~bzr-pqm/bzr/bzr.dev

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
# Copyright (C) 2005 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


import sys, os, os.path, random, time, sha, sets, types, re, shutil, tempfile
import traceback, socket, fnmatch, difflib, time
from binascii import hexlify

import bzrlib
from inventory import Inventory
from trace import mutter, note
from tree import Tree, EmptyTree, RevisionTree
from inventory import InventoryEntry, Inventory
from osutils import isdir, quotefn, isfile, uuid, sha_file, username, \
     format_date, compact_date, pumpfile, user_email, rand_bytes, splitpath, \
     joinpath, sha_file, sha_string, file_kind, local_time_offset, appendpath
from store import ImmutableStore
from revision import Revision
from errors import BzrError
from textui import show_status

BZR_BRANCH_FORMAT = "Bazaar-NG branch, format 0.0.4\n"
## TODO: Maybe include checks for common corruption of newlines, etc?



def find_branch(f, **args):
    if f and (f.startswith('http://') or f.startswith('https://')):
        import remotebranch 
        return remotebranch.RemoteBranch(f, **args)
    else:
        return Branch(f, **args)



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.

    Lifted out of Branch.relpath for ease of testing.

    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."""
    rp = os.path.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:
        from errors import NotBranchError
        raise NotBranchError("path %r is not within branch %r" % (rp, base))

    return os.sep.join(s)
        

def find_branch_root(f=None):
    """Find the branch root enclosing f, or pwd.

    f may be a filename or a URL.

    It is not necessary that f exists.

    Basically we keep looking up until we find the control directory or
    run into the root."""
    if f == None:
        f = os.getcwd()
    elif hasattr(os.path, 'realpath'):
        f = os.path.realpath(f)
    else:
        f = os.path.abspath(f)
    if not os.path.exists(f):
        raise BzrError('%r does not exist' % f)
        

    orig_f = f

    while True:
        if os.path.exists(os.path.join(f, bzrlib.BZRDIR)):
            return f
        head, tail = os.path.split(f)
        if head == f:
            # reached the root, whatever that may be
            raise BzrError('%r is not in a branch' % orig_f)
        f = head
    
class DivergedBranches(Exception):
    def __init__(self, branch1, branch2):
        self.branch1 = branch1
        self.branch2 = branch2
        Exception.__init__(self, "These branches have diverged.")

######################################################################
# branch objects

class Branch(object):
    """Branch holding a history of revisions.

    base
        Base directory of the branch.

    _lock_mode
        None, or 'r' or 'w'

    _lock_count
        If _lock_mode is true, a positive count of the number of times the
        lock has been taken.

    _lock
        Lock object from bzrlib.lock.
    """
    base = None
    _lock_mode = None
    _lock_count = None
    _lock = None
    
    def __init__(self, base, init=False, find_root=True):
        """Create new branch object at a particular location.

        base -- Base directory for the branch.
        
        init -- If True, create new control files in a previously
             unversioned directory.  If False, the branch must already
             be versioned.

        find_root -- If true and init is false, find the root of the
             existing branch containing base.

        In the test suite, creation of new trees is tested using the
        `ScratchBranch` class.
        """
        if init:
            self.base = os.path.realpath(base)
            self._make_control()
        elif find_root:
            self.base = find_branch_root(base)
        else:
            self.base = os.path.realpath(base)
            if not isdir(self.controlfilename('.')):
                from errors import NotBranchError
                raise NotBranchError("not a bzr branch: %s" % quotefn(base),
                                     ['use "bzr init" to initialize a new working tree',
                                      'current bzr can only operate from top-of-tree'])
        self._check_format()

        self.text_store = ImmutableStore(self.controlfilename('text-store'))
        self.revision_store = ImmutableStore(self.controlfilename('revision-store'))
        self.inventory_store = ImmutableStore(self.controlfilename('inventory-store'))


    def __str__(self):
        return '%s(%r)' % (self.__class__.__name__, self.base)


    __repr__ = __str__


    def __del__(self):
        if self._lock_mode or self._lock:
            from warnings import warn
            warn("branch %r was not explicitly unlocked" % self)
            self._lock.unlock()



    def lock_write(self):
        if self._lock_mode:
            if self._lock_mode != 'w':
                from errors import LockError
                raise LockError("can't upgrade to a write lock from %r" %
                                self._lock_mode)
            self._lock_count += 1
        else:
            from bzrlib.lock import WriteLock

            self._lock = WriteLock(self.controlfilename('branch-lock'))
            self._lock_mode = 'w'
            self._lock_count = 1



    def lock_read(self):
        if self._lock_mode:
            assert self._lock_mode in ('r', 'w'), \
                   "invalid lock mode %r" % self._lock_mode
            self._lock_count += 1
        else:
            from bzrlib.lock import ReadLock

            self._lock = ReadLock(self.controlfilename('branch-lock'))
            self._lock_mode = 'r'
            self._lock_count = 1
                        

            
    def unlock(self):
        if not self._lock_mode:
            from errors import LockError
            raise LockError('branch %r is not locked' % (self))

        if self._lock_count > 1:
            self._lock_count -= 1
        else:
            self._lock.unlock()
            self._lock = None
            self._lock_mode = self._lock_count = None


    def abspath(self, name):
        """Return absolute filename for something in the branch"""
        return os.path.join(self.base, name)


    def relpath(self, path):
        """Return path relative to this branch of something inside it.

        Raises an error if path is not in this branch."""
        return _relpath(self.base, path)


    def controlfilename(self, file_or_path):
        """Return location relative to branch."""
        if isinstance(file_or_path, types.StringTypes):
            file_or_path = [file_or_path]
        return os.path.join(self.base, bzrlib.BZRDIR, *file_or_path)


    def controlfile(self, file_or_path, mode='r'):
        """Open a control file for this branch.

        There are two classes of file in the control directory: text
        and binary.  binary files are untranslated byte streams.  Text
        control files are stored with Unix newlines and in UTF-8, even
        if the platform or locale defaults are different.

        Controlfiles should almost never be opened in write mode but
        rather should be atomically copied and replaced using atomicfile.
        """

        fn = self.controlfilename(file_or_path)

        if mode == 'rb' or mode == 'wb':
            return file(fn, mode)
        elif mode == 'r' or mode == 'w':
            # open in binary mode anyhow so there's no newline translation;
            # codecs uses line buffering by default; don't want that.
            import codecs
            return codecs.open(fn, mode + 'b', 'utf-8',
                               buffering=60000)
        else:
            raise BzrError("invalid controlfile mode %r" % mode)



    def _make_control(self):
        os.mkdir(self.controlfilename([]))
        self.controlfile('README', 'w').write(
            "This is a Bazaar-NG control directory.\n"
            "Do not change any files in this directory.")
        self.controlfile('branch-format', 'w').write(BZR_BRANCH_FORMAT)
        for d in ('text-store', 'inventory-store', 'revision-store'):
            os.mkdir(self.controlfilename(d))
        for f in ('revision-history', 'merged-patches',
                  'pending-merged-patches', 'branch-name',
                  'branch-lock'):
            self.controlfile(f, 'w').write('')
        mutter('created control directory in ' + self.base)
        Inventory().write_xml(self.controlfile('inventory','w'))


    def _check_format(self):
        """Check this branch format is supported.

        The current tool only supports the current unstable format.

        In the future, we might need different in-memory Branch
        classes to support downlevel branches.  But not yet.
        """
        # This ignores newlines so that we can open branches created
        # on Windows from Linux and so on.  I think it might be better
        # to always make all internal files in unix format.
        fmt = self.controlfile('branch-format', 'r').read()
        fmt.replace('\r\n', '')
        if fmt != BZR_BRANCH_FORMAT:
            raise BzrError('sorry, branch format %r not supported' % fmt,
                           ['use a different bzr version',
                            'or remove the .bzr directory and "bzr init" again'])



    def read_working_inventory(self):
        """Read the working inventory."""
        before = time.time()
        # ElementTree does its own conversion from UTF-8, so open in
        # binary.
        self.lock_read()
        try:
            inv = Inventory.read_xml(self.controlfile('inventory', 'rb'))
            mutter("loaded inventory of %d items in %f"
                   % (len(inv), time.time() - before))
            return inv
        finally:
            self.unlock()
            

    def _write_inventory(self, inv):
        """Update the working inventory.

        That is to say, the inventory describing changes underway, that
        will be committed to the next revision.
        """
        ## TODO: factor out to atomicfile?  is rename safe on windows?
        ## TODO: Maybe some kind of clean/dirty marker on inventory?
        tmpfname = self.controlfilename('inventory.tmp')
        tmpf = file(tmpfname, 'wb')
        inv.write_xml(tmpf)
        tmpf.close()
        inv_fname = self.controlfilename('inventory')
        if sys.platform == 'win32':
            os.remove(inv_fname)
        os.rename(tmpfname, inv_fname)
        mutter('wrote working inventory')
            

    inventory = property(read_working_inventory, _write_inventory, None,
                         """Inventory for the working copy.""")


    def add(self, files, verbose=False, ids=None):
        """Make files versioned.

        Note that the command line normally calls smart_add instead.

        This puts the files in the Added state, so that they will be
        recorded by the next commit.

        files
            List of paths to add, relative to the base of the tree.

        ids
            If set, use these instead of automatically generated ids.
            Must be the same length as the list of files, but may
            contain None for ids that are to be autogenerated.

        TODO: Perhaps have an option to add the ids even if the files do
              not (yet) exist.

        TODO: Perhaps return the ids of the files?  But then again it
              is easy to retrieve them if they're needed.

        TODO: Adding a directory should optionally recurse down and
              add all non-ignored children.  Perhaps do that in a
              higher-level method.
        """
        # TODO: Re-adding a file that is removed in the working copy
        # should probably put it back with the previous ID.
        if isinstance(files, types.StringTypes):
            assert(ids is None or isinstance(ids, types.StringTypes))
            files = [files]
            if ids is not None:
                ids = [ids]

        if ids is None:
            ids = [None] * len(files)
        else:
            assert(len(ids) == len(files))

        self.lock_write()
        try:
            inv = self.read_working_inventory()
            for f,file_id in zip(files, ids):
                if is_control_file(f):
                    raise BzrError("cannot add control file %s" % quotefn(f))

                fp = splitpath(f)

                if len(fp) == 0:
                    raise BzrError("cannot add top-level %r" % f)

                fullpath = os.path.normpath(self.abspath(f))

                try:
                    kind = file_kind(fullpath)
                except OSError:
                    # maybe something better?
                    raise BzrError('cannot add: not a regular file or directory: %s' % quotefn(f))

                if kind != 'file' and kind != 'directory':
                    raise BzrError('cannot add: not a regular file or directory: %s' % quotefn(f))

                if file_id is None:
                    file_id = gen_file_id(f)
                inv.add_path(f, kind=kind, file_id=file_id)

                if verbose:
                    show_status('A', kind, quotefn(f))

                mutter("add file %s file_id:{%s} kind=%r" % (f, file_id, kind))

            self._write_inventory(inv)
        finally:
            self.unlock()
            

    def print_file(self, file, revno):
        """Print `file` to stdout."""
        self.lock_read()
        try:
            tree = self.revision_tree(self.lookup_revision(revno))
            # use inventory as it was in that revision
            file_id = tree.inventory.path2id(file)
            if not file_id:
                raise BzrError("%r is not present in revision %d" % (file, revno))
            tree.print_file(file_id)
        finally:
            self.unlock()


    def remove(self, files, verbose=False):
        """Mark nominated files for removal from the inventory.

        This does not remove their text.  This does not run on 

        TODO: Refuse to remove modified files unless --force is given?

        TODO: Do something useful with directories.

        TODO: Should this remove the text or not?  Tough call; not
        removing may be useful and the user can just use use rm, and
        is the opposite of add.  Removing it is consistent with most
        other tools.  Maybe an option.
        """
        ## TODO: Normalize names
        ## TODO: Remove nested loops; better scalability
        if isinstance(files, types.StringTypes):
            files = [files]

        self.lock_write()

        try:
            tree = self.working_tree()
            inv = tree.inventory

            # do this before any modifications
            for f in files:
                fid = inv.path2id(f)
                if not fid:
                    raise BzrError("cannot remove unversioned file %s" % quotefn(f))
                mutter("remove inventory entry %s {%s}" % (quotefn(f), fid))
                if verbose:
                    # having remove it, it must be either ignored or unknown
                    if tree.is_ignored(f):
                        new_status = 'I'
                    else:
                        new_status = '?'
                    show_status(new_status, inv[fid].kind, quotefn(f))
                del inv[fid]

            self._write_inventory(inv)
        finally:
            self.unlock()


    # FIXME: this doesn't need to be a branch method
    def set_inventory(self, new_inventory_list):
        inv = Inventory()
        for path, file_id, parent, kind in new_inventory_list:
            name = os.path.basename(path)
            if name == "":
                continue
            inv.add(InventoryEntry(file_id, name, kind, parent))
        self._write_inventory(inv)


    def unknowns(self):
        """Return all unknown files.

        These are files in the working directory that are not versioned or
        control files or ignored.
        
        >>> b = ScratchBranch(files=['foo', 'foo~'])
        >>> list(b.unknowns())
        ['foo']
        >>> b.add('foo')
        >>> list(b.unknowns())
        []
        >>> b.remove('foo')
        >>> list(b.unknowns())
        ['foo']
        """
        return self.working_tree().unknowns()


    def append_revision(self, revision_id):
        mutter("add {%s} to revision-history" % revision_id)
        rev_history = self.revision_history()

        tmprhname = self.controlfilename('revision-history.tmp')
        rhname = self.controlfilename('revision-history')
        
        f = file(tmprhname, 'wt')
        rev_history.append(revision_id)
        f.write('\n'.join(rev_history))
        f.write('\n')
        f.close()

        if sys.platform == 'win32':
            os.remove(rhname)
        os.rename(tmprhname, rhname)
        


    def get_revision(self, revision_id):
        """Return the Revision object for a named revision"""
        if not revision_id or not isinstance(revision_id, basestring):
            raise ValueError('invalid revision-id: %r' % revision_id)
        r = Revision.read_xml(self.revision_store[revision_id])
        assert r.revision_id == revision_id
        return r

    def get_revision_sha1(self, revision_id):
        """Hash the stored value of a revision, and return it."""
        # In the future, revision entries will be signed. At that
        # point, it is probably best *not* to include the signature
        # in the revision hash. Because that lets you re-sign
        # the revision, (add signatures/remove signatures) and still
        # have all hash pointers stay consistent.
        # But for now, just hash the contents.
        return sha_file(self.revision_store[revision_id])


    def get_inventory(self, inventory_id):
        """Get Inventory object by hash.

        TODO: Perhaps for this and similar methods, take a revision
               parameter which can be either an integer revno or a
               string hash."""
        i = Inventory.read_xml(self.inventory_store[inventory_id])
        return i

    def get_inventory_sha1(self, inventory_id):
        """Return the sha1 hash of the inventory entry
        """
        return sha_file(self.inventory_store[inventory_id])


    def get_revision_inventory(self, revision_id):
        """Return inventory of a past revision."""
        if revision_id == None:
            return Inventory()
        else:
            return self.get_inventory(self.get_revision(revision_id).inventory_id)


    def revision_history(self):
        """Return sequence of revision hashes on to this branch.

        >>> ScratchBranch().revision_history()
        []
        """
        self.lock_read()
        try:
            return [l.rstrip('\r\n') for l in
                    self.controlfile('revision-history', 'r').readlines()]
        finally:
            self.unlock()


    def common_ancestor(self, other, self_revno=None, other_revno=None):
        """
        >>> import commit
        >>> sb = ScratchBranch(files=['foo', 'foo~'])
        >>> sb.common_ancestor(sb) == (None, None)
        True
        >>> commit.commit(sb, "Committing first revision", verbose=False)
        >>> sb.common_ancestor(sb)[0]
        1
        >>> clone = sb.clone()
        >>> commit.commit(sb, "Committing second revision", verbose=False)
        >>> sb.common_ancestor(sb)[0]
        2
        >>> sb.common_ancestor(clone)[0]
        1
        >>> commit.commit(clone, "Committing divergent second revision", 
        ...               verbose=False)
        >>> sb.common_ancestor(clone)[0]
        1
        >>> sb.common_ancestor(clone) == clone.common_ancestor(sb)
        True
        >>> sb.common_ancestor(sb) != clone.common_ancestor(clone)
        True
        >>> clone2 = sb.clone()
        >>> sb.common_ancestor(clone2)[0]
        2
        >>> sb.common_ancestor(clone2, self_revno=1)[0]
        1
        >>> sb.common_ancestor(clone2, other_revno=1)[0]
        1
        """
        my_history = self.revision_history()
        other_history = other.revision_history()
        if self_revno is None:
            self_revno = len(my_history)
        if other_revno is None:
            other_revno = len(other_history)
        indices = range(min((self_revno, other_revno)))
        indices.reverse()
        for r in indices:
            if my_history[r] == other_history[r]:
                return r+1, my_history[r]
        return None, None

    def enum_history(self, direction):
        """Return (revno, revision_id) for history of branch.

        direction
            'forward' is from earliest to latest
            'reverse' is from latest to earliest
        """
        rh = self.revision_history()
        if direction == 'forward':
            i = 1
            for rid in rh:
                yield i, rid
                i += 1
        elif direction == 'reverse':
            i = len(rh)
            while i > 0:
                yield i, rh[i-1]
                i -= 1
        else:
            raise ValueError('invalid history direction', direction)


    def revno(self):
        """Return current revision number for this branch.

        That is equivalent to the number of revisions committed to
        this branch.
        """
        return len(self.revision_history())


    def last_patch(self):
        """Return last patch hash, or None if no history.
        """
        ph = self.revision_history()
        if ph:
            return ph[-1]
        else:
            return None


    def missing_revisions(self, other):
        """
        If self and other have not diverged, return a list of the revisions
        present in other, but missing from self.

        >>> from bzrlib.commit import commit
        >>> bzrlib.trace.silent = True
        >>> br1 = ScratchBranch()
        >>> br2 = ScratchBranch()
        >>> br1.missing_revisions(br2)
        []
        >>> commit(br2, "lala!", rev_id="REVISION-ID-1")
        >>> br1.missing_revisions(br2)
        [u'REVISION-ID-1']
        >>> br2.missing_revisions(br1)
        []
        >>> commit(br1, "lala!", rev_id="REVISION-ID-1")
        >>> br1.missing_revisions(br2)
        []
        >>> commit(br2, "lala!", rev_id="REVISION-ID-2A")
        >>> br1.missing_revisions(br2)
        [u'REVISION-ID-2A']
        >>> commit(br1, "lala!", rev_id="REVISION-ID-2B")
        >>> br1.missing_revisions(br2)
        Traceback (most recent call last):
        DivergedBranches: These branches have diverged.
        """
        self_history = self.revision_history()
        self_len = len(self_history)
        other_history = other.revision_history()
        other_len = len(other_history)
        common_index = min(self_len, other_len) -1
        if common_index >= 0 and \
            self_history[common_index] != other_history[common_index]:
            raise DivergedBranches(self, other)
        if self_len < other_len:
            return other_history[self_len:]
        return []


    def update_revisions(self, other):
        """Pull in all new revisions from other branch.
        
        >>> from bzrlib.commit import commit
        >>> bzrlib.trace.silent = True
        >>> br1 = ScratchBranch(files=['foo', 'bar'])
        >>> br1.add('foo')
        >>> br1.add('bar')
        >>> commit(br1, "lala!", rev_id="REVISION-ID-1", verbose=False)
        >>> br2 = ScratchBranch()
        >>> br2.update_revisions(br1)
        Added 2 texts.
        Added 1 inventories.
        Added 1 revisions.
        >>> br2.revision_history()
        [u'REVISION-ID-1']
        >>> br2.update_revisions(br1)
        Added 0 texts.
        Added 0 inventories.
        Added 0 revisions.
        >>> br1.text_store.total_size() == br2.text_store.total_size()
        True
        """
        from bzrlib.progress import ProgressBar

        pb = ProgressBar()

        pb.update('comparing histories')
        revision_ids = self.missing_revisions(other)
        revisions = []
        needed_texts = sets.Set()
        i = 0
        for rev_id in revision_ids:
            i += 1
            pb.update('fetching revision', i, len(revision_ids))
            rev = other.get_revision(rev_id)
            revisions.append(rev)
            inv = other.get_inventory(str(rev.inventory_id))
            for key, entry in inv.iter_entries():
                if entry.text_id is None:
                    continue
                if entry.text_id not in self.text_store:
                    needed_texts.add(entry.text_id)

        pb.clear()
                    
        count = self.text_store.copy_multi(other.text_store, needed_texts)
        print "Added %d texts." % count 
        inventory_ids = [ f.inventory_id for f in revisions ]
        count = self.inventory_store.copy_multi(other.inventory_store, 
                                                inventory_ids)
        print "Added %d inventories." % count 
        revision_ids = [ f.revision_id for f in revisions]
        count = self.revision_store.copy_multi(other.revision_store, 
                                               revision_ids)
        for revision_id in revision_ids:
            self.append_revision(revision_id)
        print "Added %d revisions." % count
                    
        
    def commit(self, *args, **kw):
        """Deprecated"""
        from bzrlib.commit import commit
        commit(self, *args, **kw)
        

    def lookup_revision(self, revno):
        """Return revision hash for revision number."""
        if revno == 0:
            return None

        try:
            # list is 0-based; revisions are 1-based
            return self.revision_history()[revno-1]
        except IndexError:
            raise BzrError("no such revision %s" % revno)


    def revision_tree(self, revision_id):
        """Return Tree for a revision on this branch.

        `revision_id` may be None for the null revision, in which case
        an `EmptyTree` is returned."""
        # TODO: refactor this to use an existing revision object
        # so we don't need to read it in twice.
        if revision_id == None:
            return EmptyTree()
        else:
            inv = self.get_revision_inventory(revision_id)
            return RevisionTree(self.text_store, inv)


    def working_tree(self):
        """Return a `Tree` for the working copy."""
        from workingtree import WorkingTree
        return WorkingTree(self.base, self.read_working_inventory())


    def basis_tree(self):
        """Return `Tree` object for last revision.

        If there are no revisions yet, return an `EmptyTree`.
        """
        r = self.last_patch()
        if r == None:
            return EmptyTree()
        else:
            return RevisionTree(self.text_store, self.get_revision_inventory(r))



    def rename_one(self, from_rel, to_rel):
        """Rename one file.

        This can change the directory or the filename or both.
        """
        self.lock_write()
        try:
            tree = self.working_tree()
            inv = tree.inventory
            if not tree.has_filename(from_rel):
                raise BzrError("can't rename: old working file %r does not exist" % from_rel)
            if tree.has_filename(to_rel):
                raise BzrError("can't rename: new working file %r already exists" % to_rel)

            file_id = inv.path2id(from_rel)
            if file_id == None:
                raise BzrError("can't rename: old name %r is not versioned" % from_rel)

            if inv.path2id(to_rel):
                raise BzrError("can't rename: new name %r is already versioned" % to_rel)

            to_dir, to_tail = os.path.split(to_rel)
            to_dir_id = inv.path2id(to_dir)
            if to_dir_id == None and to_dir != '':
                raise BzrError("can't determine destination directory id for %r" % to_dir)

            mutter("rename_one:")
            mutter("  file_id    {%s}" % file_id)
            mutter("  from_rel   %r" % from_rel)
            mutter("  to_rel     %r" % to_rel)
            mutter("  to_dir     %r" % to_dir)
            mutter("  to_dir_id  {%s}" % to_dir_id)

            inv.rename(file_id, to_dir_id, to_tail)

            print "%s => %s" % (from_rel, to_rel)

            from_abs = self.abspath(from_rel)
            to_abs = self.abspath(to_rel)
            try:
                os.rename(from_abs, to_abs)
            except OSError, e:
                raise BzrError("failed to rename %r to %r: %s"
                        % (from_abs, to_abs, e[1]),
                        ["rename rolled back"])

            self._write_inventory(inv)
        finally:
            self.unlock()


    def move(self, from_paths, to_name):
        """Rename files.

        to_name must exist as a versioned directory.

        If to_name exists and is a directory, the files are moved into
        it, keeping their old names.  If it is a directory, 

        Note that to_name is only the last component of the new name;
        this doesn't change the directory.
        """
        self.lock_write()
        try:
            ## TODO: Option to move IDs only
            assert not isinstance(from_paths, basestring)
            tree = self.working_tree()
            inv = tree.inventory
            to_abs = self.abspath(to_name)
            if not isdir(to_abs):
                raise BzrError("destination %r is not a directory" % to_abs)
            if not tree.has_filename(to_name):
                raise BzrError("destination %r not in working directory" % to_abs)
            to_dir_id = inv.path2id(to_name)
            if to_dir_id == None and to_name != '':
                raise BzrError("destination %r is not a versioned directory" % to_name)
            to_dir_ie = inv[to_dir_id]
            if to_dir_ie.kind not in ('directory', 'root_directory'):
                raise BzrError("destination %r is not a directory" % to_abs)

            to_idpath = inv.get_idpath(to_dir_id)

            for f in from_paths:
                if not tree.has_filename(f):
                    raise BzrError("%r does not exist in working tree" % f)
                f_id = inv.path2id(f)
                if f_id == None:
                    raise BzrError("%r is not versioned" % f)
                name_tail = splitpath(f)[-1]
                dest_path = appendpath(to_name, name_tail)
                if tree.has_filename(dest_path):
                    raise BzrError("destination %r already exists" % dest_path)
                if f_id in to_idpath:
                    raise BzrError("can't move %r to a subdirectory of itself" % f)

            # OK, so there's a race here, it's possible that someone will
            # create a file in this interval and then the rename might be
            # left half-done.  But we should have caught most problems.

            for f in from_paths:
                name_tail = splitpath(f)[-1]
                dest_path = appendpath(to_name, name_tail)
                print "%s => %s" % (f, dest_path)
                inv.rename(inv.path2id(f), to_dir_id, name_tail)
                try:
                    os.rename(self.abspath(f), self.abspath(dest_path))
                except OSError, e:
                    raise BzrError("failed to rename %r to %r: %s" % (f, dest_path, e[1]),
                            ["rename rolled back"])

            self._write_inventory(inv)
        finally:
            self.unlock()



class ScratchBranch(Branch):
    """Special test class: a branch that cleans up after itself.

    >>> b = ScratchBranch()
    >>> isdir(b.base)
    True
    >>> bd = b.base
    >>> b.destroy()
    >>> isdir(bd)
    False
    """
    def __init__(self, files=[], dirs=[], base=None):
        """Make a test branch.

        This creates a temporary directory and runs init-tree in it.

        If any files are listed, they are created in the working copy.
        """
        init = False
        if base is None:
            base = tempfile.mkdtemp()
            init = True
        Branch.__init__(self, base, init=init)
        for d in dirs:
            os.mkdir(self.abspath(d))
            
        for f in files:
            file(os.path.join(self.base, f), 'w').write('content of %s' % f)


    def clone(self):
        """
        >>> orig = ScratchBranch(files=["file1", "file2"])
        >>> clone = orig.clone()
        >>> os.path.samefile(orig.base, clone.base)
        False
        >>> os.path.isfile(os.path.join(clone.base, "file1"))
        True
        """
        base = tempfile.mkdtemp()
        os.rmdir(base)
        shutil.copytree(self.base, base, symlinks=True)
        return ScratchBranch(base=base)
        
    def __del__(self):
        self.destroy()

    def destroy(self):
        """Destroy the test branch, removing the scratch directory."""
        try:
            if self.base:
                mutter("delete ScratchBranch %s" % self.base)
                shutil.rmtree(self.base)
        except OSError, e:
            # Work around for shutil.rmtree failing on Windows when
            # readonly files are encountered
            mutter("hit exception in destroying ScratchBranch: %s" % e)
            for root, dirs, files in os.walk(self.base, topdown=False):
                for name in files:
                    os.chmod(os.path.join(root, name), 0700)
            shutil.rmtree(self.base)
        self.base = None

    

######################################################################
# predicates


def is_control_file(filename):
    ## FIXME: better check
    filename = os.path.normpath(filename)
    while filename != '':
        head, tail = os.path.split(filename)
        ## mutter('check %r for control file' % ((head, tail), ))
        if tail == bzrlib.BZRDIR:
            return True
        if filename == head:
            break
        filename = head
    return False



def gen_file_id(name):
    """Return new file id.

    This should probably generate proper UUIDs, but for the moment we
    cope with just randomness because running uuidgen every time is
    slow."""
    import re

    # get last component
    idx = name.rfind('/')
    if idx != -1:
        name = name[idx+1 : ]
    idx = name.rfind('\\')
    if idx != -1:
        name = name[idx+1 : ]

    # make it not a hidden file
    name = name.lstrip('.')

    # remove any wierd characters; we don't escape them but rather
    # just pull them out
    name = re.sub(r'[^\w.]', '', name)

    s = hexlify(rand_bytes(8))
    return '-'.join((name, compact_date(time.time()), s))