


def isfile(f):
    """True if f is a regular file."""
    try:
        return S_ISREG(os.lstat(f)[ST_MODE])
    except OSError:
        return False

def islink(f):
    """True if f is a symlink."""
    try:
        return S_ISLNK(os.lstat(f)[ST_MODE])
    except OSError:
        return False

def is_inside(dir, fname):
    """True if fname is inside dir.
    
    The parameters should typically be passed to osutils.normpath first, so
    that . and .. and repeated slashes are eliminated, and the separators
    are canonical for the platform.
    
    The empty string as a dir name is taken as top-of-tree and matches 
    everything.
    """
    # XXX: Most callers of this can actually do something smarter by 
    # looking at the inventory
    if dir == fname:
        return True
    
    if dir == '':
        return True

    if dir[-1] != '/':
        dir += '/'

    return fname.startswith(dir)


def is_inside_any(dir_list, fname):
    """True if fname is inside any of given dirs."""
    for dirname in dir_list:
        if is_inside(dirname, fname):
            return True
    else:
        return False


def is_inside_or_parent_of_any(dir_list, fname):
    """True if fname is a child or a parent of any of the given files."""
    for dirname in dir_list:
        if is_inside(dirname, fname) or is_inside(fname, dirname):
            return True
    else:
        return False


def pumpfile(fromfile, tofile):
    """Copy contents of one file to another."""
    BUFSIZE = 32768
    while True:
        b = fromfile.read(BUFSIZE)
        if not b:
            break
        tofile.write(b)


def file_iterator(input_file, readsize=32768):
    while True:
        b = input_file.read(readsize)
        if len(b) == 0:
            break
        yield b


def sha_file(f):
    if getattr(f, 'tell', None) is not None:
        assert f.tell() == 0
    s = sha.new()
    BUFSIZE = 128<<10
    while True:
        b = f.read(BUFSIZE)
        if not b:
            break
        s.update(b)
    return s.hexdigest()



def sha_strings(strings):
    """Return the sha-1 of concatenation of strings"""
    s = sha.new()
    map(s.update, strings)
    return s.hexdigest()


def sha_string(f):
    s = sha.new()
    s.update(f)
    return s.hexdigest()


def fingerprint_file(f):
    s = sha.new()
    b = f.read()
    s.update(b)
    size = len(b)
    return {'size': size,
            'sha1': s.hexdigest()}


def compare_files(a, b):
    """Returns true if equal in contents"""
    BUFSIZE = 4096
    while True:
        ai = a.read(BUFSIZE)
        bi = b.read(BUFSIZE)
        if ai != bi:
            return False
        if ai == '':
            return True


def local_time_offset(t=None):
    """Return offset of local zone from GMT, either at present or at time t."""
    if t is None:
        t = time.time()
    offset = datetime.fromtimestamp(t) - datetime.utcfromtimestamp(t)
    return offset.days * 86400 + offset.seconds

    
def format_date(t, offset=0, timezone='original', date_fmt=None, 
                show_offset=True):
    ## TODO: Perhaps a global option to use either universal or local time?
    ## Or perhaps just let people set $TZ?
    assert isinstance(t, float)
    
    if timezone == 'utc':
        tt = time.gmtime(t)
        offset = 0
    elif timezone == 'original':
        if offset is None:
            offset = 0
        tt = time.gmtime(t + offset)
    elif timezone == 'local':
        tt = time.localtime(t)
        offset = local_time_offset(t)
    else:
        raise errors.BzrError("unsupported timezone format %r" % timezone,
                              ['options are "utc", "original", "local"'])
    if date_fmt is None:
        date_fmt = "%a %Y-%m-%d %H:%M:%S"
    if show_offset:
        offset_str = ' %+03d%02d' % (offset / 3600, (offset / 60) % 60)
    else:
        offset_str = ''
    return (time.strftime(date_fmt, tt) +  offset_str)


def compact_date(when):
    return time.strftime('%Y%m%d%H%M%S', time.gmtime(when))
    

def format_delta(delta):
    """Get a nice looking string for a time delta.

    :param delta: The time difference in seconds, can be positive or negative.
        positive indicates time in the past, negative indicates time in the
        future. (usually time.time() - stored_time)
    :return: String formatted to show approximate resolution
    """
    delta = int(delta)
    if delta >= 0:
        direction = 'ago'
    else:
        direction = 'in the future'
        delta = -delta

    seconds = delta
    if seconds < 90: # print seconds up to 90 seconds
        if seconds == 1:
            return '%d second %s' % (seconds, direction,)
        else:
            return '%d seconds %s' % (seconds, direction)

    minutes = int(seconds / 60)
    seconds -= 60 * minutes
    if seconds == 1:
        plural_seconds = ''
    else:
        plural_seconds = 's'
    if minutes < 90: # print minutes, seconds up to 90 minutes
        if minutes == 1:
            return '%d minute, %d second%s %s' % (
                    minutes, seconds, plural_seconds, direction)
        else:
            return '%d minutes, %d second%s %s' % (
                    minutes, seconds, plural_seconds, direction)

    hours = int(minutes / 60)
    minutes -= 60 * hours
    if minutes == 1:
        plural_minutes = ''
    else:
        plural_minutes = 's'

    if hours == 1:
        return '%d hour, %d minute%s %s' % (hours, minutes,
                                            plural_minutes, direction)
    return '%d hours, %d minute%s %s' % (hours, minutes,
                                         plural_minutes, direction)

def filesize(f):
    """Return size of given open file."""
    return os.fstat(f.fileno())[ST_SIZE]


# Define rand_bytes based on platform.
try:
    # Python 2.4 and later have os.urandom,
    # but it doesn't work on some arches
    os.urandom(1)
    rand_bytes = os.urandom
except (NotImplementedError, AttributeError):
    # If python doesn't have os.urandom, or it doesn't work,
    # then try to first pull random data from /dev/urandom
    try:
        rand_bytes = file('/dev/urandom', 'rb').read
    # Otherwise, use this hack as a last resort
    except (IOError, OSError):
        # not well seeded, but better than nothing
        def rand_bytes(n):
            import random
            s = ''
            while n:
                s += chr(random.randint(0, 255))
                n -= 1
            return s


ALNUM = '0123456789abcdefghijklmnopqrstuvwxyz'
def rand_chars(num):
    """Return a random string of num alphanumeric characters
    
    The result only contains lowercase chars because it may be used on 
    case-insensitive filesystems.
    """
    s = ''
    for raw_byte in rand_bytes(num):
        s += ALNUM[ord(raw_byte) % 36]
    return s


## TODO: We could later have path objects that remember their list
## decomposition (might be too tricksy though.)

def splitpath(p):
    """Turn string into list of parts."""
    assert isinstance(p, basestring)

    # split on either delimiter because people might use either on
    # Windows
    ps = re.split(r'[\\/]', p)

    rps = []
    for f in ps:
        if f == '..':
            raise errors.BzrError("sorry, %r not allowed in path" % f)
        elif (f == '.') or (f == ''):
            pass
        else:
            rps.append(f)
    return rps

def joinpath(p):
    assert isinstance(p, (list, tuple))
    for f in p:
        if (f == '..') or (f is None) or (f == ''):
            raise errors.BzrError("sorry, %r not allowed in path" % f)
    return pathjoin(*p)


@deprecated_function(zero_nine)
def appendpath(p1, p2):
    if p1 == '':
        return p2
    else:
        return pathjoin(p1, p2)
    

def split_lines(s):
    """Split s into lines, but without removing the newline characters."""
    lines = s.split('\n')
    result = [line + '\n' for line in lines[:-1]]
    if lines[-1]:
        result.append(lines[-1])
    return result


def hardlinks_good():
    return sys.platform not in ('win32', 'cygwin', 'darwin')


def link_or_copy(src, dest):
    """Hardlink a file, or copy it if it can't be hardlinked."""
    if not hardlinks_good():
        shutil.copyfile(src, dest)
        return
    try:
        os.link(src, dest)
    except (OSError, IOError), e:
        if e.errno != errno.EXDEV:
            raise
        shutil.copyfile(src, dest)

def delete_any(full_path):
    """Delete a file or directory."""
    try:
        os.unlink(full_path)
    except OSError, e:
    # We may be renaming a dangling inventory id
        if e.errno not in (errno.EISDIR, errno.EACCES, errno.EPERM):
            raise
        os.rmdir(full_path)


def has_symlinks():
    if getattr(os, 'symlink', None) is not None:
        return True
    else:
        return False
        

def contains_whitespace(s):
    """True if there are any whitespace characters in s."""
    # string.whitespace can include '\xa0' in certain locales, because it is
    # considered "non-breaking-space" as part of ISO-8859-1. But it
    # 1) Isn't a breaking whitespace
    # 2) Isn't one of ' \t\r\n' which are characters we sometimes use as
    #    separators
    # 3) '\xa0' isn't unicode safe since it is >128.

    # This should *not* be a unicode set of characters in case the source
    # string is not a Unicode string. We can auto-up-cast the characters since
    # they are ascii, but we don't want to auto-up-cast the string in case it
    # is utf-8
    for ch in ' \t\n\r\v\f':
        if ch in s:
            return True
    else:
        return False


def contains_linebreaks(s):
    """True if there is any vertical whitespace in s."""
    for ch in '\f\n\r':
        if ch in s:
            return True
    else:
        return False


def relpath(base, path):
    """Return path relative to base, or raise exception.

    The path may be either an absolute path or a path relative to the
    current working directory.

    os.path.commonprefix (python2.4) has a bad bug that it works just
    on string prefixes, assuming that '/u' is a prefix of '/u2'.  This
    avoids that problem.
    """

    assert len(base) >= MIN_ABS_PATHLENGTH, ('Length of base must be equal or'
        ' exceed the platform minimum length (which is %d)' % 
        MIN_ABS_PATHLENGTH)

    rp = abspath(path)

    s = []
    head = rp
    while len(head) >= len(base):
        if head == base:
            break
        head, tail = os.path.split(head)
        if tail:
            s.insert(0, tail)
    else:
        raise errors.PathNotChild(rp, base)

    if s:
        return pathjoin(*s)
    else:
        return ''


def safe_unicode(unicode_or_utf8_string):
    """Coerce unicode_or_utf8_string into unicode.

    If it is unicode, it is returned.
    Otherwise it is decoded from utf-8. If a decoding error
    occurs, it is wrapped as a If the decoding fails, the exception is wrapped 
    as a BzrBadParameter exception.
    """
    if isinstance(unicode_or_utf8_string, unicode):
        return unicode_or_utf8_string
    try:
        return unicode_or_utf8_string.decode('utf8')
    except UnicodeDecodeError:
        raise errors.BzrBadParameterNotUnicode(unicode_or_utf8_string)


def safe_utf8(unicode_or_utf8_string):
    """Coerce unicode_or_utf8_string to a utf8 string.

    If it is a str, it is returned.
    If it is Unicode, it is encoded into a utf-8 string.
    """
    if isinstance(unicode_or_utf8_string, str):
        # TODO: jam 20070209 This is overkill, and probably has an impact on
        #       performance if we are dealing with lots of apis that want a
        #       utf-8 revision id
        try:
            # Make sure it is a valid utf-8 string
            unicode_or_utf8_string.decode('utf-8')
        except UnicodeDecodeError:
            raise errors.BzrBadParameterNotUnicode(unicode_or_utf8_string)
        return unicode_or_utf8_string
    return unicode_or_utf8_string.encode('utf-8')


_revision_id_warning = ('Unicode revision ids were deprecated in bzr 0.15.'
                        ' Revision id generators should be creating utf8'
                        ' revision ids.')


def safe_revision_id(unicode_or_utf8_string, warn=True):
    """Revision ids should now be utf8, but at one point they were unicode.

    :param unicode_or_utf8_string: A possibly Unicode revision_id. (can also be
        utf8 or None).
    :param warn: Functions that are sanitizing user data can set warn=False
    :return: None or a utf8 revision id.
    """
    if (unicode_or_utf8_string is None
        or unicode_or_utf8_string.__class__ == str):
        return unicode_or_utf8_string
    if warn:
        symbol_versioning.warn(_revision_id_warning, DeprecationWarning,
                               stacklevel=2)
    return cache_utf8.encode(unicode_or_utf8_string)


_file_id_warning = ('Unicode file ids were deprecated in bzr 0.15. File id'
                    ' generators should be creating utf8 file ids.')


def safe_file_id(unicode_or_utf8_string, warn=True):
    """File ids should now be utf8, but at one point they were unicode.

    This is the same as safe_utf8, except it uses the cached encode functions
    to save a little bit of performance.

    :param unicode_or_utf8_string: A possibly Unicode file_id. (can also be
        utf8 or None).
    :param warn: Functions that are sanitizing user data can set warn=False
    :return: None or a utf8 file id.
    """
    if (unicode_or_utf8_string is None
        or unicode_or_utf8_string.__class__ == str):
        return unicode_or_utf8_string
    if warn:
        symbol_versioning.warn(_file_id_warning, DeprecationWarning,
                               stacklevel=2)
    return cache_utf8.encode(unicode_or_utf8_string)


_platform_normalizes_filenames = False
if sys.platform == 'darwin':
    _platform_normalizes_filenames = True


def normalizes_filenames():
    """Return True if this platform normalizes unicode filenames.

    Mac OSX does, Windows/Linux do not.
    """
    return _platform_normalizes_filenames


def _accessible_normalized_filename(path):
    """Get the unicode normalized path, and if you can access the file.

    On platforms where the system normalizes filenames (Mac OSX),
    you can access a file by any path which will normalize correctly.
    On platforms where the system does not normalize filenames 
    (Windows, Linux), you have to access a file by its exact path.

    Internally, bzr only supports NFC/NFKC normalization, since that is 
    the standard for XML documents.

    So return the normalized path, and a flag indicating if the file
    can be accessed by that path.
    """

    return unicodedata.normalize('NFKC', unicode(path)), True


def _inaccessible_normalized_filename(path):
    __doc__ = _accessible_normalized_filename.__doc__

    normalized = unicodedata.normalize('NFKC', unicode(path))
    return normalized, normalized == path


if _platform_normalizes_filenames:
    normalized_filename = _accessible_normalized_filename
else:
    normalized_filename = _inaccessible_normalized_filename


def terminal_width():
    """Return estimated terminal width."""
    if sys.platform == 'win32':
        return win32utils.get_console_size()[0]
    width = 0
    try:
        import struct, fcntl, termios
        s = struct.pack('HHHH', 0, 0, 0, 0)
        x = fcntl.ioctl(1, termios.TIOCGWINSZ, s)
        width = struct.unpack('HHHH', x)[1]
    except IOError:
        pass
    if width <= 0:
        try:
            width = int(os.environ['COLUMNS'])
        except:
            pass
    if width <= 0:
        width = 80

    return width


def supports_executable():
    return sys.platform != "win32"


def supports_posix_readonly():
    """Return True if 'readonly' has POSIX semantics, False otherwise.

    Notably, a win32 readonly file cannot be deleted, unlike POSIX where the
    directory controls creation/deletion, etc.

    And under win32, readonly means that the directory itself cannot be
    deleted.  The contents of a readonly directory can be changed, unlike POSIX
    where files in readonly directories cannot be added, deleted or renamed.
    """
    return sys.platform != "win32"


def set_or_unset_env(env_variable, value):
    """Modify the environment, setting or removing the env_variable.

    :param env_variable: The environment variable in question
    :param value: The value to set the environment to. If None, then
        the variable will be removed.
    :return: The original value of the environment variable.
    """
    orig_val = os.environ.get(env_variable)
    if value is None:
        if orig_val is not None:
            del os.environ[env_variable]
    else:
        if isinstance(value, unicode):
            value = value.encode(bzrlib.user_encoding)
        os.environ[env_variable] = value
    return orig_val


_validWin32PathRE = re.compile(r'^([A-Za-z]:[/\\])?[^:<>*"?\|]*$')


def check_legal_path(path):
    """Check whether the supplied path is legal.  
    This is only required on Windows, so we don't test on other platforms
    right now.
    """
    if sys.platform != "win32":
        return
    if _validWin32PathRE.match(path) is None:
        raise errors.IllegalPath(path)


def walkdirs(top, prefix=""):
    """Yield data about all the directories in a tree.
    
    This yields all the data about the contents of a directory at a time.
    After each directory has been yielded, if the caller has mutated the list
    to exclude some directories, they are then not descended into.
    
    The data yielded is of the form:
    ((directory-relpath, directory-path-from-top),
    [(directory-relpath, basename, kind, lstat, path-from-top), ...]),
     - directory-relpath is the relative path of the directory being returned
       with respect to top. prefix is prepended to this.
     - directory-path-from-root is the path including top for this directory. 
       It is suitable for use with os functions.
     - relpath is the relative path within the subtree being walked.
     - basename is the basename of the path
     - kind is the kind of the file now. If unknown then the file is not
       present within the tree - but it may be recorded as versioned. See
       versioned_kind.
     - lstat is the stat data *if* the file was statted.
     - planned, not implemented: 
       path_from_tree_root is the path from the root of the tree.

    :param prefix: Prefix the relpaths that are yielded with 'prefix'. This 
        allows one to walk a subtree but get paths that are relative to a tree
        rooted higher up.
    :return: an iterator over the dirs.
    """
    #TODO there is a bit of a smell where the results of the directory-
    # summary in this, and the path from the root, may not agree 
    # depending on top and prefix - i.e. ./foo and foo as a pair leads to
    # potentially confusing output. We should make this more robust - but
    # not at a speed cost. RBC 20060731
    _lstat = os.lstat
    _directory = _directory_kind
    _listdir = os.listdir
    _kind_from_mode = _formats.get
    pending = [(safe_unicode(prefix), "", _directory, None, safe_unicode(top))]
    while pending:
        # 0 - relpath, 1- basename, 2- kind, 3- stat, 4-toppath
        relroot, _, _, _, top = pending.pop()
        if relroot:
            relprefix = relroot + u'/'
        else:
            relprefix = ''
        top_slash = top + u'/'

        dirblock = []
        append = dirblock.append
        for name in sorted(_listdir(top)):
            abspath = top_slash + name
            statvalue = _lstat(abspath)
            kind = _kind_from_mode(statvalue.st_mode & 0170000, 'unknown')
            append((relprefix + name, name, kind, statvalue, abspath))
        yield (relroot, top), dirblock

        # push the user specified dirs from dirblock
        pending.extend(d for d in reversed(dirblock) if d[2] == _directory)


def _walkdirs_utf8(top, prefix=""):
    """Yield data about all the directories in a tree.

    This yields the same information as walkdirs() only each entry is yielded
    in utf-8. On platforms which have a filesystem encoding of utf8 the paths
    are returned as exact byte-strings.

    :return: yields a tuple of (dir_info, [file_info])
        dir_info is (utf8_relpath, path-from-top)
        file_info is (utf8_relpath, utf8_name, kind, lstat, path-from-top)
        if top is an absolute path, path-from-top is also an absolute path.
        path-from-top might be unicode or utf8, but it is the correct path to
        pass to os functions to affect the file in question. (such as os.lstat)
    """
    fs_encoding = sys.getfilesystemencoding()
    if (sys.platform == 'win32' or
        fs_encoding not in ('UTF-8', 'US-ASCII', 'ANSI_X3.4-1968')): # ascii
        return _walkdirs_unicode_to_utf8(top, prefix=prefix)
    else:
        return _walkdirs_fs_utf8(top, prefix=prefix)


def _walkdirs_fs_utf8(top, prefix=""):
    """See _walkdirs_utf8.

    This sub-function is called when we know the filesystem is already in utf8
    encoding. So we don't need to transcode filenames.
    """
    _lstat = os.lstat
    _directory = _directory_kind
    _listdir = os.listdir
    _kind_from_mode = _formats.get

    # 0 - relpath, 1- basename, 2- kind, 3- stat, 4-toppath
    # But we don't actually uses 1-3 in pending, so set them to None
    pending = [(safe_utf8(prefix), None, None, None, safe_utf8(top))]
    while pending:
        relroot, _, _, _, top = pending.pop()
        if relroot:
            relprefix = relroot + '/'
        else:
            relprefix = ''
        top_slash = top + '/'

        dirblock = []
        append = dirblock.append
        for name in sorted(_listdir(top)):
            abspath = top_slash + name
            statvalue = _lstat(abspath)
            kind = _kind_from_mode(statvalue.st_mode & 0170000, 'unknown')
            append((relprefix + name, name, kind, statvalue, abspath))
        yield (relroot, top), dirblock

        # push the user specified dirs from dirblock
        pending.extend(d for d in reversed(dirblock) if d[2] == _directory)


def _walkdirs_unicode_to_utf8(top, prefix=""):
    """See _walkdirs_utf8

    Because Win32 has a Unicode api, all of the 'path-from-top' entries will be
    Unicode paths.
    This is currently the fallback code path when the filesystem encoding is
    not UTF-8. It may be better to implement an alternative so that we can
    safely handle paths that are not properly decodable in the current
    encoding.
    """
    _utf8_encode = codecs.getencoder('utf8')
    _lstat = os.lstat
    _directory = _directory_kind
    _listdir = os.listdir
    _kind_from_mode = _formats.get

    pending = [(safe_utf8(prefix), None, None, None, safe_unicode(top))]
    while pending:
        relroot, _, _, _, top = pending.pop()
        if relroot:
            relprefix = relroot + '/'
        else:
            relprefix = ''
        top_slash = top + u'/'

        dirblock = []
        append = dirblock.append
        for name in sorted(_listdir(top)):
            name_utf8 = _utf8_encode(name)[0]
            abspath = top_slash + name
            statvalue = _lstat(abspath)
            kind = _kind_from_mode(statvalue.st_mode & 0170000, 'unknown')
            append((relprefix + name_utf8, name_utf8, kind, statvalue, abspath))
        yield (relroot, top), dirblock

        # push the user specified dirs from dirblock
        pending.extend(d for d in reversed(dirblock) if d[2] == _directory)


def copy_tree(from_path, to_path, handlers={}):
    """Copy all of the entries in from_path into to_path.

    :param from_path: The base directory to copy. 
    :param to_path: The target directory. If it does not exist, it will
        be created.
    :param handlers: A dictionary of functions, which takes a source and
        destinations for files, directories, etc.
        It is keyed on the file kind, such as 'directory', 'symlink', or 'file'
        'file', 'directory', and 'symlink' should always exist.
        If they are missing, they will be replaced with 'os.mkdir()',
        'os.readlink() + os.symlink()', and 'shutil.copy2()', respectively.
    """
    # Now, just copy the existing cached tree to the new location
    # We use a cheap trick here.
    # Absolute paths are prefixed with the first parameter
    # relative paths are prefixed with the second.
    # So we can get both the source and target returned
    # without any extra work.

    def copy_dir(source, dest):
        os.mkdir(dest)

    def copy_link(source, dest):
        """Copy the contents of a symlink"""
        link_to = os.readlink(source)
        os.symlink(link_to, dest)

    real_handlers = {'file':shutil.copy2,
                     'symlink':copy_link,
                     'directory':copy_dir,
                    }
    real_handlers.update(handlers)

    if not os.path.exists(to_path):
        real_handlers['directory'](from_path, to_path)

    for dir_info, entries in walkdirs(from_path, prefix=to_path):
        for relpath, name, kind, st, abspath in entries:
            real_handlers[kind](abspath, relpath)


def path_prefix_key(path):
    """Generate a prefix-order path key for path.

    This can be used to sort paths in the same way that walkdirs does.
    """
    return (dirname(path) , path)


def compare_paths_prefix_order(path_a, path_b):
    """Compare path_a and path_b to generate the same order walkdirs uses."""
    key_a = path_prefix_key(path_a)
    key_b = path_prefix_key(path_b)
    return cmp(key_a, key_b)


_cached_user_encoding = None


def get_user_encoding(use_cache=True):
    """Find out what the preferred user encoding is.

    This is generally the encoding that is used for command line parameters
    and file contents. This may be different from the terminal encoding
    or the filesystem encoding.

    :param  use_cache:  Enable cache for detected encoding.
                        (This parameter is turned on by default,
                        and required only for selftesting)

    :return: A string defining the preferred user encoding
    """
    global _cached_user_encoding
    if _cached_user_encoding is not None and use_cache:
        return _cached_user_encoding

    if sys.platform == 'darwin':
        # work around egregious python 2.4 bug
        sys.platform = 'posix'
        try:
            import locale
        finally:
            sys.platform = 'darwin'
    else:
        import locale

    try:
        user_encoding = locale.getpreferredencoding()
    except locale.Error, e:
        sys.stderr.write('bzr: warning: %s\n'
                         '  Could not determine what text encoding to use.\n'
                         '  This error usually means your Python interpreter\n'
                         '  doesn\'t support the locale set by $LANG (%s)\n'
                         "  Continuing with ascii encoding.\n"
                         % (e, os.environ.get('LANG')))
        user_encoding = 'ascii'

    # Windows returns 'cp0' to indicate there is no code page. So we'll just
    # treat that as ASCII, and not support printing unicode characters to the
    # console.
    if user_encoding in (None, 'cp0'):
        user_encoding = 'ascii'
    else:
        # check encoding
        try:
            codecs.lookup(user_encoding)
        except LookupError:
            sys.stderr.write('bzr: warning:'
                             ' unknown encoding %s.'
                             ' Continuing with ascii encoding.\n'
                             % user_encoding
                            )
            user_encoding = 'ascii'

    if use_cache:
        _cached_user_encoding = user_encoding

    return user_encoding


def recv_all(socket, bytes):
    """Receive an exact number of bytes.

    Regular Socket.recv() may return less than the requested number of bytes,
    dependning on what's in the OS buffer.  MSG_WAITALL is not available
    on all platforms, but this should work everywhere.  This will return
    less than the requested amount if the remote end closes.

    This isn't optimized and is intended mostly for use in testing.
    """
    b = ''
    while len(b) < bytes:
        new = socket.recv(bytes - len(b))
        if new == '':
            break # eof
        b += new
    return b

def dereference_path(path):
    """Determine the real path to a file.

    All parent elements are dereferenced.  But the file itself is not
    dereferenced.
    :param path: The original path.  May be absolute or relative.
    :return: the real path *to* the file
    """
    parent, base = os.path.split(path)
    # The pathjoin for '.' is a workaround for Python bug #1213894.
    # (initial path components aren't dereferenced)
    return pathjoin(realpath(pathjoin('.', parent)), base)