~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
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
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
2432
2433
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
# Copyright (C) 2005-2010 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., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA

# FIXME: This refactoring of the workingtree code doesn't seem to keep
# the WorkingTree's copy of the inventory in sync with the branch.  The
# branch modifies its working inventory when it does a commit to make
# missing files permanently removed.

# TODO: Maybe also keep the full path of the entry, and the children?
# But those depend on its position within a particular inventory, and
# it would be nice not to need to hold the backpointer here.

# This should really be an id randomly assigned when the tree is
# created, but it's not for now.
ROOT_ID = "TREE_ROOT"

from bzrlib.lazy_import import lazy_import
lazy_import(globals(), """
import collections
import copy
import os
import re
import tarfile

import bzrlib
from bzrlib import (
    chk_map,
    errors,
    generate_ids,
    osutils,
    symbol_versioning,
    )
""")

from bzrlib.errors import (
    BzrCheckError,
    BzrError,
    )
from bzrlib.symbol_versioning import deprecated_in, deprecated_method
from bzrlib.trace import mutter
from bzrlib.static_tuple import StaticTuple


class InventoryEntry(object):
    """Description of a versioned file.

    An InventoryEntry has the following fields, which are also
    present in the XML inventory-entry element:

    file_id

    name
        (within the parent directory)

    parent_id
        file_id of the parent directory, or ROOT_ID

    revision
        the revision_id in which this variation of this file was
        introduced.

    executable
        Indicates that this file should be executable on systems
        that support it.

    text_sha1
        sha-1 of the text of the file

    text_size
        size in bytes of the text of the file

    (reading a version 4 tree created a text_id field.)

    >>> i = Inventory()
    >>> i.path2id('')
    'TREE_ROOT'
    >>> i.add(InventoryDirectory('123', 'src', ROOT_ID))
    InventoryDirectory('123', 'src', parent_id='TREE_ROOT', revision=None)
    >>> i.add(InventoryFile('2323', 'hello.c', parent_id='123'))
    InventoryFile('2323', 'hello.c', parent_id='123', sha1=None, len=None, revision=None)
    >>> shouldbe = {0: '', 1: 'src', 2: 'src/hello.c'}
    >>> for ix, j in enumerate(i.iter_entries()):
    ...   print (j[0] == shouldbe[ix], j[1])
    ...
    (True, InventoryDirectory('TREE_ROOT', u'', parent_id=None, revision=None))
    (True, InventoryDirectory('123', 'src', parent_id='TREE_ROOT', revision=None))
    (True, InventoryFile('2323', 'hello.c', parent_id='123', sha1=None, len=None, revision=None))
    >>> i.add(InventoryFile('2324', 'bye.c', '123'))
    InventoryFile('2324', 'bye.c', parent_id='123', sha1=None, len=None, revision=None)
    >>> i.add(InventoryDirectory('2325', 'wibble', '123'))
    InventoryDirectory('2325', 'wibble', parent_id='123', revision=None)
    >>> i.path2id('src/wibble')
    '2325'
    >>> '2325' in i
    True
    >>> i.add(InventoryFile('2326', 'wibble.c', '2325'))
    InventoryFile('2326', 'wibble.c', parent_id='2325', sha1=None, len=None, revision=None)
    >>> i['2326']
    InventoryFile('2326', 'wibble.c', parent_id='2325', sha1=None, len=None, revision=None)
    >>> for path, entry in i.iter_entries():
    ...     print path
    ...
    <BLANKLINE>
    src
    src/bye.c
    src/hello.c
    src/wibble
    src/wibble/wibble.c
    >>> i.id2path('2326')
    'src/wibble/wibble.c'
    """

    # Constants returned by describe_change()
    #
    # TODO: These should probably move to some kind of FileChangeDescription
    # class; that's like what's inside a TreeDelta but we want to be able to
    # generate them just for one file at a time.
    RENAMED = 'renamed'
    MODIFIED_AND_RENAMED = 'modified and renamed'

    __slots__ = []

    def detect_changes(self, old_entry):
        """Return a (text_modified, meta_modified) from this to old_entry.

        _read_tree_state must have been called on self and old_entry prior to
        calling detect_changes.
        """
        return False, False

    def _diff(self, text_diff, from_label, tree, to_label, to_entry, to_tree,
             output_to, reverse=False):
        """Perform a diff between two entries of the same kind."""

    def parent_candidates(self, previous_inventories):
        """Find possible per-file graph parents.

        This is currently defined by:
         - Select the last changed revision in the parent inventory.
         - Do deal with a short lived bug in bzr 0.8's development two entries
           that have the same last changed but different 'x' bit settings are
           changed in-place.
        """
        # revision:ie mapping for each ie found in previous_inventories.
        candidates = {}
        # identify candidate head revision ids.
        for inv in previous_inventories:
            if self.file_id in inv:
                ie = inv[self.file_id]
                if ie.revision in candidates:
                    # same revision value in two different inventories:
                    # correct possible inconsistencies:
                    #     * there was a bug in revision updates with 'x' bit
                    #       support.
                    try:
                        if candidates[ie.revision].executable != ie.executable:
                            candidates[ie.revision].executable = False
                            ie.executable = False
                    except AttributeError:
                        pass
                else:
                    # add this revision as a candidate.
                    candidates[ie.revision] = ie
        return candidates

    @deprecated_method(deprecated_in((1, 6, 0)))
    def get_tar_item(self, root, dp, now, tree):
        """Get a tarfile item and a file stream for its content."""
        item = tarfile.TarInfo(osutils.pathjoin(root, dp).encode('utf8'))
        # TODO: would be cool to actually set it to the timestamp of the
        # revision it was last changed
        item.mtime = now
        fileobj = self._put_in_tar(item, tree)
        return item, fileobj

    def has_text(self):
        """Return true if the object this entry represents has textual data.

        Note that textual data includes binary content.

        Also note that all entries get weave files created for them.
        This attribute is primarily used when upgrading from old trees that
        did not have the weave index for all inventory entries.
        """
        return False

    def __init__(self, file_id, name, parent_id, text_id=None):
        """Create an InventoryEntry

        The filename must be a single component, relative to the
        parent directory; it cannot be a whole path or relative name.

        >>> e = InventoryFile('123', 'hello.c', ROOT_ID)
        >>> e.name
        'hello.c'
        >>> e.file_id
        '123'
        >>> e = InventoryFile('123', 'src/hello.c', ROOT_ID)
        Traceback (most recent call last):
        InvalidEntryName: Invalid entry name: src/hello.c
        """
        if '/' in name or '\\' in name:
            raise errors.InvalidEntryName(name=name)
        self.executable = False
        self.revision = None
        self.text_sha1 = None
        self.text_size = None
        self.file_id = file_id
        self.name = name
        self.text_id = text_id
        self.parent_id = parent_id
        self.symlink_target = None
        self.reference_revision = None

    def kind_character(self):
        """Return a short kind indicator useful for appending to names."""
        raise BzrError('unknown kind %r' % self.kind)

    known_kinds = ('file', 'directory', 'symlink')

    def _put_in_tar(self, item, tree):
        """populate item for stashing in a tar, and return the content stream.

        If no content is available, return None.
        """
        raise BzrError("don't know how to export {%s} of kind %r" %
                       (self.file_id, self.kind))

    @deprecated_method(deprecated_in((1, 6, 0)))
    def put_on_disk(self, dest, dp, tree):
        """Create a representation of self on disk in the prefix dest.

        This is a template method - implement _put_on_disk in subclasses.
        """
        fullpath = osutils.pathjoin(dest, dp)
        self._put_on_disk(fullpath, tree)
        # mutter("  export {%s} kind %s to %s", self.file_id,
        #         self.kind, fullpath)

    def _put_on_disk(self, fullpath, tree):
        """Put this entry onto disk at fullpath, from tree tree."""
        raise BzrError("don't know how to export {%s} of kind %r" % (self.file_id, self.kind))

    def sorted_children(self):
        return sorted(self.children.items())

    @staticmethod
    def versionable_kind(kind):
        return (kind in ('file', 'directory', 'symlink', 'tree-reference'))

    def check(self, checker, rev_id, inv):
        """Check this inventory entry is intact.

        This is a template method, override _check for kind specific
        tests.

        :param checker: Check object providing context for the checks;
             can be used to find out what parts of the repository have already
             been checked.
        :param rev_id: Revision id from which this InventoryEntry was loaded.
             Not necessarily the last-changed revision for this file.
        :param inv: Inventory from which the entry was loaded.
        """
        if self.parent_id is not None:
            if not inv.has_id(self.parent_id):
                raise BzrCheckError('missing parent {%s} in inventory for revision {%s}'
                        % (self.parent_id, rev_id))
        checker._add_entry_to_text_key_references(inv, self)
        self._check(checker, rev_id)

    def _check(self, checker, rev_id):
        """Check this inventory entry for kind specific errors."""
        checker._report_items.append(
            'unknown entry kind %r in revision {%s}' % (self.kind, rev_id))

    def copy(self):
        """Clone this inventory entry."""
        raise NotImplementedError

    @staticmethod
    def describe_change(old_entry, new_entry):
        """Describe the change between old_entry and this.

        This smells of being an InterInventoryEntry situation, but as its
        the first one, we're making it a static method for now.

        An entry with a different parent, or different name is considered
        to be renamed. Reparenting is an internal detail.
        Note that renaming the parent does not trigger a rename for the
        child entry itself.
        """
        # TODO: Perhaps return an object rather than just a string
        if old_entry is new_entry:
            # also the case of both being None
            return 'unchanged'
        elif old_entry is None:
            return 'added'
        elif new_entry is None:
            return 'removed'
        if old_entry.kind != new_entry.kind:
            return 'modified'
        text_modified, meta_modified = new_entry.detect_changes(old_entry)
        if text_modified or meta_modified:
            modified = True
        else:
            modified = False
        # TODO 20060511 (mbp, rbc) factor out 'detect_rename' here.
        if old_entry.parent_id != new_entry.parent_id:
            renamed = True
        elif old_entry.name != new_entry.name:
            renamed = True
        else:
            renamed = False
        if renamed and not modified:
            return InventoryEntry.RENAMED
        if modified and not renamed:
            return 'modified'
        if modified and renamed:
            return InventoryEntry.MODIFIED_AND_RENAMED
        return 'unchanged'

    def __repr__(self):
        return ("%s(%r, %r, parent_id=%r, revision=%r)"
                % (self.__class__.__name__,
                   self.file_id,
                   self.name,
                   self.parent_id,
                   self.revision))

    def __eq__(self, other):
        if other is self:
            # For the case when objects are cached
            return True
        if not isinstance(other, InventoryEntry):
            return NotImplemented

        return ((self.file_id == other.file_id)
                and (self.name == other.name)
                and (other.symlink_target == self.symlink_target)
                and (self.text_sha1 == other.text_sha1)
                and (self.text_size == other.text_size)
                and (self.text_id == other.text_id)
                and (self.parent_id == other.parent_id)
                and (self.kind == other.kind)
                and (self.revision == other.revision)
                and (self.executable == other.executable)
                and (self.reference_revision == other.reference_revision)
                )

    def __ne__(self, other):
        return not (self == other)

    def __hash__(self):
        raise ValueError('not hashable')

    def _unchanged(self, previous_ie):
        """Has this entry changed relative to previous_ie.

        This method should be overridden in child classes.
        """
        compatible = True
        # different inv parent
        if previous_ie.parent_id != self.parent_id:
            compatible = False
        # renamed
        elif previous_ie.name != self.name:
            compatible = False
        elif previous_ie.kind != self.kind:
            compatible = False
        return compatible

    def _read_tree_state(self, path, work_tree):
        """Populate fields in the inventory entry from the given tree.

        Note that this should be modified to be a noop on virtual trees
        as all entries created there are prepopulated.
        """
        # TODO: Rather than running this manually, we should check the
        # working sha1 and other expensive properties when they're
        # first requested, or preload them if they're already known
        pass            # nothing to do by default

    def _forget_tree_state(self):
        pass


class RootEntry(InventoryEntry):

    __slots__ = ['text_sha1', 'text_size', 'file_id', 'name', 'kind',
                 'text_id', 'parent_id', 'children', 'executable',
                 'revision', 'symlink_target', 'reference_revision']

    def _check(self, checker, rev_id):
        """See InventoryEntry._check"""

    def __init__(self, file_id):
        self.file_id = file_id
        self.children = {}
        self.kind = 'directory'
        self.parent_id = None
        self.name = u''
        self.revision = None
        symbol_versioning.warn('RootEntry is deprecated as of bzr 0.10.'
                               '  Please use InventoryDirectory instead.',
                               DeprecationWarning, stacklevel=2)

    def __eq__(self, other):
        if not isinstance(other, RootEntry):
            return NotImplemented

        return (self.file_id == other.file_id) \
               and (self.children == other.children)


class InventoryDirectory(InventoryEntry):
    """A directory in an inventory."""

    __slots__ = ['text_sha1', 'text_size', 'file_id', 'name', 'kind',
                 'text_id', 'parent_id', 'children', 'executable',
                 'revision', 'symlink_target', 'reference_revision']

    def _check(self, checker, rev_id):
        """See InventoryEntry._check"""
        if (self.text_sha1 is not None or self.text_size is not None or
            self.text_id is not None):
            checker._report_items.append('directory {%s} has text in revision {%s}'
                                % (self.file_id, rev_id))
        # In non rich root repositories we do not expect a file graph for the
        # root.
        if self.name == '' and not checker.rich_roots:
            return
        # Directories are stored as an empty file, but the file should exist
        # to provide a per-fileid log. The hash of every directory content is
        # "da..." below (the sha1sum of '').
        checker.add_pending_item(rev_id,
            ('texts', self.file_id, self.revision), 'text',
             'da39a3ee5e6b4b0d3255bfef95601890afd80709')

    def copy(self):
        other = InventoryDirectory(self.file_id, self.name, self.parent_id)
        other.revision = self.revision
        # note that children are *not* copied; they're pulled across when
        # others are added
        return other

    def __init__(self, file_id, name, parent_id):
        super(InventoryDirectory, self).__init__(file_id, name, parent_id)
        self.children = {}
        self.kind = 'directory'

    def kind_character(self):
        """See InventoryEntry.kind_character."""
        return '/'

    def _put_in_tar(self, item, tree):
        """See InventoryEntry._put_in_tar."""
        item.type = tarfile.DIRTYPE
        fileobj = None
        item.name += '/'
        item.size = 0
        item.mode = 0755
        return fileobj

    def _put_on_disk(self, fullpath, tree):
        """See InventoryEntry._put_on_disk."""
        os.mkdir(fullpath)


class InventoryFile(InventoryEntry):
    """A file in an inventory."""

    __slots__ = ['text_sha1', 'text_size', 'file_id', 'name', 'kind',
                 'text_id', 'parent_id', 'children', 'executable',
                 'revision', 'symlink_target', 'reference_revision']

    def _check(self, checker, tree_revision_id):
        """See InventoryEntry._check"""
        # TODO: check size too.
        checker.add_pending_item(tree_revision_id,
            ('texts', self.file_id, self.revision), 'text',
             self.text_sha1)
        if self.text_size is None:
            checker._report_items.append(
                'fileid {%s} in {%s} has None for text_size' % (self.file_id,
                tree_revision_id))

    def copy(self):
        other = InventoryFile(self.file_id, self.name, self.parent_id)
        other.executable = self.executable
        other.text_id = self.text_id
        other.text_sha1 = self.text_sha1
        other.text_size = self.text_size
        other.revision = self.revision
        return other

    def detect_changes(self, old_entry):
        """See InventoryEntry.detect_changes."""
        text_modified = (self.text_sha1 != old_entry.text_sha1)
        meta_modified = (self.executable != old_entry.executable)
        return text_modified, meta_modified

    def _diff(self, text_diff, from_label, tree, to_label, to_entry, to_tree,
             output_to, reverse=False):
        """See InventoryEntry._diff."""
        from bzrlib.diff import DiffText
        from_file_id = self.file_id
        if to_entry:
            to_file_id = to_entry.file_id
        else:
            to_file_id = None
        if reverse:
            to_file_id, from_file_id = from_file_id, to_file_id
            tree, to_tree = to_tree, tree
            from_label, to_label = to_label, from_label
        differ = DiffText(tree, to_tree, output_to, 'utf-8', '', '',
                          text_diff)
        return differ.diff_text(from_file_id, to_file_id, from_label, to_label)

    def has_text(self):
        """See InventoryEntry.has_text."""
        return True

    def __init__(self, file_id, name, parent_id):
        super(InventoryFile, self).__init__(file_id, name, parent_id)
        self.kind = 'file'

    def kind_character(self):
        """See InventoryEntry.kind_character."""
        return ''

    def _put_in_tar(self, item, tree):
        """See InventoryEntry._put_in_tar."""
        item.type = tarfile.REGTYPE
        fileobj = tree.get_file(self.file_id)
        item.size = self.text_size
        if tree.is_executable(self.file_id):
            item.mode = 0755
        else:
            item.mode = 0644
        return fileobj

    def _put_on_disk(self, fullpath, tree):
        """See InventoryEntry._put_on_disk."""
        osutils.pumpfile(tree.get_file(self.file_id), file(fullpath, 'wb'))
        if tree.is_executable(self.file_id):
            os.chmod(fullpath, 0755)

    def _read_tree_state(self, path, work_tree):
        """See InventoryEntry._read_tree_state."""
        self.text_sha1 = work_tree.get_file_sha1(self.file_id, path=path)
        # FIXME: 20050930 probe for the text size when getting sha1
        # in _read_tree_state
        self.executable = work_tree.is_executable(self.file_id, path=path)

    def __repr__(self):
        return ("%s(%r, %r, parent_id=%r, sha1=%r, len=%s, revision=%s)"
                % (self.__class__.__name__,
                   self.file_id,
                   self.name,
                   self.parent_id,
                   self.text_sha1,
                   self.text_size,
                   self.revision))

    def _forget_tree_state(self):
        self.text_sha1 = None

    def _unchanged(self, previous_ie):
        """See InventoryEntry._unchanged."""
        compatible = super(InventoryFile, self)._unchanged(previous_ie)
        if self.text_sha1 != previous_ie.text_sha1:
            compatible = False
        else:
            # FIXME: 20050930 probe for the text size when getting sha1
            # in _read_tree_state
            self.text_size = previous_ie.text_size
        if self.executable != previous_ie.executable:
            compatible = False
        return compatible


class InventoryLink(InventoryEntry):
    """A file in an inventory."""

    __slots__ = ['text_sha1', 'text_size', 'file_id', 'name', 'kind',
                 'text_id', 'parent_id', 'children', 'executable',
                 'revision', 'symlink_target', 'reference_revision']

    def _check(self, checker, tree_revision_id):
        """See InventoryEntry._check"""
        if self.text_sha1 is not None or self.text_size is not None or self.text_id is not None:
            checker._report_items.append(
               'symlink {%s} has text in revision {%s}'
                    % (self.file_id, tree_revision_id))
        if self.symlink_target is None:
            checker._report_items.append(
                'symlink {%s} has no target in revision {%s}'
                    % (self.file_id, tree_revision_id))
        # Symlinks are stored as ''
        checker.add_pending_item(tree_revision_id,
            ('texts', self.file_id, self.revision), 'text',
             'da39a3ee5e6b4b0d3255bfef95601890afd80709')

    def copy(self):
        other = InventoryLink(self.file_id, self.name, self.parent_id)
        other.symlink_target = self.symlink_target
        other.revision = self.revision
        return other

    def detect_changes(self, old_entry):
        """See InventoryEntry.detect_changes."""
        # FIXME: which _modified field should we use ? RBC 20051003
        text_modified = (self.symlink_target != old_entry.symlink_target)
        if text_modified:
            mutter("    symlink target changed")
        meta_modified = False
        return text_modified, meta_modified

    def _diff(self, text_diff, from_label, tree, to_label, to_entry, to_tree,
             output_to, reverse=False):
        """See InventoryEntry._diff."""
        from bzrlib.diff import DiffSymlink
        old_target = self.symlink_target
        if to_entry is not None:
            new_target = to_entry.symlink_target
        else:
            new_target = None
        if not reverse:
            old_tree = tree
            new_tree = to_tree
        else:
            old_tree = to_tree
            new_tree = tree
            new_target, old_target = old_target, new_target
        differ = DiffSymlink(old_tree, new_tree, output_to)
        return differ.diff_symlink(old_target, new_target)

    def __init__(self, file_id, name, parent_id):
        super(InventoryLink, self).__init__(file_id, name, parent_id)
        self.kind = 'symlink'

    def kind_character(self):
        """See InventoryEntry.kind_character."""
        return ''

    def _put_in_tar(self, item, tree):
        """See InventoryEntry._put_in_tar."""
        item.type = tarfile.SYMTYPE
        fileobj = None
        item.size = 0
        item.mode = 0755
        item.linkname = self.symlink_target
        return fileobj

    def _put_on_disk(self, fullpath, tree):
        """See InventoryEntry._put_on_disk."""
        try:
            os.symlink(self.symlink_target, fullpath)
        except OSError,e:
            raise BzrError("Failed to create symlink %r -> %r, error: %s" % (fullpath, self.symlink_target, e))

    def _read_tree_state(self, path, work_tree):
        """See InventoryEntry._read_tree_state."""
        self.symlink_target = work_tree.get_symlink_target(self.file_id)

    def _forget_tree_state(self):
        self.symlink_target = None

    def _unchanged(self, previous_ie):
        """See InventoryEntry._unchanged."""
        compatible = super(InventoryLink, self)._unchanged(previous_ie)
        if self.symlink_target != previous_ie.symlink_target:
            compatible = False
        return compatible


class TreeReference(InventoryEntry):

    kind = 'tree-reference'

    def __init__(self, file_id, name, parent_id, revision=None,
                 reference_revision=None):
        InventoryEntry.__init__(self, file_id, name, parent_id)
        self.revision = revision
        self.reference_revision = reference_revision

    def copy(self):
        return TreeReference(self.file_id, self.name, self.parent_id,
                             self.revision, self.reference_revision)

    def _read_tree_state(self, path, work_tree):
        """Populate fields in the inventory entry from the given tree.
        """
        self.reference_revision = work_tree.get_reference_revision(
            self.file_id, path)

    def _forget_tree_state(self):
        self.reference_revision = None

    def _unchanged(self, previous_ie):
        """See InventoryEntry._unchanged."""
        compatible = super(TreeReference, self)._unchanged(previous_ie)
        if self.reference_revision != previous_ie.reference_revision:
            compatible = False
        return compatible


class CommonInventory(object):
    """Basic inventory logic, defined in terms of primitives like has_id.

    An inventory is the metadata about the contents of a tree.

    This is broadly a map from file_id to entries such as directories, files,
    symlinks and tree references. Each entry maintains its own metadata like
    SHA1 and length for files, or children for a directory.

    Entries can be looked up either by path or by file_id.

    InventoryEntry objects must not be modified after they are
    inserted, other than through the Inventory API.
    """

    def __contains__(self, file_id):
        """True if this entry contains a file with given id.

        >>> inv = Inventory()
        >>> inv.add(InventoryFile('123', 'foo.c', ROOT_ID))
        InventoryFile('123', 'foo.c', parent_id='TREE_ROOT', sha1=None, len=None, revision=None)
        >>> '123' in inv
        True
        >>> '456' in inv
        False

        Note that this method along with __iter__ are not encouraged for use as
        they are less clear than specific query methods - they may be rmeoved
        in the future.
        """
        return self.has_id(file_id)

    def has_filename(self, filename):
        return bool(self.path2id(filename))

    def id2path(self, file_id):
        """Return as a string the path to file_id.

        >>> i = Inventory()
        >>> e = i.add(InventoryDirectory('src-id', 'src', ROOT_ID))
        >>> e = i.add(InventoryFile('foo-id', 'foo.c', parent_id='src-id'))
        >>> print i.id2path('foo-id')
        src/foo.c

        :raises NoSuchId: If file_id is not present in the inventory.
        """
        # get all names, skipping root
        return '/'.join(reversed(
            [parent.name for parent in
             self._iter_file_id_parents(file_id)][:-1]))

    def iter_entries(self, from_dir=None, recursive=True):
        """Return (path, entry) pairs, in order by name.
        
        :param from_dir: if None, start from the root,
          otherwise start from this directory (either file-id or entry)
        :param recursive: recurse into directories or not
        """
        if from_dir is None:
            if self.root is None:
                return
            from_dir = self.root
            yield '', self.root
        elif isinstance(from_dir, basestring):
            from_dir = self[from_dir]

        # unrolling the recursive called changed the time from
        # 440ms/663ms (inline/total) to 116ms/116ms
        children = from_dir.children.items()
        children.sort()
        if not recursive:
            for name, ie in children:
                yield name, ie
            return
        children = collections.deque(children)
        stack = [(u'', children)]
        while stack:
            from_dir_relpath, children = stack[-1]

            while children:
                name, ie = children.popleft()

                # we know that from_dir_relpath never ends in a slash
                # and 'f' doesn't begin with one, we can do a string op, rather
                # than the checks of pathjoin(), though this means that all paths
                # start with a slash
                path = from_dir_relpath + '/' + name

                yield path[1:], ie

                if ie.kind != 'directory':
                    continue

                # But do this child first
                new_children = ie.children.items()
                new_children.sort()
                new_children = collections.deque(new_children)
                stack.append((path, new_children))
                # Break out of inner loop, so that we start outer loop with child
                break
            else:
                # if we finished all children, pop it off the stack
                stack.pop()

    def iter_entries_by_dir(self, from_dir=None, specific_file_ids=None,
        yield_parents=False):
        """Iterate over the entries in a directory first order.

        This returns all entries for a directory before returning
        the entries for children of a directory. This is not
        lexicographically sorted order, and is a hybrid between
        depth-first and breadth-first.

        :param yield_parents: If True, yield the parents from the root leading
            down to specific_file_ids that have been requested. This has no
            impact if specific_file_ids is None.
        :return: This yields (path, entry) pairs
        """
        if specific_file_ids and not isinstance(specific_file_ids, set):
            specific_file_ids = set(specific_file_ids)
        # TODO? Perhaps this should return the from_dir so that the root is
        # yielded? or maybe an option?
        if from_dir is None:
            if self.root is None:
                return
            # Optimize a common case
            if (not yield_parents and specific_file_ids is not None and
                len(specific_file_ids) == 1):
                file_id = list(specific_file_ids)[0]
                if file_id in self:
                    yield self.id2path(file_id), self[file_id]
                return
            from_dir = self.root
            if (specific_file_ids is None or yield_parents or
                self.root.file_id in specific_file_ids):
                yield u'', self.root
        elif isinstance(from_dir, basestring):
            from_dir = self[from_dir]

        if specific_file_ids is not None:
            # TODO: jam 20070302 This could really be done as a loop rather
            #       than a bunch of recursive calls.
            parents = set()
            byid = self
            def add_ancestors(file_id):
                if file_id not in byid:
                    return
                parent_id = byid[file_id].parent_id
                if parent_id is None:
                    return
                if parent_id not in parents:
                    parents.add(parent_id)
                    add_ancestors(parent_id)
            for file_id in specific_file_ids:
                add_ancestors(file_id)
        else:
            parents = None

        stack = [(u'', from_dir)]
        while stack:
            cur_relpath, cur_dir = stack.pop()

            child_dirs = []
            for child_name, child_ie in sorted(cur_dir.children.iteritems()):

                child_relpath = cur_relpath + child_name

                if (specific_file_ids is None or
                    child_ie.file_id in specific_file_ids or
                    (yield_parents and child_ie.file_id in parents)):
                    yield child_relpath, child_ie

                if child_ie.kind == 'directory':
                    if parents is None or child_ie.file_id in parents:
                        child_dirs.append((child_relpath+'/', child_ie))
            stack.extend(reversed(child_dirs))

    def _make_delta(self, old):
        """Make an inventory delta from two inventories."""
        old_ids = set(old)
        new_ids = set(self)
        adds = new_ids - old_ids
        deletes = old_ids - new_ids
        common = old_ids.intersection(new_ids)
        delta = []
        for file_id in deletes:
            delta.append((old.id2path(file_id), None, file_id, None))
        for file_id in adds:
            delta.append((None, self.id2path(file_id), file_id, self[file_id]))
        for file_id in common:
            if old[file_id] != self[file_id]:
                delta.append((old.id2path(file_id), self.id2path(file_id),
                    file_id, self[file_id]))
        return delta

    def _get_mutable_inventory(self):
        """Returns a mutable copy of the object.

        Some inventories are immutable, yet working trees, for example, needs
        to mutate exisiting inventories instead of creating a new one.
        """
        raise NotImplementedError(self._get_mutable_inventory)

    def make_entry(self, kind, name, parent_id, file_id=None):
        """Simple thunk to bzrlib.inventory.make_entry."""
        return make_entry(kind, name, parent_id, file_id)

    def entries(self):
        """Return list of (path, ie) for all entries except the root.

        This may be faster than iter_entries.
        """
        accum = []
        def descend(dir_ie, dir_path):
            kids = dir_ie.children.items()
            kids.sort()
            for name, ie in kids:
                child_path = osutils.pathjoin(dir_path, name)
                accum.append((child_path, ie))
                if ie.kind == 'directory':
                    descend(ie, child_path)

        descend(self.root, u'')
        return accum

    def directories(self):
        """Return (path, entry) pairs for all directories, including the root.
        """
        accum = []
        def descend(parent_ie, parent_path):
            accum.append((parent_path, parent_ie))

            kids = [(ie.name, ie) for ie in parent_ie.children.itervalues() if ie.kind == 'directory']
            kids.sort()

            for name, child_ie in kids:
                child_path = osutils.pathjoin(parent_path, name)
                descend(child_ie, child_path)
        descend(self.root, u'')
        return accum

    def path2id(self, relpath):
        """Walk down through directories to return entry of last component.

        :param relpath: may be either a list of path components, or a single
            string, in which case it is automatically split.

        This returns the entry of the last component in the path,
        which may be either a file or a directory.

        Returns None IFF the path is not found.
        """
        if isinstance(relpath, basestring):
            names = osutils.splitpath(relpath)
        else:
            names = relpath

        try:
            parent = self.root
        except errors.NoSuchId:
            # root doesn't exist yet so nothing else can
            return None
        if parent is None:
            return None
        for f in names:
            try:
                children = getattr(parent, 'children', None)
                if children is None:
                    return None
                cie = children[f]
                parent = cie
            except KeyError:
                # or raise an error?
                return None

        return parent.file_id

    def filter(self, specific_fileids):
        """Get an inventory view filtered against a set of file-ids.

        Children of directories and parents are included.

        The result may or may not reference the underlying inventory
        so it should be treated as immutable.
        """
        interesting_parents = set()
        for fileid in specific_fileids:
            try:
                interesting_parents.update(self.get_idpath(fileid))
            except errors.NoSuchId:
                # This fileid is not in the inventory - that's ok
                pass
        entries = self.iter_entries()
        if self.root is None:
            return Inventory(root_id=None)
        other = Inventory(entries.next()[1].file_id)
        other.root.revision = self.root.revision
        other.revision_id = self.revision_id
        directories_to_expand = set()
        for path, entry in entries:
            file_id = entry.file_id
            if (file_id in specific_fileids
                or entry.parent_id in directories_to_expand):
                if entry.kind == 'directory':
                    directories_to_expand.add(file_id)
            elif file_id not in interesting_parents:
                continue
            other.add(entry.copy())
        return other

    def get_idpath(self, file_id):
        """Return a list of file_ids for the path to an entry.

        The list contains one element for each directory followed by
        the id of the file itself.  So the length of the returned list
        is equal to the depth of the file in the tree, counting the
        root directory as depth 1.
        """
        p = []
        for parent in self._iter_file_id_parents(file_id):
            p.insert(0, parent.file_id)
        return p


class Inventory(CommonInventory):
    """Mutable dict based in-memory inventory.

    We never store the full path to a file, because renaming a directory
    implicitly moves all of its contents.  This class internally maintains a
    lookup tree that allows the children under a directory to be
    returned quickly.

    >>> inv = Inventory()
    >>> inv.add(InventoryFile('123-123', 'hello.c', ROOT_ID))
    InventoryFile('123-123', 'hello.c', parent_id='TREE_ROOT', sha1=None, len=None, revision=None)
    >>> inv['123-123'].name
    'hello.c'

    Id's may be looked up from paths:

    >>> inv.path2id('hello.c')
    '123-123'
    >>> '123-123' in inv
    True

    There are iterators over the contents:

    >>> [entry[0] for entry in inv.iter_entries()]
    ['', u'hello.c']
    """

    def __init__(self, root_id=ROOT_ID, revision_id=None):
        """Create or read an inventory.

        If a working directory is specified, the inventory is read
        from there.  If the file is specified, read from that. If not,
        the inventory is created empty.

        The inventory is created with a default root directory, with
        an id of None.
        """
        if root_id is not None:
            self._set_root(InventoryDirectory(root_id, u'', None))
        else:
            self.root = None
            self._byid = {}
        self.revision_id = revision_id

    def __repr__(self):
        # More than one page of ouput is not useful anymore to debug
        max_len = 2048
        closing = '...}'
        contents = repr(self._byid)
        if len(contents) > max_len:
            contents = contents[:(max_len-len(closing))] + closing
        return "<Inventory object at %x, contents=%r>" % (id(self), contents)

    def apply_delta(self, delta):
        """Apply a delta to this inventory.

        See the inventory developers documentation for the theory behind
        inventory deltas.

        If delta application fails the inventory is left in an indeterminate
        state and must not be used.

        :param delta: A list of changes to apply. After all the changes are
            applied the final inventory must be internally consistent, but it
            is ok to supply changes which, if only half-applied would have an
            invalid result - such as supplying two changes which rename two
            files, 'A' and 'B' with each other : [('A', 'B', 'A-id', a_entry),
            ('B', 'A', 'B-id', b_entry)].

            Each change is a tuple, of the form (old_path, new_path, file_id,
            new_entry).

            When new_path is None, the change indicates the removal of an entry
            from the inventory and new_entry will be ignored (using None is
            appropriate). If new_path is not None, then new_entry must be an
            InventoryEntry instance, which will be incorporated into the
            inventory (and replace any existing entry with the same file id).

            When old_path is None, the change indicates the addition of
            a new entry to the inventory.

            When neither new_path nor old_path are None, the change is a
            modification to an entry, such as a rename, reparent, kind change
            etc.

            The children attribute of new_entry is ignored. This is because
            this method preserves children automatically across alterations to
            the parent of the children, and cases where the parent id of a
            child is changing require the child to be passed in as a separate
            change regardless. E.g. in the recursive deletion of a directory -
            the directory's children must be included in the delta, or the
            final inventory will be invalid.

            Note that a file_id must only appear once within a given delta.
            An AssertionError is raised otherwise.
        """
        # Check that the delta is legal. It would be nice if this could be
        # done within the loops below but it's safer to validate the delta
        # before starting to mutate the inventory, as there isn't a rollback
        # facility.
        list(_check_delta_unique_ids(_check_delta_unique_new_paths(
            _check_delta_unique_old_paths(_check_delta_ids_match_entry(
            _check_delta_ids_are_valid(
            _check_delta_new_path_entry_both_or_None(
            delta)))))))

        children = {}
        # Remove all affected items which were in the original inventory,
        # starting with the longest paths, thus ensuring parents are examined
        # after their children, which means that everything we examine has no
        # modified children remaining by the time we examine it.
        for old_path, file_id in sorted(((op, f) for op, np, f, e in delta
                                        if op is not None), reverse=True):
            # Preserve unaltered children of file_id for later reinsertion.
            file_id_children = getattr(self[file_id], 'children', {})
            if len(file_id_children):
                children[file_id] = file_id_children
            if self.id2path(file_id) != old_path:
                raise errors.InconsistentDelta(old_path, file_id,
                    "Entry was at wrong other path %r." % self.id2path(file_id))
            # Remove file_id and the unaltered children. If file_id is not
            # being deleted it will be reinserted back later.
            self.remove_recursive_id(file_id)
        # Insert all affected which should be in the new inventory, reattaching
        # their children if they had any. This is done from shortest path to
        # longest, ensuring that items which were modified and whose parents in
        # the resulting inventory were also modified, are inserted after their
        # parents.
        for new_path, f, new_entry in sorted((np, f, e) for op, np, f, e in
                                          delta if np is not None):
            if new_entry.kind == 'directory':
                # Pop the child which to allow detection of children whose
                # parents were deleted and which were not reattached to a new
                # parent.
                replacement = InventoryDirectory(new_entry.file_id,
                    new_entry.name, new_entry.parent_id)
                replacement.revision = new_entry.revision
                replacement.children = children.pop(replacement.file_id, {})
                new_entry = replacement
            try:
                self.add(new_entry)
            except errors.DuplicateFileId:
                raise errors.InconsistentDelta(new_path, new_entry.file_id,
                    "New id is already present in target.")
            except AttributeError:
                raise errors.InconsistentDelta(new_path, new_entry.file_id,
                    "Parent is not a directory.")
            if self.id2path(new_entry.file_id) != new_path:
                raise errors.InconsistentDelta(new_path, new_entry.file_id,
                    "New path is not consistent with parent path.")
        if len(children):
            # Get the parent id that was deleted
            parent_id, children = children.popitem()
            raise errors.InconsistentDelta("<deleted>", parent_id,
                "The file id was deleted but its children were not deleted.")

    def create_by_apply_delta(self, inventory_delta, new_revision_id,
                              propagate_caches=False):
        """See CHKInventory.create_by_apply_delta()"""
        new_inv = self.copy()
        new_inv.apply_delta(inventory_delta)
        new_inv.revision_id = new_revision_id
        return new_inv

    def _set_root(self, ie):
        self.root = ie
        self._byid = {self.root.file_id: self.root}

    def copy(self):
        # TODO: jam 20051218 Should copy also copy the revision_id?
        entries = self.iter_entries()
        if self.root is None:
            return Inventory(root_id=None)
        other = Inventory(entries.next()[1].file_id)
        other.root.revision = self.root.revision
        # copy recursively so we know directories will be added before
        # their children.  There are more efficient ways than this...
        for path, entry in entries:
            other.add(entry.copy())
        return other

    def _get_mutable_inventory(self):
        """See CommonInventory._get_mutable_inventory."""
        return copy.deepcopy(self)

    def __iter__(self):
        """Iterate over all file-ids."""
        return iter(self._byid)

    def iter_just_entries(self):
        """Iterate over all entries.
        
        Unlike iter_entries(), just the entries are returned (not (path, ie))
        and the order of entries is undefined.

        XXX: We may not want to merge this into bzr.dev.
        """
        if self.root is None:
            return
        for _, ie in self._byid.iteritems():
            yield ie

    def __len__(self):
        """Returns number of entries."""
        return len(self._byid)

    def __getitem__(self, file_id):
        """Return the entry for given file_id.

        >>> inv = Inventory()
        >>> inv.add(InventoryFile('123123', 'hello.c', ROOT_ID))
        InventoryFile('123123', 'hello.c', parent_id='TREE_ROOT', sha1=None, len=None, revision=None)
        >>> inv['123123'].name
        'hello.c'
        """
        try:
            return self._byid[file_id]
        except KeyError:
            # really we're passing an inventory, not a tree...
            raise errors.NoSuchId(self, file_id)

    def get_file_kind(self, file_id):
        return self._byid[file_id].kind

    def get_child(self, parent_id, filename):
        return self[parent_id].children.get(filename)

    def _add_child(self, entry):
        """Add an entry to the inventory, without adding it to its parent"""
        if entry.file_id in self._byid:
            raise BzrError("inventory already contains entry with id {%s}" %
                           entry.file_id)
        self._byid[entry.file_id] = entry
        for child in getattr(entry, 'children', {}).itervalues():
            self._add_child(child)
        return entry

    def add(self, entry):
        """Add entry to inventory.

        To add  a file to a branch ready to be committed, use Branch.add,
        which calls this.

        :return: entry
        """
        if entry.file_id in self._byid:
            raise errors.DuplicateFileId(entry.file_id,
                                         self._byid[entry.file_id])
        if entry.parent_id is None:
            self.root = entry
        else:
            try:
                parent = self._byid[entry.parent_id]
            except KeyError:
                raise errors.InconsistentDelta("<unknown>", entry.parent_id,
                    "Parent not in inventory.")
            if entry.name in parent.children:
                raise errors.InconsistentDelta(
                    self.id2path(parent.children[entry.name].file_id),
                    entry.file_id,
                    "Path already versioned")
            parent.children[entry.name] = entry
        return self._add_child(entry)

    def add_path(self, relpath, kind, file_id=None, parent_id=None):
        """Add entry from a path.

        The immediate parent must already be versioned.

        Returns the new entry object."""

        parts = osutils.splitpath(relpath)

        if len(parts) == 0:
            if file_id is None:
                file_id = generate_ids.gen_root_id()
            self.root = InventoryDirectory(file_id, '', None)
            self._byid = {self.root.file_id: self.root}
            return self.root
        else:
            parent_path = parts[:-1]
            parent_id = self.path2id(parent_path)
            if parent_id is None:
                raise errors.NotVersionedError(path=parent_path)
        ie = make_entry(kind, parts[-1], parent_id, file_id)
        return self.add(ie)

    def __delitem__(self, file_id):
        """Remove entry by id.

        >>> inv = Inventory()
        >>> inv.add(InventoryFile('123', 'foo.c', ROOT_ID))
        InventoryFile('123', 'foo.c', parent_id='TREE_ROOT', sha1=None, len=None, revision=None)
        >>> '123' in inv
        True
        >>> del inv['123']
        >>> '123' in inv
        False
        """
        ie = self[file_id]
        del self._byid[file_id]
        if ie.parent_id is not None:
            del self[ie.parent_id].children[ie.name]

    def __eq__(self, other):
        """Compare two sets by comparing their contents.

        >>> i1 = Inventory()
        >>> i2 = Inventory()
        >>> i1 == i2
        True
        >>> i1.add(InventoryFile('123', 'foo', ROOT_ID))
        InventoryFile('123', 'foo', parent_id='TREE_ROOT', sha1=None, len=None, revision=None)
        >>> i1 == i2
        False
        >>> i2.add(InventoryFile('123', 'foo', ROOT_ID))
        InventoryFile('123', 'foo', parent_id='TREE_ROOT', sha1=None, len=None, revision=None)
        >>> i1 == i2
        True
        """
        if not isinstance(other, Inventory):
            return NotImplemented

        return self._byid == other._byid

    def __ne__(self, other):
        return not self.__eq__(other)

    def __hash__(self):
        raise ValueError('not hashable')

    def _iter_file_id_parents(self, file_id):
        """Yield the parents of file_id up to the root."""
        while file_id is not None:
            try:
                ie = self._byid[file_id]
            except KeyError:
                raise errors.NoSuchId(tree=None, file_id=file_id)
            yield ie
            file_id = ie.parent_id

    def has_id(self, file_id):
        return (file_id in self._byid)

    def _make_delta(self, old):
        """Make an inventory delta from two inventories."""
        old_getter = getattr(old, '_byid', old)
        new_getter = self._byid
        old_ids = set(old_getter)
        new_ids = set(new_getter)
        adds = new_ids - old_ids
        deletes = old_ids - new_ids
        if not adds and not deletes:
            common = new_ids
        else:
            common = old_ids.intersection(new_ids)
        delta = []
        for file_id in deletes:
            delta.append((old.id2path(file_id), None, file_id, None))
        for file_id in adds:
            delta.append((None, self.id2path(file_id), file_id, self[file_id]))
        for file_id in common:
            new_ie = new_getter[file_id]
            old_ie = old_getter[file_id]
            # If xml_serializer returns the cached InventoryEntries (rather
            # than always doing .copy()), inlining the 'is' check saves 2.7M
            # calls to __eq__.  Under lsprof this saves 20s => 6s.
            # It is a minor improvement without lsprof.
            if old_ie is new_ie or old_ie == new_ie:
                continue
            else:
                delta.append((old.id2path(file_id), self.id2path(file_id),
                              file_id, new_ie))
        return delta

    def remove_recursive_id(self, file_id):
        """Remove file_id, and children, from the inventory.

        :param file_id: A file_id to remove.
        """
        to_find_delete = [self._byid[file_id]]
        to_delete = []
        while to_find_delete:
            ie = to_find_delete.pop()
            to_delete.append(ie.file_id)
            if ie.kind == 'directory':
                to_find_delete.extend(ie.children.values())
        for file_id in reversed(to_delete):
            ie = self[file_id]
            del self._byid[file_id]
        if ie.parent_id is not None:
            del self[ie.parent_id].children[ie.name]
        else:
            self.root = None

    def rename(self, file_id, new_parent_id, new_name):
        """Move a file within the inventory.

        This can change either the name, or the parent, or both.

        This does not move the working file.
        """
        new_name = ensure_normalized_name(new_name)
        if not is_valid_name(new_name):
            raise BzrError("not an acceptable filename: %r" % new_name)

        new_parent = self._byid[new_parent_id]
        if new_name in new_parent.children:
            raise BzrError("%r already exists in %r" % (new_name, self.id2path(new_parent_id)))

        new_parent_idpath = self.get_idpath(new_parent_id)
        if file_id in new_parent_idpath:
            raise BzrError("cannot move directory %r into a subdirectory of itself, %r"
                    % (self.id2path(file_id), self.id2path(new_parent_id)))

        file_ie = self._byid[file_id]
        old_parent = self._byid[file_ie.parent_id]

        # TODO: Don't leave things messed up if this fails

        del old_parent.children[file_ie.name]
        new_parent.children[new_name] = file_ie

        file_ie.name = new_name
        file_ie.parent_id = new_parent_id

    def is_root(self, file_id):
        return self.root is not None and file_id == self.root.file_id


class CHKInventory(CommonInventory):
    """An inventory persisted in a CHK store.

    By design, a CHKInventory is immutable so many of the methods
    supported by Inventory - add, rename, apply_delta, etc - are *not*
    supported. To create a new CHKInventory, use create_by_apply_delta()
    or from_inventory(), say.

    Internally, a CHKInventory has one or two CHKMaps:

    * id_to_entry - a map from (file_id,) => InventoryEntry as bytes
    * parent_id_basename_to_file_id - a map from (parent_id, basename_utf8)
        => file_id as bytes

    The second map is optional and not present in early CHkRepository's.

    No caching is performed: every method call or item access will perform
    requests to the storage layer. As such, keep references to objects you
    want to reuse.
    """

    def __init__(self, search_key_name):
        CommonInventory.__init__(self)
        self._fileid_to_entry_cache = {}
        self._path_to_fileid_cache = {}
        self._search_key_name = search_key_name
        self.root_id = None

    def __eq__(self, other):
        """Compare two sets by comparing their contents."""
        if not isinstance(other, CHKInventory):
            return NotImplemented

        this_key = self.id_to_entry.key()
        other_key = other.id_to_entry.key()
        this_pid_key = self.parent_id_basename_to_file_id.key()
        other_pid_key = other.parent_id_basename_to_file_id.key()
        if None in (this_key, this_pid_key, other_key, other_pid_key):
            return False
        return this_key == other_key and this_pid_key == other_pid_key

    def _entry_to_bytes(self, entry):
        """Serialise entry as a single bytestring.

        :param Entry: An inventory entry.
        :return: A bytestring for the entry.

        The BNF:
        ENTRY ::= FILE | DIR | SYMLINK | TREE
        FILE ::= "file: " COMMON SEP SHA SEP SIZE SEP EXECUTABLE
        DIR ::= "dir: " COMMON
        SYMLINK ::= "symlink: " COMMON SEP TARGET_UTF8
        TREE ::= "tree: " COMMON REFERENCE_REVISION
        COMMON ::= FILE_ID SEP PARENT_ID SEP NAME_UTF8 SEP REVISION
        SEP ::= "\n"
        """
        if entry.parent_id is not None:
            parent_str = entry.parent_id
        else:
            parent_str = ''
        name_str = entry.name.encode("utf8")
        if entry.kind == 'file':
            if entry.executable:
                exec_str = "Y"
            else:
                exec_str = "N"
            return "file: %s\n%s\n%s\n%s\n%s\n%d\n%s" % (
                entry.file_id, parent_str, name_str, entry.revision,
                entry.text_sha1, entry.text_size, exec_str)
        elif entry.kind == 'directory':
            return "dir: %s\n%s\n%s\n%s" % (
                entry.file_id, parent_str, name_str, entry.revision)
        elif entry.kind == 'symlink':
            return "symlink: %s\n%s\n%s\n%s\n%s" % (
                entry.file_id, parent_str, name_str, entry.revision,
                entry.symlink_target.encode("utf8"))
        elif entry.kind == 'tree-reference':
            return "tree: %s\n%s\n%s\n%s\n%s" % (
                entry.file_id, parent_str, name_str, entry.revision,
                entry.reference_revision)
        else:
            raise ValueError("unknown kind %r" % entry.kind)

    def _expand_fileids_to_parents_and_children(self, file_ids):
        """Give a more wholistic view starting with the given file_ids.

        For any file_id which maps to a directory, we will include all children
        of that directory. We will also include all directories which are
        parents of the given file_ids, but we will not include their children.

        eg:
          /     # TREE_ROOT
          foo/  # foo-id
            baz # baz-id
            frob/ # frob-id
              fringle # fringle-id
          bar/  # bar-id
            bing # bing-id

        if given [foo-id] we will include
            TREE_ROOT as interesting parents
        and 
            foo-id, baz-id, frob-id, fringle-id
        As interesting ids.
        """
        interesting = set()
        # TODO: Pre-pass over the list of fileids to see if anything is already
        #       deserialized in self._fileid_to_entry_cache

        directories_to_expand = set()
        children_of_parent_id = {}
        # It is okay if some of the fileids are missing
        for entry in self._getitems(file_ids):
            if entry.kind == 'directory':
                directories_to_expand.add(entry.file_id)
            interesting.add(entry.parent_id)
            children_of_parent_id.setdefault(entry.parent_id, []
                                             ).append(entry.file_id)

        # Now, interesting has all of the direct parents, but not the
        # parents of those parents. It also may have some duplicates with
        # specific_fileids
        remaining_parents = interesting.difference(file_ids)
        # When we hit the TREE_ROOT, we'll get an interesting parent of None,
        # but we don't actually want to recurse into that
        interesting.add(None) # this will auto-filter it in the loop
        remaining_parents.discard(None) 
        while remaining_parents:
            next_parents = set()
            for entry in self._getitems(remaining_parents):
                next_parents.add(entry.parent_id)
                children_of_parent_id.setdefault(entry.parent_id, []
                                                 ).append(entry.file_id)
            # Remove any search tips we've already processed
            remaining_parents = next_parents.difference(interesting)
            interesting.update(remaining_parents)
            # We should probably also .difference(directories_to_expand)
        interesting.update(file_ids)
        interesting.discard(None)
        while directories_to_expand:
            # Expand directories by looking in the
            # parent_id_basename_to_file_id map
            keys = [StaticTuple(f,).intern() for f in directories_to_expand]
            directories_to_expand = set()
            items = self.parent_id_basename_to_file_id.iteritems(keys)
            next_file_ids = set([item[1] for item in items])
            next_file_ids = next_file_ids.difference(interesting)
            interesting.update(next_file_ids)
            for entry in self._getitems(next_file_ids):
                if entry.kind == 'directory':
                    directories_to_expand.add(entry.file_id)
                children_of_parent_id.setdefault(entry.parent_id, []
                                                 ).append(entry.file_id)
        return interesting, children_of_parent_id

    def filter(self, specific_fileids):
        """Get an inventory view filtered against a set of file-ids.

        Children of directories and parents are included.

        The result may or may not reference the underlying inventory
        so it should be treated as immutable.
        """
        (interesting,
         parent_to_children) = self._expand_fileids_to_parents_and_children(
                                specific_fileids)
        # There is some overlap here, but we assume that all interesting items
        # are in the _fileid_to_entry_cache because we had to read them to
        # determine if they were a dir we wanted to recurse, or just a file
        # This should give us all the entries we'll want to add, so start
        # adding
        other = Inventory(self.root_id)
        other.root.revision = self.root.revision
        other.revision_id = self.revision_id
        if not interesting or not parent_to_children:
            # empty filter, or filtering entrys that don't exist
            # (if even 1 existed, then we would have populated
            # parent_to_children with at least the tree root.)
            return other
        cache = self._fileid_to_entry_cache
        remaining_children = collections.deque(parent_to_children[self.root_id])
        while remaining_children:
            file_id = remaining_children.popleft()
            ie = cache[file_id]
            if ie.kind == 'directory':
                ie = ie.copy() # We create a copy to depopulate the .children attribute
            # TODO: depending on the uses of 'other' we should probably alwyas
            #       '.copy()' to prevent someone from mutating other and
            #       invaliding our internal cache
            other.add(ie)
            if file_id in parent_to_children:
                remaining_children.extend(parent_to_children[file_id])
        return other

    @staticmethod
    def _bytes_to_utf8name_key(bytes):
        """Get the file_id, revision_id key out of bytes."""
        # We don't normally care about name, except for times when we want
        # to filter out empty names because of non rich-root...
        sections = bytes.split('\n')
        kind, file_id = sections[0].split(': ')
        return (sections[2], intern(file_id), intern(sections[3]))

    def _bytes_to_entry(self, bytes):
        """Deserialise a serialised entry."""
        sections = bytes.split('\n')
        if sections[0].startswith("file: "):
            result = InventoryFile(sections[0][6:],
                sections[2].decode('utf8'),
                sections[1])
            result.text_sha1 = sections[4]
            result.text_size = int(sections[5])
            result.executable = sections[6] == "Y"
        elif sections[0].startswith("dir: "):
            result = CHKInventoryDirectory(sections[0][5:],
                sections[2].decode('utf8'),
                sections[1], self)
        elif sections[0].startswith("symlink: "):
            result = InventoryLink(sections[0][9:],
                sections[2].decode('utf8'),
                sections[1])
            result.symlink_target = sections[4].decode('utf8')
        elif sections[0].startswith("tree: "):
            result = TreeReference(sections[0][6:],
                sections[2].decode('utf8'),
                sections[1])
            result.reference_revision = sections[4]
        else:
            raise ValueError("Not a serialised entry %r" % bytes)
        result.file_id = intern(result.file_id)
        result.revision = intern(sections[3])
        if result.parent_id == '':
            result.parent_id = None
        self._fileid_to_entry_cache[result.file_id] = result
        return result

    def _get_mutable_inventory(self):
        """See CommonInventory._get_mutable_inventory."""
        entries = self.iter_entries()
        inv = Inventory(None, self.revision_id)
        for path, inv_entry in entries:
            inv.add(inv_entry.copy())
        return inv

    def create_by_apply_delta(self, inventory_delta, new_revision_id,
        propagate_caches=False):
        """Create a new CHKInventory by applying inventory_delta to this one.

        See the inventory developers documentation for the theory behind
        inventory deltas.

        :param inventory_delta: The inventory delta to apply. See
            Inventory.apply_delta for details.
        :param new_revision_id: The revision id of the resulting CHKInventory.
        :param propagate_caches: If True, the caches for this inventory are
          copied to and updated for the result.
        :return: The new CHKInventory.
        """
        split = osutils.split
        result = CHKInventory(self._search_key_name)
        if propagate_caches:
            # Just propagate the path-to-fileid cache for now
            result._path_to_fileid_cache = dict(self._path_to_fileid_cache.iteritems())
        search_key_func = chk_map.search_key_registry.get(self._search_key_name)
        self.id_to_entry._ensure_root()
        maximum_size = self.id_to_entry._root_node.maximum_size
        result.revision_id = new_revision_id
        result.id_to_entry = chk_map.CHKMap(
            self.id_to_entry._store,
            self.id_to_entry.key(),
            search_key_func=search_key_func)
        result.id_to_entry._ensure_root()
        result.id_to_entry._root_node.set_maximum_size(maximum_size)
        # Change to apply to the parent_id_basename delta. The dict maps
        # (parent_id, basename) -> (old_key, new_value). We use a dict because
        # when a path has its id replaced (e.g. the root is changed, or someone
        # does bzr mv a b, bzr mv c a, we should output a single change to this
        # map rather than two.
        parent_id_basename_delta = {}
        if self.parent_id_basename_to_file_id is not None:
            result.parent_id_basename_to_file_id = chk_map.CHKMap(
                self.parent_id_basename_to_file_id._store,
                self.parent_id_basename_to_file_id.key(),
                search_key_func=search_key_func)
            result.parent_id_basename_to_file_id._ensure_root()
            self.parent_id_basename_to_file_id._ensure_root()
            result_p_id_root = result.parent_id_basename_to_file_id._root_node
            p_id_root = self.parent_id_basename_to_file_id._root_node
            result_p_id_root.set_maximum_size(p_id_root.maximum_size)
            result_p_id_root._key_width = p_id_root._key_width
        else:
            result.parent_id_basename_to_file_id = None
        result.root_id = self.root_id
        id_to_entry_delta = []
        # inventory_delta is only traversed once, so we just update the
        # variable.
        # Check for repeated file ids
        inventory_delta = _check_delta_unique_ids(inventory_delta)
        # Repeated old paths
        inventory_delta = _check_delta_unique_old_paths(inventory_delta)
        # Check for repeated new paths
        inventory_delta = _check_delta_unique_new_paths(inventory_delta)
        # Check for entries that don't match the fileid
        inventory_delta = _check_delta_ids_match_entry(inventory_delta)
        # Check for nonsense fileids
        inventory_delta = _check_delta_ids_are_valid(inventory_delta)
        # Check for new_path <-> entry consistency
        inventory_delta = _check_delta_new_path_entry_both_or_None(
            inventory_delta)
        # All changed entries need to have their parents be directories and be
        # at the right path. This set contains (path, id) tuples.
        parents = set()
        # When we delete an item, all the children of it must be either deleted
        # or altered in their own right. As we batch process the change via
        # CHKMap.apply_delta, we build a set of things to use to validate the
        # delta.
        deletes = set()
        altered = set()
        for old_path, new_path, file_id, entry in inventory_delta:
            # file id changes
            if new_path == '':
                result.root_id = file_id
            if new_path is None:
                # Make a delete:
                new_key = None
                new_value = None
                # Update caches
                if propagate_caches:
                    try:
                        del result._path_to_fileid_cache[old_path]
                    except KeyError:
                        pass
                deletes.add(file_id)
            else:
                new_key = StaticTuple(file_id,)
                new_value = result._entry_to_bytes(entry)
                # Update caches. It's worth doing this whether
                # we're propagating the old caches or not.
                result._path_to_fileid_cache[new_path] = file_id
                parents.add((split(new_path)[0], entry.parent_id))
            if old_path is None:
                old_key = None
            else:
                old_key = StaticTuple(file_id,)
                if self.id2path(file_id) != old_path:
                    raise errors.InconsistentDelta(old_path, file_id,
                        "Entry was at wrong other path %r." %
                        self.id2path(file_id))
                altered.add(file_id)
            id_to_entry_delta.append(StaticTuple(old_key, new_key, new_value))
            if result.parent_id_basename_to_file_id is not None:
                # parent_id, basename changes
                if old_path is None:
                    old_key = None
                else:
                    old_entry = self[file_id]
                    old_key = self._parent_id_basename_key(old_entry)
                if new_path is None:
                    new_key = None
                    new_value = None
                else:
                    new_key = self._parent_id_basename_key(entry)
                    new_value = file_id
                # If the two keys are the same, the value will be unchanged
                # as its always the file id for this entry.
                if old_key != new_key:
                    # Transform a change into explicit delete/add preserving
                    # a possible match on the key from a different file id.
                    if old_key is not None:
                        parent_id_basename_delta.setdefault(
                            old_key, [None, None])[0] = old_key
                    if new_key is not None:
                        parent_id_basename_delta.setdefault(
                            new_key, [None, None])[1] = new_value
        # validate that deletes are complete.
        for file_id in deletes:
            entry = self[file_id]
            if entry.kind != 'directory':
                continue
            # This loop could potentially be better by using the id_basename
            # map to just get the child file ids.
            for child in entry.children.values():
                if child.file_id not in altered:
                    raise errors.InconsistentDelta(self.id2path(child.file_id),
                        child.file_id, "Child not deleted or reparented when "
                        "parent deleted.")
        result.id_to_entry.apply_delta(id_to_entry_delta)
        if parent_id_basename_delta:
            # Transform the parent_id_basename delta data into a linear delta
            # with only one record for a given key. Optimally this would allow
            # re-keying, but its simpler to just output that as a delete+add
            # to spend less time calculating the delta.
            delta_list = []
            for key, (old_key, value) in parent_id_basename_delta.iteritems():
                if value is not None:
                    delta_list.append((old_key, key, value))
                else:
                    delta_list.append((old_key, None, None))
            result.parent_id_basename_to_file_id.apply_delta(delta_list)
        parents.discard(('', None))
        for parent_path, parent in parents:
            try:
                if result[parent].kind != 'directory':
                    raise errors.InconsistentDelta(result.id2path(parent), parent,
                        'Not a directory, but given children')
            except errors.NoSuchId:
                raise errors.InconsistentDelta("<unknown>", parent,
                    "Parent is not present in resulting inventory.")
            if result.path2id(parent_path) != parent:
                raise errors.InconsistentDelta(parent_path, parent,
                    "Parent has wrong path %r." % result.path2id(parent_path))
        return result

    @classmethod
    def deserialise(klass, chk_store, bytes, expected_revision_id):
        """Deserialise a CHKInventory.

        :param chk_store: A CHK capable VersionedFiles instance.
        :param bytes: The serialised bytes.
        :param expected_revision_id: The revision ID we think this inventory is
            for.
        :return: A CHKInventory
        """
        lines = bytes.split('\n')
        if lines[-1] != '':
            raise AssertionError('bytes to deserialize must end with an eol')
        lines.pop()
        if lines[0] != 'chkinventory:':
            raise ValueError("not a serialised CHKInventory: %r" % bytes)
        info = {}
        allowed_keys = frozenset(['root_id', 'revision_id', 'search_key_name',
                                  'parent_id_basename_to_file_id',
                                  'id_to_entry'])
        for line in lines[1:]:
            key, value = line.split(': ', 1)
            if key not in allowed_keys:
                raise errors.BzrError('Unknown key in inventory: %r\n%r'
                                      % (key, bytes))
            if key in info:
                raise errors.BzrError('Duplicate key in inventory: %r\n%r'
                                      % (key, bytes))
            info[key] = value
        revision_id = intern(info['revision_id'])
        root_id = intern(info['root_id'])
        search_key_name = intern(info.get('search_key_name', 'plain'))
        parent_id_basename_to_file_id = intern(info.get(
            'parent_id_basename_to_file_id', None))
        if not parent_id_basename_to_file_id.startswith('sha1:'):
            raise ValueError('parent_id_basename_to_file_id should be a sha1'
                             ' key not %r' % (parent_id_basename_to_file_id,))
        id_to_entry = info['id_to_entry']
        if not id_to_entry.startswith('sha1:'):
            raise ValueError('id_to_entry should be a sha1'
                             ' key not %r' % (id_to_entry,))

        result = CHKInventory(search_key_name)
        result.revision_id = revision_id
        result.root_id = root_id
        search_key_func = chk_map.search_key_registry.get(
                            result._search_key_name)
        if parent_id_basename_to_file_id is not None:
            result.parent_id_basename_to_file_id = chk_map.CHKMap(
                chk_store, StaticTuple(parent_id_basename_to_file_id,),
                search_key_func=search_key_func)
        else:
            result.parent_id_basename_to_file_id = None

        result.id_to_entry = chk_map.CHKMap(chk_store,
                                            StaticTuple(id_to_entry,),
                                            search_key_func=search_key_func)
        if (result.revision_id,) != expected_revision_id:
            raise ValueError("Mismatched revision id and expected: %r, %r" %
                (result.revision_id, expected_revision_id))
        return result

    @classmethod
    def from_inventory(klass, chk_store, inventory, maximum_size=0, search_key_name='plain'):
        """Create a CHKInventory from an existing inventory.

        The content of inventory is copied into the chk_store, and a
        CHKInventory referencing that is returned.

        :param chk_store: A CHK capable VersionedFiles instance.
        :param inventory: The inventory to copy.
        :param maximum_size: The CHKMap node size limit.
        :param search_key_name: The identifier for the search key function
        """
        result = klass(search_key_name)
        result.revision_id = inventory.revision_id
        result.root_id = inventory.root.file_id

        entry_to_bytes = result._entry_to_bytes
        parent_id_basename_key = result._parent_id_basename_key
        id_to_entry_dict = {}
        parent_id_basename_dict = {}
        for path, entry in inventory.iter_entries():
            key = StaticTuple(entry.file_id,).intern()
            id_to_entry_dict[key] = entry_to_bytes(entry)
            p_id_key = parent_id_basename_key(entry)
            parent_id_basename_dict[p_id_key] = entry.file_id

        result._populate_from_dicts(chk_store, id_to_entry_dict,
            parent_id_basename_dict, maximum_size=maximum_size)
        return result

    def _populate_from_dicts(self, chk_store, id_to_entry_dict,
                             parent_id_basename_dict, maximum_size):
        search_key_func = chk_map.search_key_registry.get(self._search_key_name)
        root_key = chk_map.CHKMap.from_dict(chk_store, id_to_entry_dict,
                   maximum_size=maximum_size, key_width=1,
                   search_key_func=search_key_func)
        self.id_to_entry = chk_map.CHKMap(chk_store, root_key,
                                          search_key_func)
        root_key = chk_map.CHKMap.from_dict(chk_store,
                   parent_id_basename_dict,
                   maximum_size=maximum_size, key_width=2,
                   search_key_func=search_key_func)
        self.parent_id_basename_to_file_id = chk_map.CHKMap(chk_store,
                                                    root_key, search_key_func)

    def _parent_id_basename_key(self, entry):
        """Create a key for a entry in a parent_id_basename_to_file_id index."""
        if entry.parent_id is not None:
            parent_id = entry.parent_id
        else:
            parent_id = ''
        return StaticTuple(parent_id, entry.name.encode('utf8')).intern()

    def __getitem__(self, file_id):
        """map a single file_id -> InventoryEntry."""
        if file_id is None:
            raise errors.NoSuchId(self, file_id)
        result = self._fileid_to_entry_cache.get(file_id, None)
        if result is not None:
            return result
        try:
            return self._bytes_to_entry(
                self.id_to_entry.iteritems([StaticTuple(file_id,)]).next()[1])
        except StopIteration:
            # really we're passing an inventory, not a tree...
            raise errors.NoSuchId(self, file_id)

    def _getitems(self, file_ids):
        """Similar to __getitem__, but lets you query for multiple.
        
        The returned order is undefined. And currently if an item doesn't
        exist, it isn't included in the output.
        """
        result = []
        remaining = []
        for file_id in file_ids:
            entry = self._fileid_to_entry_cache.get(file_id, None)
            if entry is None:
                remaining.append(file_id)
            else:
                result.append(entry)
        file_keys = [StaticTuple(f,).intern() for f in remaining]
        for file_key, value in self.id_to_entry.iteritems(file_keys):
            entry = self._bytes_to_entry(value)
            result.append(entry)
            self._fileid_to_entry_cache[entry.file_id] = entry
        return result

    def has_id(self, file_id):
        # Perhaps have an explicit 'contains' method on CHKMap ?
        if self._fileid_to_entry_cache.get(file_id, None) is not None:
            return True
        return len(list(
            self.id_to_entry.iteritems([StaticTuple(file_id,)]))) == 1

    def is_root(self, file_id):
        return file_id == self.root_id

    def _iter_file_id_parents(self, file_id):
        """Yield the parents of file_id up to the root."""
        while file_id is not None:
            try:
                ie = self[file_id]
            except KeyError:
                raise errors.NoSuchId(tree=self, file_id=file_id)
            yield ie
            file_id = ie.parent_id

    def __iter__(self):
        """Iterate over all file-ids."""
        for key, _ in self.id_to_entry.iteritems():
            yield key[-1]

    def iter_just_entries(self):
        """Iterate over all entries.
        
        Unlike iter_entries(), just the entries are returned (not (path, ie))
        and the order of entries is undefined.

        XXX: We may not want to merge this into bzr.dev.
        """
        for key, entry in self.id_to_entry.iteritems():
            file_id = key[0]
            ie = self._fileid_to_entry_cache.get(file_id, None)
            if ie is None:
                ie = self._bytes_to_entry(entry)
                self._fileid_to_entry_cache[file_id] = ie
            yield ie

    def iter_changes(self, basis):
        """Generate a Tree.iter_changes change list between this and basis.

        :param basis: Another CHKInventory.
        :return: An iterator over the changes between self and basis, as per
            tree.iter_changes().
        """
        # We want: (file_id, (path_in_source, path_in_target),
        # changed_content, versioned, parent, name, kind,
        # executable)
        for key, basis_value, self_value in \
            self.id_to_entry.iter_changes(basis.id_to_entry):
            file_id = key[0]
            if basis_value is not None:
                basis_entry = basis._bytes_to_entry(basis_value)
                path_in_source = basis.id2path(file_id)
                basis_parent = basis_entry.parent_id
                basis_name = basis_entry.name
                basis_executable = basis_entry.executable
            else:
                path_in_source = None
                basis_parent = None
                basis_name = None
                basis_executable = None
            if self_value is not None:
                self_entry = self._bytes_to_entry(self_value)
                path_in_target = self.id2path(file_id)
                self_parent = self_entry.parent_id
                self_name = self_entry.name
                self_executable = self_entry.executable
            else:
                path_in_target = None
                self_parent = None
                self_name = None
                self_executable = None
            if basis_value is None:
                # add
                kind = (None, self_entry.kind)
                versioned = (False, True)
            elif self_value is None:
                # delete
                kind = (basis_entry.kind, None)
                versioned = (True, False)
            else:
                kind = (basis_entry.kind, self_entry.kind)
                versioned = (True, True)
            changed_content = False
            if kind[0] != kind[1]:
                changed_content = True
            elif kind[0] == 'file':
                if (self_entry.text_size != basis_entry.text_size or
                    self_entry.text_sha1 != basis_entry.text_sha1):
                    changed_content = True
            elif kind[0] == 'symlink':
                if self_entry.symlink_target != basis_entry.symlink_target:
                    changed_content = True
            elif kind[0] == 'tree-reference':
                if (self_entry.reference_revision !=
                    basis_entry.reference_revision):
                    changed_content = True
            parent = (basis_parent, self_parent)
            name = (basis_name, self_name)
            executable = (basis_executable, self_executable)
            if (not changed_content
                and parent[0] == parent[1]
                and name[0] == name[1]
                and executable[0] == executable[1]):
                # Could happen when only the revision changed for a directory
                # for instance.
                continue
            yield (file_id, (path_in_source, path_in_target), changed_content,
                versioned, parent, name, kind, executable)

    def __len__(self):
        """Return the number of entries in the inventory."""
        return len(self.id_to_entry)

    def _make_delta(self, old):
        """Make an inventory delta from two inventories."""
        if type(old) != CHKInventory:
            return CommonInventory._make_delta(self, old)
        delta = []
        for key, old_value, self_value in \
            self.id_to_entry.iter_changes(old.id_to_entry):
            file_id = key[0]
            if old_value is not None:
                old_path = old.id2path(file_id)
            else:
                old_path = None
            if self_value is not None:
                entry = self._bytes_to_entry(self_value)
                self._fileid_to_entry_cache[file_id] = entry
                new_path = self.id2path(file_id)
            else:
                entry = None
                new_path = None
            delta.append((old_path, new_path, file_id, entry))
        return delta

    def path2id(self, relpath):
        """See CommonInventory.path2id()."""
        # TODO: perhaps support negative hits?
        result = self._path_to_fileid_cache.get(relpath, None)
        if result is not None:
            return result
        if isinstance(relpath, basestring):
            names = osutils.splitpath(relpath)
        else:
            names = relpath
        current_id = self.root_id
        if current_id is None:
            return None
        parent_id_index = self.parent_id_basename_to_file_id
        cur_path = None
        for basename in names:
            if cur_path is None:
                cur_path = basename
            else:
                cur_path = cur_path + '/' + basename
            basename_utf8 = basename.encode('utf8')
            file_id = self._path_to_fileid_cache.get(cur_path, None)
            if file_id is None:
                key_filter = [StaticTuple(current_id, basename_utf8)]
                items = parent_id_index.iteritems(key_filter)
                for (parent_id, name_utf8), file_id in items:
                    if parent_id != current_id or name_utf8 != basename_utf8:
                        raise errors.BzrError("corrupt inventory lookup! "
                            "%r %r %r %r" % (parent_id, current_id, name_utf8,
                            basename_utf8))
                if file_id is None:
                    return None
                else:
                    self._path_to_fileid_cache[cur_path] = file_id
            current_id = file_id
        return current_id

    def to_lines(self):
        """Serialise the inventory to lines."""
        lines = ["chkinventory:\n"]
        if self._search_key_name != 'plain':
            # custom ordering grouping things that don't change together
            lines.append('search_key_name: %s\n' % (self._search_key_name,))
            lines.append("root_id: %s\n" % self.root_id)
            lines.append('parent_id_basename_to_file_id: %s\n' %
                (self.parent_id_basename_to_file_id.key()[0],))
            lines.append("revision_id: %s\n" % self.revision_id)
            lines.append("id_to_entry: %s\n" % (self.id_to_entry.key()[0],))
        else:
            lines.append("revision_id: %s\n" % self.revision_id)
            lines.append("root_id: %s\n" % self.root_id)
            if self.parent_id_basename_to_file_id is not None:
                lines.append('parent_id_basename_to_file_id: %s\n' %
                    (self.parent_id_basename_to_file_id.key()[0],))
            lines.append("id_to_entry: %s\n" % (self.id_to_entry.key()[0],))
        return lines

    @property
    def root(self):
        """Get the root entry."""
        return self[self.root_id]


class CHKInventoryDirectory(InventoryDirectory):
    """A directory in an inventory."""

    __slots__ = ['text_sha1', 'text_size', 'file_id', 'name', 'kind',
                 'text_id', 'parent_id', '_children', 'executable',
                 'revision', 'symlink_target', 'reference_revision',
                 '_chk_inventory']

    def __init__(self, file_id, name, parent_id, chk_inventory):
        # Don't call InventoryDirectory.__init__ - it isn't right for this
        # class.
        InventoryEntry.__init__(self, file_id, name, parent_id)
        self._children = None
        self.kind = 'directory'
        self._chk_inventory = chk_inventory

    @property
    def children(self):
        """Access the list of children of this directory.

        With a parent_id_basename_to_file_id index, loads all the children,
        without loads the entire index. Without is bad. A more sophisticated
        proxy object might be nice, to allow partial loading of children as
        well when specific names are accessed. (So path traversal can be
        written in the obvious way but not examine siblings.).
        """
        if self._children is not None:
            return self._children
        # No longer supported
        if self._chk_inventory.parent_id_basename_to_file_id is None:
            raise AssertionError("Inventories without"
                " parent_id_basename_to_file_id are no longer supported")
        result = {}
        # XXX: Todo - use proxy objects for the children rather than loading
        # all when the attribute is referenced.
        parent_id_index = self._chk_inventory.parent_id_basename_to_file_id
        child_keys = set()
        for (parent_id, name_utf8), file_id in parent_id_index.iteritems(
            key_filter=[StaticTuple(self.file_id,)]):
            child_keys.add(StaticTuple(file_id,))
        cached = set()
        for file_id_key in child_keys:
            entry = self._chk_inventory._fileid_to_entry_cache.get(
                file_id_key[0], None)
            if entry is not None:
                result[entry.name] = entry
                cached.add(file_id_key)
        child_keys.difference_update(cached)
        # populate; todo: do by name
        id_to_entry = self._chk_inventory.id_to_entry
        for file_id_key, bytes in id_to_entry.iteritems(child_keys):
            entry = self._chk_inventory._bytes_to_entry(bytes)
            result[entry.name] = entry
            self._chk_inventory._fileid_to_entry_cache[file_id_key[0]] = entry
        self._children = result
        return result

entry_factory = {
    'directory': InventoryDirectory,
    'file': InventoryFile,
    'symlink': InventoryLink,
    'tree-reference': TreeReference
}

def make_entry(kind, name, parent_id, file_id=None):
    """Create an inventory entry.

    :param kind: the type of inventory entry to create.
    :param name: the basename of the entry.
    :param parent_id: the parent_id of the entry.
    :param file_id: the file_id to use. if None, one will be created.
    """
    if file_id is None:
        file_id = generate_ids.gen_file_id(name)
    name = ensure_normalized_name(name)
    try:
        factory = entry_factory[kind]
    except KeyError:
        raise errors.BadFileKindError(name, kind)
    return factory(file_id, name, parent_id)


def ensure_normalized_name(name):
    """Normalize name.

    :raises InvalidNormalization: When name is not normalized, and cannot be
        accessed on this platform by the normalized path.
    :return: The NFC normalised version of name.
    """
    #------- This has been copied to bzrlib.dirstate.DirState.add, please
    # keep them synchronised.
    # we dont import normalized_filename directly because we want to be
    # able to change the implementation at runtime for tests.
    norm_name, can_access = osutils.normalized_filename(name)
    if norm_name != name:
        if can_access:
            return norm_name
        else:
            # TODO: jam 20060701 This would probably be more useful
            #       if the error was raised with the full path
            raise errors.InvalidNormalization(name)
    return name


_NAME_RE = None

def is_valid_name(name):
    global _NAME_RE
    if _NAME_RE is None:
        _NAME_RE = re.compile(r'^[^/\\]+$')

    return bool(_NAME_RE.match(name))


def _check_delta_unique_ids(delta):
    """Decorate a delta and check that the file ids in it are unique.

    :return: A generator over delta.
    """
    ids = set()
    for item in delta:
        length = len(ids) + 1
        ids.add(item[2])
        if len(ids) != length:
            raise errors.InconsistentDelta(item[0] or item[1], item[2],
                "repeated file_id")
        yield item


def _check_delta_unique_new_paths(delta):
    """Decorate a delta and check that the new paths in it are unique.

    :return: A generator over delta.
    """
    paths = set()
    for item in delta:
        length = len(paths) + 1
        path = item[1]
        if path is not None:
            paths.add(path)
            if len(paths) != length:
                raise errors.InconsistentDelta(path, item[2], "repeated path")
        yield item


def _check_delta_unique_old_paths(delta):
    """Decorate a delta and check that the old paths in it are unique.

    :return: A generator over delta.
    """
    paths = set()
    for item in delta:
        length = len(paths) + 1
        path = item[0]
        if path is not None:
            paths.add(path)
            if len(paths) != length:
                raise errors.InconsistentDelta(path, item[2], "repeated path")
        yield item


def _check_delta_ids_are_valid(delta):
    """Decorate a delta and check that the ids in it are valid.

    :return: A generator over delta.
    """
    for item in delta:
        entry = item[3]
        if item[2] is None:
            raise errors.InconsistentDelta(item[0] or item[1], item[2],
                "entry with file_id None %r" % entry)
        if type(item[2]) != str:
            raise errors.InconsistentDelta(item[0] or item[1], item[2],
                "entry with non bytes file_id %r" % entry)
        yield item


def _check_delta_ids_match_entry(delta):
    """Decorate a delta and check that the ids in it match the entry.file_id.

    :return: A generator over delta.
    """
    for item in delta:
        entry = item[3]
        if entry is not None:
            if entry.file_id != item[2]:
                raise errors.InconsistentDelta(item[0] or item[1], item[2],
                    "mismatched id with %r" % entry)
        yield item


def _check_delta_new_path_entry_both_or_None(delta):
    """Decorate a delta and check that the new_path and entry are paired.

    :return: A generator over delta.
    """
    for item in delta:
        new_path = item[1]
        entry = item[3]
        if new_path is None and entry is not None:
            raise errors.InconsistentDelta(item[0], item[1],
                "Entry with no new_path")
        if new_path is not None and entry is None:
            raise errors.InconsistentDelta(new_path, item[1],
                "new_path with no entry")
        yield item