~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
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
2479
2480
2481
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
2533
2534
2535
2536
2537
2538
2539
2540
2541
2542
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
2566
2567
2568
2569
2570
2571
2572
2573
2574
2575
2576
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
2587
2588
2589
2590
2591
2592
2593
2594
2595
2596
2597
2598
2599
2600
2601
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
2630
2631
2632
2633
2634
2635
2636
2637
2638
2639
2640
2641
2642
2643
2644
2645
2646
2647
2648
2649
2650
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
2668
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
2687
2688
2689
2690
2691
2692
2693
2694
2695
2696
2697
2698
2699
2700
2701
2702
2703
2704
2705
2706
2707
2708
2709
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
2735
2736
2737
2738
2739
2740
2741
2742
2743
2744
2745
2746
2747
2748
2749
2750
2751
2752
2753
2754
2755
2756
2757
2758
2759
2760
2761
2762
2763
2764
2765
2766
2767
2768
2769
2770
2771
2772
2773
2774
2775
2776
2777
2778
2779
2780
2781
2782
2783
2784
2785
2786
2787
2788
2789
2790
2791
2792
2793
2794
2795
2796
2797
2798
2799
2800
2801
2802
2803
2804
2805
2806
2807
2808
2809
2810
2811
2812
2813
2814
2815
2816
2817
2818
2819
2820
2821
2822
2823
2824
2825
2826
2827
2828
2829
2830
2831
2832
2833
2834
2835
2836
2837
2838
2839
2840
2841
2842
2843
2844
2845
2846
2847
2848
2849
2850
2851
2852
2853
2854
2855
2856
2857
2858
2859
2860
2861
2862
2863
2864
2865
2866
2867
2868
2869
2870
2871
2872
2873
2874
2875
2876
2877
2878
2879
2880
2881
2882
2883
2884
2885
2886
2887
2888
2889
2890
2891
2892
2893
2894
2895
2896
2897
2898
2899
2900
2901
2902
2903
2904
2905
2906
2907
2908
2909
2910
2911
2912
2913
2914
2915
2916
2917
2918
2919
2920
2921
2922
2923
2924
2925
2926
2927
2928
2929
2930
2931
2932
2933
2934
2935
2936
2937
2938
2939
2940
2941
2942
2943
2944
2945
2946
2947
2948
2949
2950
2951
2952
2953
2954
2955
2956
2957
2958
2959
2960
2961
2962
2963
2964
2965
2966
2967
2968
2969
2970
2971
2972
2973
2974
2975
2976
2977
2978
2979
2980
2981
2982
2983
2984
2985
2986
2987
2988
2989
2990
2991
2992
2993
2994
2995
2996
2997
2998
2999
3000
3001
3002
3003
3004
3005
3006
3007
3008
3009
3010
3011
3012
3013
3014
3015
3016
3017
3018
3019
3020
3021
3022
3023
3024
3025
3026
3027
3028
3029
3030
3031
3032
3033
3034
3035
3036
3037
3038
3039
3040
3041
3042
3043
3044
3045
3046
3047
3048
3049
3050
3051
3052
3053
3054
3055
3056
3057
3058
3059
3060
# Copyright (C) 2005, 2006, 2007, 2008, 2009 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

"""WorkingTree object and friends.

A WorkingTree represents the editable working copy of a branch.
Operations which represent the WorkingTree are also done here,
such as renaming or adding files.  The WorkingTree has an inventory
which is updated by these operations.  A commit produces a
new revision based on the workingtree and its inventory.

At the moment every WorkingTree has its own branch.  Remote
WorkingTrees aren't supported.

To get a WorkingTree, call bzrdir.open_workingtree() or
WorkingTree.open(dir).
"""

# TODO: Give the workingtree sole responsibility for the working inventory;
# remove the variable and references to it from the branch.  This may require
# updating the commit code so as to update the inventory within the working
# copy, and making sure there's only one WorkingTree for any directory on disk.
# At the moment they may alias the inventory and have old copies of it in
# memory.  (Now done? -- mbp 20060309)

from cStringIO import StringIO
import os
import sys

from bzrlib.lazy_import import lazy_import
lazy_import(globals(), """
from bisect import bisect_left
import collections
import errno
import itertools
import operator
import stat
import re

import bzrlib
from bzrlib import (
    branch,
    bzrdir,
    conflicts as _mod_conflicts,
    errors,
    generate_ids,
    globbing,
    graph as _mod_graph,
    hashcache,
    ignores,
    inventory,
    merge,
    revision as _mod_revision,
    revisiontree,
    trace,
    transform,
    ui,
    views,
    xml5,
    xml7,
    )
import bzrlib.branch
from bzrlib.transport import get_transport
from bzrlib.workingtree_4 import (
    WorkingTreeFormat4,
    WorkingTreeFormat5,
    WorkingTreeFormat6,
    )
""")

from bzrlib import symbol_versioning
from bzrlib.decorators import needs_read_lock, needs_write_lock
from bzrlib.lockable_files import LockableFiles
from bzrlib.lockdir import LockDir
import bzrlib.mutabletree
from bzrlib.mutabletree import needs_tree_write_lock
from bzrlib import osutils
from bzrlib.osutils import (
    file_kind,
    isdir,
    normpath,
    pathjoin,
    realpath,
    safe_unicode,
    splitpath,
    supports_executable,
    )
from bzrlib.filters import filtered_input_file
from bzrlib.trace import mutter, note
from bzrlib.transport.local import LocalTransport
from bzrlib.progress import DummyProgress, ProgressPhase
from bzrlib.revision import CURRENT_REVISION
from bzrlib.rio import RioReader, rio_file, Stanza
from bzrlib.symbol_versioning import (
    deprecated_passed,
    DEPRECATED_PARAMETER,
    )


MERGE_MODIFIED_HEADER_1 = "BZR merge-modified list format 1"
CONFLICT_HEADER_1 = "BZR conflict list format 1"

ERROR_PATH_NOT_FOUND = 3    # WindowsError errno code, equivalent to ENOENT


class TreeEntry(object):
    """An entry that implements the minimum interface used by commands.

    This needs further inspection, it may be better to have
    InventoryEntries without ids - though that seems wrong. For now,
    this is a parallel hierarchy to InventoryEntry, and needs to become
    one of several things: decorates to that hierarchy, children of, or
    parents of it.
    Another note is that these objects are currently only used when there is
    no InventoryEntry available - i.e. for unversioned objects.
    Perhaps they should be UnversionedEntry et al. ? - RBC 20051003
    """

    def __eq__(self, other):
        # yes, this us ugly, TODO: best practice __eq__ style.
        return (isinstance(other, TreeEntry)
                and other.__class__ == self.__class__)

    def kind_character(self):
        return "???"


class TreeDirectory(TreeEntry):
    """See TreeEntry. This is a directory in a working tree."""

    def __eq__(self, other):
        return (isinstance(other, TreeDirectory)
                and other.__class__ == self.__class__)

    def kind_character(self):
        return "/"


class TreeFile(TreeEntry):
    """See TreeEntry. This is a regular file in a working tree."""

    def __eq__(self, other):
        return (isinstance(other, TreeFile)
                and other.__class__ == self.__class__)

    def kind_character(self):
        return ''


class TreeLink(TreeEntry):
    """See TreeEntry. This is a symlink in a working tree."""

    def __eq__(self, other):
        return (isinstance(other, TreeLink)
                and other.__class__ == self.__class__)

    def kind_character(self):
        return ''


class WorkingTree(bzrlib.mutabletree.MutableTree):
    """Working copy tree.

    The inventory is held in the `Branch` working-inventory, and the
    files are in a directory on disk.

    It is possible for a `WorkingTree` to have a filename which is
    not listed in the Inventory and vice versa.
    """

    # override this to set the strategy for storing views
    def _make_views(self):
        return views.DisabledViews(self)

    def __init__(self, basedir='.',
                 branch=DEPRECATED_PARAMETER,
                 _inventory=None,
                 _control_files=None,
                 _internal=False,
                 _format=None,
                 _bzrdir=None):
        """Construct a WorkingTree instance. This is not a public API.

        :param branch: A branch to override probing for the branch.
        """
        self._format = _format
        self.bzrdir = _bzrdir
        if not _internal:
            raise errors.BzrError("Please use bzrdir.open_workingtree or "
                "WorkingTree.open() to obtain a WorkingTree.")
        basedir = safe_unicode(basedir)
        mutter("opening working tree %r", basedir)
        if deprecated_passed(branch):
            self._branch = branch
        else:
            self._branch = self.bzrdir.open_branch()
        self.basedir = realpath(basedir)
        # if branch is at our basedir and is a format 6 or less
        if isinstance(self._format, WorkingTreeFormat2):
            # share control object
            self._control_files = self.branch.control_files
        else:
            # assume all other formats have their own control files.
            self._control_files = _control_files
        self._transport = self._control_files._transport
        # update the whole cache up front and write to disk if anything changed;
        # in the future we might want to do this more selectively
        # two possible ways offer themselves : in self._unlock, write the cache
        # if needed, or, when the cache sees a change, append it to the hash
        # cache file, and have the parser take the most recent entry for a
        # given path only.
        wt_trans = self.bzrdir.get_workingtree_transport(None)
        cache_filename = wt_trans.local_abspath('stat-cache')
        self._hashcache = hashcache.HashCache(basedir, cache_filename,
            self.bzrdir._get_file_mode(),
            self._content_filter_stack_provider())
        hc = self._hashcache
        hc.read()
        # is this scan needed ? it makes things kinda slow.
        #hc.scan()

        if hc.needs_write:
            mutter("write hc")
            hc.write()

        if _inventory is None:
            # This will be acquired on lock_read() or lock_write()
            self._inventory_is_modified = False
            self._inventory = None
        else:
            # the caller of __init__ has provided an inventory,
            # we assume they know what they are doing - as its only
            # the Format factory and creation methods that are
            # permitted to do this.
            self._set_inventory(_inventory, dirty=False)
        self._detect_case_handling()
        self._rules_searcher = None
        self.views = self._make_views()

    def _detect_case_handling(self):
        wt_trans = self.bzrdir.get_workingtree_transport(None)
        try:
            wt_trans.stat("FoRMaT")
        except errors.NoSuchFile:
            self.case_sensitive = True
        else:
            self.case_sensitive = False

        self._setup_directory_is_tree_reference()

    branch = property(
        fget=lambda self: self._branch,
        doc="""The branch this WorkingTree is connected to.

            This cannot be set - it is reflective of the actual disk structure
            the working tree has been constructed from.
            """)

    def break_lock(self):
        """Break a lock if one is present from another instance.

        Uses the ui factory to ask for confirmation if the lock may be from
        an active process.

        This will probe the repository for its lock as well.
        """
        self._control_files.break_lock()
        self.branch.break_lock()

    def _get_check_refs(self):
        """Return the references needed to perform a check of this tree.
        
        The default implementation returns no refs, and is only suitable for
        trees that have no local caching and can commit on ghosts at any time.

        :seealso: bzrlib.check for details about check_refs.
        """
        return []

    def requires_rich_root(self):
        return self._format.requires_rich_root

    def supports_tree_reference(self):
        return False

    def supports_content_filtering(self):
        return self._format.supports_content_filtering()

    def supports_views(self):
        return self.views.supports_views()

    def _set_inventory(self, inv, dirty):
        """Set the internal cached inventory.

        :param inv: The inventory to set.
        :param dirty: A boolean indicating whether the inventory is the same
            logical inventory as whats on disk. If True the inventory is not
            the same and should be written to disk or data will be lost, if
            False then the inventory is the same as that on disk and any
            serialisation would be unneeded overhead.
        """
        self._inventory = inv
        self._inventory_is_modified = dirty

    @staticmethod
    def open(path=None, _unsupported=False):
        """Open an existing working tree at path.

        """
        if path is None:
            path = osutils.getcwd()
        control = bzrdir.BzrDir.open(path, _unsupported)
        return control.open_workingtree(_unsupported)

    @staticmethod
    def open_containing(path=None):
        """Open an existing working tree which has its root about path.

        This probes for a working tree at path and searches upwards from there.

        Basically we keep looking up until we find the control directory or
        run into /.  If there isn't one, raises NotBranchError.
        TODO: give this a new exception.
        If there is one, it is returned, along with the unused portion of path.

        :return: The WorkingTree that contains 'path', and the rest of path
        """
        if path is None:
            path = osutils.getcwd()
        control, relpath = bzrdir.BzrDir.open_containing(path)

        return control.open_workingtree(), relpath

    @staticmethod
    def open_downlevel(path=None):
        """Open an unsupported working tree.

        Only intended for advanced situations like upgrading part of a bzrdir.
        """
        return WorkingTree.open(path, _unsupported=True)

    @staticmethod
    def find_trees(location):
        def list_current(transport):
            return [d for d in transport.list_dir('') if d != '.bzr']
        def evaluate(bzrdir):
            try:
                tree = bzrdir.open_workingtree()
            except errors.NoWorkingTree:
                return True, None
            else:
                return True, tree
        transport = get_transport(location)
        iterator = bzrdir.BzrDir.find_bzrdirs(transport, evaluate=evaluate,
                                              list_current=list_current)
        return [t for t in iterator if t is not None]

    # should be deprecated - this is slow and in any case treating them as a
    # container is (we now know) bad style -- mbp 20070302
    ## @deprecated_method(zero_fifteen)
    def __iter__(self):
        """Iterate through file_ids for this tree.

        file_ids are in a WorkingTree if they are in the working inventory
        and the working file exists.
        """
        inv = self._inventory
        for path, ie in inv.iter_entries():
            if osutils.lexists(self.abspath(path)):
                yield ie.file_id

    def all_file_ids(self):
        """See Tree.iter_all_file_ids"""
        return set(self.inventory)

    def __repr__(self):
        return "<%s of %s>" % (self.__class__.__name__,
                               getattr(self, 'basedir', None))

    def abspath(self, filename):
        return pathjoin(self.basedir, filename)

    def basis_tree(self):
        """Return RevisionTree for the current last revision.

        If the left most parent is a ghost then the returned tree will be an
        empty tree - one obtained by calling
        repository.revision_tree(NULL_REVISION).
        """
        try:
            revision_id = self.get_parent_ids()[0]
        except IndexError:
            # no parents, return an empty revision tree.
            # in the future this should return the tree for
            # 'empty:' - the implicit root empty tree.
            return self.branch.repository.revision_tree(
                       _mod_revision.NULL_REVISION)
        try:
            return self.revision_tree(revision_id)
        except errors.NoSuchRevision:
            pass
        # No cached copy available, retrieve from the repository.
        # FIXME? RBC 20060403 should we cache the inventory locally
        # at this point ?
        try:
            return self.branch.repository.revision_tree(revision_id)
        except (errors.RevisionNotPresent, errors.NoSuchRevision):
            # the basis tree *may* be a ghost or a low level error may have
            # occurred. If the revision is present, its a problem, if its not
            # its a ghost.
            if self.branch.repository.has_revision(revision_id):
                raise
            # the basis tree is a ghost so return an empty tree.
            return self.branch.repository.revision_tree(
                       _mod_revision.NULL_REVISION)

    def _cleanup(self):
        self._flush_ignore_list_cache()

    def relpath(self, path):
        """Return the local path portion from a given path.

        The path may be absolute or relative. If its a relative path it is
        interpreted relative to the python current working directory.
        """
        return osutils.relpath(self.basedir, path)

    def has_filename(self, filename):
        return osutils.lexists(self.abspath(filename))

    def get_file(self, file_id, path=None, filtered=True):
        return self.get_file_with_stat(file_id, path, filtered=filtered)[0]

    def get_file_with_stat(self, file_id, path=None, filtered=True,
        _fstat=os.fstat):
        """See Tree.get_file_with_stat."""
        if path is None:
            path = self.id2path(file_id)
        file_obj = self.get_file_byname(path, filtered=False)
        stat_value = _fstat(file_obj.fileno())
        if filtered and self.supports_content_filtering():
            filters = self._content_filter_stack(path)
            file_obj = filtered_input_file(file_obj, filters)
        return (file_obj, stat_value)

    def get_file_text(self, file_id, path=None, filtered=True):
        return self.get_file(file_id, path=path, filtered=filtered).read()

    def get_file_byname(self, filename, filtered=True):
        path = self.abspath(filename)
        f = file(path, 'rb')
        if filtered and self.supports_content_filtering():
            filters = self._content_filter_stack(filename)
            return filtered_input_file(f, filters)
        else:
            return f

    def get_file_lines(self, file_id, path=None, filtered=True):
        """See Tree.get_file_lines()"""
        file = self.get_file(file_id, path, filtered=filtered)
        try:
            return file.readlines()
        finally:
            file.close()

    @needs_read_lock
    def annotate_iter(self, file_id, default_revision=CURRENT_REVISION):
        """See Tree.annotate_iter

        This implementation will use the basis tree implementation if possible.
        Lines not in the basis are attributed to CURRENT_REVISION

        If there are pending merges, lines added by those merges will be
        incorrectly attributed to CURRENT_REVISION (but after committing, the
        attribution will be correct).
        """
        maybe_file_parent_keys = []
        for parent_id in self.get_parent_ids():
            try:
                parent_tree = self.revision_tree(parent_id)
            except errors.NoSuchRevisionInTree:
                parent_tree = self.branch.repository.revision_tree(parent_id)
            parent_tree.lock_read()
            try:
                if file_id not in parent_tree:
                    continue
                ie = parent_tree.inventory[file_id]
                if ie.kind != 'file':
                    # Note: this is slightly unnecessary, because symlinks and
                    # directories have a "text" which is the empty text, and we
                    # know that won't mess up annotations. But it seems cleaner
                    continue
                parent_text_key = (file_id, ie.revision)
                if parent_text_key not in maybe_file_parent_keys:
                    maybe_file_parent_keys.append(parent_text_key)
            finally:
                parent_tree.unlock()
        graph = _mod_graph.Graph(self.branch.repository.texts)
        heads = graph.heads(maybe_file_parent_keys)
        file_parent_keys = []
        for key in maybe_file_parent_keys:
            if key in heads:
                file_parent_keys.append(key)

        # Now we have the parents of this content
        annotator = self.branch.repository.texts.get_annotator()
        text = self.get_file(file_id).read()
        this_key =(file_id, default_revision)
        annotator.add_special_text(this_key, file_parent_keys, text)
        annotations = [(key[-1], line)
                       for key, line in annotator.annotate_flat(this_key)]
        return annotations

    def _get_ancestors(self, default_revision):
        ancestors = set([default_revision])
        for parent_id in self.get_parent_ids():
            ancestors.update(self.branch.repository.get_ancestry(
                             parent_id, topo_sorted=False))
        return ancestors

    def get_parent_ids(self):
        """See Tree.get_parent_ids.

        This implementation reads the pending merges list and last_revision
        value and uses that to decide what the parents list should be.
        """
        last_rev = _mod_revision.ensure_null(self._last_revision())
        if _mod_revision.NULL_REVISION == last_rev:
            parents = []
        else:
            parents = [last_rev]
        try:
            merges_bytes = self._transport.get_bytes('pending-merges')
        except errors.NoSuchFile:
            pass
        else:
            for l in osutils.split_lines(merges_bytes):
                revision_id = l.rstrip('\n')
                parents.append(revision_id)
        return parents

    @needs_read_lock
    def get_root_id(self):
        """Return the id of this trees root"""
        return self._inventory.root.file_id

    def _get_store_filename(self, file_id):
        ## XXX: badly named; this is not in the store at all
        return self.abspath(self.id2path(file_id))

    @needs_read_lock
    def clone(self, to_bzrdir, revision_id=None):
        """Duplicate this working tree into to_bzr, including all state.

        Specifically modified files are kept as modified, but
        ignored and unknown files are discarded.

        If you want to make a new line of development, see bzrdir.sprout()

        revision
            If not None, the cloned tree will have its last revision set to
            revision, and difference between the source trees last revision
            and this one merged in.
        """
        # assumes the target bzr dir format is compatible.
        result = to_bzrdir.create_workingtree()
        self.copy_content_into(result, revision_id)
        return result

    @needs_read_lock
    def copy_content_into(self, tree, revision_id=None):
        """Copy the current content and user files of this tree into tree."""
        tree.set_root_id(self.get_root_id())
        if revision_id is None:
            merge.transform_tree(tree, self)
        else:
            # TODO now merge from tree.last_revision to revision (to preserve
            # user local changes)
            merge.transform_tree(tree, self)
            tree.set_parent_ids([revision_id])

    def id2abspath(self, file_id):
        return self.abspath(self.id2path(file_id))

    def has_id(self, file_id):
        # files that have been deleted are excluded
        inv = self.inventory
        if not inv.has_id(file_id):
            return False
        path = inv.id2path(file_id)
        return osutils.lexists(self.abspath(path))

    def has_or_had_id(self, file_id):
        if file_id == self.inventory.root.file_id:
            return True
        return self.inventory.has_id(file_id)

    __contains__ = has_id

    def get_file_size(self, file_id):
        """See Tree.get_file_size"""
        # XXX: this returns the on-disk size; it should probably return the
        # canonical size
        try:
            return os.path.getsize(self.id2abspath(file_id))
        except OSError, e:
            if e.errno != errno.ENOENT:
                raise
            else:
                return None

    @needs_read_lock
    def get_file_sha1(self, file_id, path=None, stat_value=None):
        if not path:
            path = self._inventory.id2path(file_id)
        return self._hashcache.get_sha1(path, stat_value)

    def get_file_mtime(self, file_id, path=None):
        if not path:
            path = self.inventory.id2path(file_id)
        return os.lstat(self.abspath(path)).st_mtime

    def _is_executable_from_path_and_stat_from_basis(self, path, stat_result):
        file_id = self.path2id(path)
        if file_id is None:
            # For unversioned files on win32, we just assume they are not
            # executable
            return False
        return self._inventory[file_id].executable

    def _is_executable_from_path_and_stat_from_stat(self, path, stat_result):
        mode = stat_result.st_mode
        return bool(stat.S_ISREG(mode) and stat.S_IEXEC & mode)

    if not supports_executable():
        def is_executable(self, file_id, path=None):
            return self._inventory[file_id].executable

        _is_executable_from_path_and_stat = \
            _is_executable_from_path_and_stat_from_basis
    else:
        def is_executable(self, file_id, path=None):
            if not path:
                path = self.id2path(file_id)
            mode = os.lstat(self.abspath(path)).st_mode
            return bool(stat.S_ISREG(mode) and stat.S_IEXEC & mode)

        _is_executable_from_path_and_stat = \
            _is_executable_from_path_and_stat_from_stat

    @needs_tree_write_lock
    def _add(self, files, ids, kinds):
        """See MutableTree._add."""
        # TODO: Re-adding a file that is removed in the working copy
        # should probably put it back with the previous ID.
        # the read and write working inventory should not occur in this
        # function - they should be part of lock_write and unlock.
        inv = self.inventory
        for f, file_id, kind in zip(files, ids, kinds):
            if file_id is None:
                inv.add_path(f, kind=kind)
            else:
                inv.add_path(f, kind=kind, file_id=file_id)
            self._inventory_is_modified = True

    @needs_tree_write_lock
    def _gather_kinds(self, files, kinds):
        """See MutableTree._gather_kinds."""
        for pos, f in enumerate(files):
            if kinds[pos] is None:
                fullpath = normpath(self.abspath(f))
                try:
                    kinds[pos] = file_kind(fullpath)
                except OSError, e:
                    if e.errno == errno.ENOENT:
                        raise errors.NoSuchFile(fullpath)

    @needs_write_lock
    def add_parent_tree_id(self, revision_id, allow_leftmost_as_ghost=False):
        """Add revision_id as a parent.

        This is equivalent to retrieving the current list of parent ids
        and setting the list to its value plus revision_id.

        :param revision_id: The revision id to add to the parent list. It may
        be a ghost revision as long as its not the first parent to be added,
        or the allow_leftmost_as_ghost parameter is set True.
        :param allow_leftmost_as_ghost: Allow the first parent to be a ghost.
        """
        parents = self.get_parent_ids() + [revision_id]
        self.set_parent_ids(parents, allow_leftmost_as_ghost=len(parents) > 1
            or allow_leftmost_as_ghost)

    @needs_tree_write_lock
    def add_parent_tree(self, parent_tuple, allow_leftmost_as_ghost=False):
        """Add revision_id, tree tuple as a parent.

        This is equivalent to retrieving the current list of parent trees
        and setting the list to its value plus parent_tuple. See also
        add_parent_tree_id - if you only have a parent id available it will be
        simpler to use that api. If you have the parent already available, using
        this api is preferred.

        :param parent_tuple: The (revision id, tree) to add to the parent list.
            If the revision_id is a ghost, pass None for the tree.
        :param allow_leftmost_as_ghost: Allow the first parent to be a ghost.
        """
        parent_ids = self.get_parent_ids() + [parent_tuple[0]]
        if len(parent_ids) > 1:
            # the leftmost may have already been a ghost, preserve that if it
            # was.
            allow_leftmost_as_ghost = True
        self.set_parent_ids(parent_ids,
            allow_leftmost_as_ghost=allow_leftmost_as_ghost)

    @needs_tree_write_lock
    def add_pending_merge(self, *revision_ids):
        # TODO: Perhaps should check at this point that the
        # history of the revision is actually present?
        parents = self.get_parent_ids()
        updated = False
        for rev_id in revision_ids:
            if rev_id in parents:
                continue
            parents.append(rev_id)
            updated = True
        if updated:
            self.set_parent_ids(parents, allow_leftmost_as_ghost=True)

    def path_content_summary(self, path, _lstat=os.lstat,
        _mapper=osutils.file_kind_from_stat_mode):
        """See Tree.path_content_summary."""
        abspath = self.abspath(path)
        try:
            stat_result = _lstat(abspath)
        except OSError, e:
            if getattr(e, 'errno', None) == errno.ENOENT:
                # no file.
                return ('missing', None, None, None)
            # propagate other errors
            raise
        kind = _mapper(stat_result.st_mode)
        if kind == 'file':
            return self._file_content_summary(path, stat_result)
        elif kind == 'directory':
            # perhaps it looks like a plain directory, but it's really a
            # reference.
            if self._directory_is_tree_reference(path):
                kind = 'tree-reference'
            return kind, None, None, None
        elif kind == 'symlink':
            target = osutils.readlink(abspath)
            return ('symlink', None, None, target)
        else:
            return (kind, None, None, None)

    def _file_content_summary(self, path, stat_result):
        size = stat_result.st_size
        executable = self._is_executable_from_path_and_stat(path, stat_result)
        # try for a stat cache lookup
        return ('file', size, executable, self._sha_from_stat(
            path, stat_result))

    def _check_parents_for_ghosts(self, revision_ids, allow_leftmost_as_ghost):
        """Common ghost checking functionality from set_parent_*.

        This checks that the left hand-parent exists if there are any
        revisions present.
        """
        if len(revision_ids) > 0:
            leftmost_id = revision_ids[0]
            if (not allow_leftmost_as_ghost and not
                self.branch.repository.has_revision(leftmost_id)):
                raise errors.GhostRevisionUnusableHere(leftmost_id)

    def _set_merges_from_parent_ids(self, parent_ids):
        merges = parent_ids[1:]
        self._transport.put_bytes('pending-merges', '\n'.join(merges),
            mode=self.bzrdir._get_file_mode())

    def _filter_parent_ids_by_ancestry(self, revision_ids):
        """Check that all merged revisions are proper 'heads'.

        This will always return the first revision_id, and any merged revisions
        which are
        """
        if len(revision_ids) == 0:
            return revision_ids
        graph = self.branch.repository.get_graph()
        heads = graph.heads(revision_ids)
        new_revision_ids = revision_ids[:1]
        for revision_id in revision_ids[1:]:
            if revision_id in heads and revision_id not in new_revision_ids:
                new_revision_ids.append(revision_id)
        if new_revision_ids != revision_ids:
            trace.mutter('requested to set revision_ids = %s,'
                         ' but filtered to %s', revision_ids, new_revision_ids)
        return new_revision_ids

    @needs_tree_write_lock
    def set_parent_ids(self, revision_ids, allow_leftmost_as_ghost=False):
        """Set the parent ids to revision_ids.

        See also set_parent_trees. This api will try to retrieve the tree data
        for each element of revision_ids from the trees repository. If you have
        tree data already available, it is more efficient to use
        set_parent_trees rather than set_parent_ids. set_parent_ids is however
        an easier API to use.

        :param revision_ids: The revision_ids to set as the parent ids of this
            working tree. Any of these may be ghosts.
        """
        self._check_parents_for_ghosts(revision_ids,
            allow_leftmost_as_ghost=allow_leftmost_as_ghost)
        for revision_id in revision_ids:
            _mod_revision.check_not_reserved_id(revision_id)

        revision_ids = self._filter_parent_ids_by_ancestry(revision_ids)

        if len(revision_ids) > 0:
            self.set_last_revision(revision_ids[0])
        else:
            self.set_last_revision(_mod_revision.NULL_REVISION)

        self._set_merges_from_parent_ids(revision_ids)

    @needs_tree_write_lock
    def set_parent_trees(self, parents_list, allow_leftmost_as_ghost=False):
        """See MutableTree.set_parent_trees."""
        parent_ids = [rev for (rev, tree) in parents_list]
        for revision_id in parent_ids:
            _mod_revision.check_not_reserved_id(revision_id)

        self._check_parents_for_ghosts(parent_ids,
            allow_leftmost_as_ghost=allow_leftmost_as_ghost)

        parent_ids = self._filter_parent_ids_by_ancestry(parent_ids)

        if len(parent_ids) == 0:
            leftmost_parent_id = _mod_revision.NULL_REVISION
            leftmost_parent_tree = None
        else:
            leftmost_parent_id, leftmost_parent_tree = parents_list[0]

        if self._change_last_revision(leftmost_parent_id):
            if leftmost_parent_tree is None:
                # If we don't have a tree, fall back to reading the
                # parent tree from the repository.
                self._cache_basis_inventory(leftmost_parent_id)
            else:
                inv = leftmost_parent_tree.inventory
                xml = self._create_basis_xml_from_inventory(
                                        leftmost_parent_id, inv)
                self._write_basis_inventory(xml)
        self._set_merges_from_parent_ids(parent_ids)

    @needs_tree_write_lock
    def set_pending_merges(self, rev_list):
        parents = self.get_parent_ids()
        leftmost = parents[:1]
        new_parents = leftmost + rev_list
        self.set_parent_ids(new_parents)

    @needs_tree_write_lock
    def set_merge_modified(self, modified_hashes):
        def iter_stanzas():
            for file_id, hash in modified_hashes.iteritems():
                yield Stanza(file_id=file_id.decode('utf8'), hash=hash)
        self._put_rio('merge-hashes', iter_stanzas(), MERGE_MODIFIED_HEADER_1)

    def _sha_from_stat(self, path, stat_result):
        """Get a sha digest from the tree's stat cache.

        The default implementation assumes no stat cache is present.

        :param path: The path.
        :param stat_result: The stat result being looked up.
        """
        return None

    def _put_rio(self, filename, stanzas, header):
        self._must_be_locked()
        my_file = rio_file(stanzas, header)
        self._transport.put_file(filename, my_file,
            mode=self.bzrdir._get_file_mode())

    @needs_write_lock # because merge pulls data into the branch.
    def merge_from_branch(self, branch, to_revision=None, from_revision=None,
                          merge_type=None, force=False):
        """Merge from a branch into this working tree.

        :param branch: The branch to merge from.
        :param to_revision: If non-None, the merge will merge to to_revision,
            but not beyond it. to_revision does not need to be in the history
            of the branch when it is supplied. If None, to_revision defaults to
            branch.last_revision().
        """
        from bzrlib.merge import Merger, Merge3Merger
        pb = ui.ui_factory.nested_progress_bar()
        try:
            merger = Merger(self.branch, this_tree=self, pb=pb)
            merger.pp = ProgressPhase("Merge phase", 5, pb)
            merger.pp.next_phase()
            # check that there are no local alterations
            if not force and self.has_changes():
                raise errors.UncommittedChanges(self)
            if to_revision is None:
                to_revision = _mod_revision.ensure_null(branch.last_revision())
            merger.other_rev_id = to_revision
            if _mod_revision.is_null(merger.other_rev_id):
                raise errors.NoCommits(branch)
            self.branch.fetch(branch, last_revision=merger.other_rev_id)
            merger.other_basis = merger.other_rev_id
            merger.other_tree = self.branch.repository.revision_tree(
                merger.other_rev_id)
            merger.other_branch = branch
            merger.pp.next_phase()
            if from_revision is None:
                merger.find_base()
            else:
                merger.set_base_revision(from_revision, branch)
            if merger.base_rev_id == merger.other_rev_id:
                raise errors.PointlessMerge
            merger.backup_files = False
            if merge_type is None:
                merger.merge_type = Merge3Merger
            else:
                merger.merge_type = merge_type
            merger.set_interesting_files(None)
            merger.show_base = False
            merger.reprocess = False
            conflicts = merger.do_merge()
            merger.set_pending()
        finally:
            pb.finished()
        return conflicts

    @needs_read_lock
    def merge_modified(self):
        """Return a dictionary of files modified by a merge.

        The list is initialized by WorkingTree.set_merge_modified, which is
        typically called after we make some automatic updates to the tree
        because of a merge.

        This returns a map of file_id->sha1, containing only files which are
        still in the working inventory and have that text hash.
        """
        try:
            hashfile = self._transport.get('merge-hashes')
        except errors.NoSuchFile:
            return {}
        try:
            merge_hashes = {}
            try:
                if hashfile.next() != MERGE_MODIFIED_HEADER_1 + '\n':
                    raise errors.MergeModifiedFormatError()
            except StopIteration:
                raise errors.MergeModifiedFormatError()
            for s in RioReader(hashfile):
                # RioReader reads in Unicode, so convert file_ids back to utf8
                file_id = osutils.safe_file_id(s.get("file_id"), warn=False)
                if file_id not in self.inventory:
                    continue
                text_hash = s.get("hash")
                if text_hash == self.get_file_sha1(file_id):
                    merge_hashes[file_id] = text_hash
            return merge_hashes
        finally:
            hashfile.close()

    @needs_write_lock
    def mkdir(self, path, file_id=None):
        """See MutableTree.mkdir()."""
        if file_id is None:
            file_id = generate_ids.gen_file_id(os.path.basename(path))
        os.mkdir(self.abspath(path))
        self.add(path, file_id, 'directory')
        return file_id

    def get_symlink_target(self, file_id):
        abspath = self.id2abspath(file_id)
        target = osutils.readlink(abspath)
        return target

    @needs_write_lock
    def subsume(self, other_tree):
        def add_children(inventory, entry):
            for child_entry in entry.children.values():
                inventory._byid[child_entry.file_id] = child_entry
                if child_entry.kind == 'directory':
                    add_children(inventory, child_entry)
        if other_tree.get_root_id() == self.get_root_id():
            raise errors.BadSubsumeSource(self, other_tree,
                                          'Trees have the same root')
        try:
            other_tree_path = self.relpath(other_tree.basedir)
        except errors.PathNotChild:
            raise errors.BadSubsumeSource(self, other_tree,
                'Tree is not contained by the other')
        new_root_parent = self.path2id(osutils.dirname(other_tree_path))
        if new_root_parent is None:
            raise errors.BadSubsumeSource(self, other_tree,
                'Parent directory is not versioned.')
        # We need to ensure that the result of a fetch will have a
        # versionedfile for the other_tree root, and only fetching into
        # RepositoryKnit2 guarantees that.
        if not self.branch.repository.supports_rich_root():
            raise errors.SubsumeTargetNeedsUpgrade(other_tree)
        other_tree.lock_tree_write()
        try:
            new_parents = other_tree.get_parent_ids()
            other_root = other_tree.inventory.root
            other_root.parent_id = new_root_parent
            other_root.name = osutils.basename(other_tree_path)
            self.inventory.add(other_root)
            add_children(self.inventory, other_root)
            self._write_inventory(self.inventory)
            # normally we don't want to fetch whole repositories, but i think
            # here we really do want to consolidate the whole thing.
            for parent_id in other_tree.get_parent_ids():
                self.branch.fetch(other_tree.branch, parent_id)
                self.add_parent_tree_id(parent_id)
        finally:
            other_tree.unlock()
        other_tree.bzrdir.retire_bzrdir()

    def _setup_directory_is_tree_reference(self):
        if self._branch.repository._format.supports_tree_reference:
            self._directory_is_tree_reference = \
                self._directory_may_be_tree_reference
        else:
            self._directory_is_tree_reference = \
                self._directory_is_never_tree_reference

    def _directory_is_never_tree_reference(self, relpath):
        return False

    def _directory_may_be_tree_reference(self, relpath):
        # as a special case, if a directory contains control files then
        # it's a tree reference, except that the root of the tree is not
        return relpath and osutils.isdir(self.abspath(relpath) + u"/.bzr")
        # TODO: We could ask all the control formats whether they
        # recognize this directory, but at the moment there's no cheap api
        # to do that.  Since we probably can only nest bzr checkouts and
        # they always use this name it's ok for now.  -- mbp 20060306
        #
        # FIXME: There is an unhandled case here of a subdirectory
        # containing .bzr but not a branch; that will probably blow up
        # when you try to commit it.  It might happen if there is a
        # checkout in a subdirectory.  This can be avoided by not adding
        # it.  mbp 20070306

    @needs_tree_write_lock
    def extract(self, file_id, format=None):
        """Extract a subtree from this tree.

        A new branch will be created, relative to the path for this tree.
        """
        self.flush()
        def mkdirs(path):
            segments = osutils.splitpath(path)
            transport = self.branch.bzrdir.root_transport
            for name in segments:
                transport = transport.clone(name)
                transport.ensure_base()
            return transport

        sub_path = self.id2path(file_id)
        branch_transport = mkdirs(sub_path)
        if format is None:
            format = self.bzrdir.cloning_metadir()
        branch_transport.ensure_base()
        branch_bzrdir = format.initialize_on_transport(branch_transport)
        try:
            repo = branch_bzrdir.find_repository()
        except errors.NoRepositoryPresent:
            repo = branch_bzrdir.create_repository()
        if not repo.supports_rich_root():
            raise errors.RootNotRich()
        new_branch = branch_bzrdir.create_branch()
        new_branch.pull(self.branch)
        for parent_id in self.get_parent_ids():
            new_branch.fetch(self.branch, parent_id)
        tree_transport = self.bzrdir.root_transport.clone(sub_path)
        if tree_transport.base != branch_transport.base:
            tree_bzrdir = format.initialize_on_transport(tree_transport)
            branch.BranchReferenceFormat().initialize(tree_bzrdir, new_branch)
        else:
            tree_bzrdir = branch_bzrdir
        wt = tree_bzrdir.create_workingtree(_mod_revision.NULL_REVISION)
        wt.set_parent_ids(self.get_parent_ids())
        my_inv = self.inventory
        child_inv = inventory.Inventory(root_id=None)
        new_root = my_inv[file_id]
        my_inv.remove_recursive_id(file_id)
        new_root.parent_id = None
        child_inv.add(new_root)
        self._write_inventory(my_inv)
        wt._write_inventory(child_inv)
        return wt

    def _serialize(self, inventory, out_file):
        xml5.serializer_v5.write_inventory(self._inventory, out_file,
            working=True)

    def _deserialize(selt, in_file):
        return xml5.serializer_v5.read_inventory(in_file)

    def flush(self):
        """Write the in memory inventory to disk."""
        # TODO: Maybe this should only write on dirty ?
        if self._control_files._lock_mode != 'w':
            raise errors.NotWriteLocked(self)
        sio = StringIO()
        self._serialize(self._inventory, sio)
        sio.seek(0)
        self._transport.put_file('inventory', sio,
            mode=self.bzrdir._get_file_mode())
        self._inventory_is_modified = False

    def _kind(self, relpath):
        return osutils.file_kind(self.abspath(relpath))

    def list_files(self, include_root=False, from_dir=None, recursive=True):
        """List all files as (path, class, kind, id, entry).

        Lists, but does not descend into unversioned directories.
        This does not include files that have been deleted in this
        tree. Skips the control directory.

        :param include_root: if True, do not return an entry for the root
        :param from_dir: start from this directory or None for the root
        :param recursive: whether to recurse into subdirectories or not
        """
        # list_files is an iterator, so @needs_read_lock doesn't work properly
        # with it. So callers should be careful to always read_lock the tree.
        if not self.is_locked():
            raise errors.ObjectNotLocked(self)

        inv = self.inventory
        if from_dir is None and include_root is True:
            yield ('', 'V', 'directory', inv.root.file_id, inv.root)
        # Convert these into local objects to save lookup times
        pathjoin = osutils.pathjoin
        file_kind = self._kind

        # transport.base ends in a slash, we want the piece
        # between the last two slashes
        transport_base_dir = self.bzrdir.transport.base.rsplit('/', 2)[1]

        fk_entries = {'directory':TreeDirectory, 'file':TreeFile, 'symlink':TreeLink}

        # directory file_id, relative path, absolute path, reverse sorted children
        if from_dir is not None:
            from_dir_id = inv.path2id(from_dir)
            if from_dir_id is None:
                # Directory not versioned
                return
            from_dir_abspath = pathjoin(self.basedir, from_dir)
        else:
            from_dir_id = inv.root.file_id
            from_dir_abspath = self.basedir
        children = os.listdir(from_dir_abspath)
        children.sort()
        # jam 20060527 The kernel sized tree seems equivalent whether we
        # use a deque and popleft to keep them sorted, or if we use a plain
        # list and just reverse() them.
        children = collections.deque(children)
        stack = [(from_dir_id, u'', from_dir_abspath, children)]
        while stack:
            from_dir_id, from_dir_relpath, from_dir_abspath, children = stack[-1]

            while children:
                f = children.popleft()
                ## TODO: If we find a subdirectory with its own .bzr
                ## directory, then that is a separate tree and we
                ## should exclude it.

                # the bzrdir for this tree
                if transport_base_dir == f:
                    continue

                # we know that from_dir_relpath and from_dir_abspath never end in a slash
                # and 'f' doesn't begin with one, we can do a string op, rather
                # than the checks of pathjoin(), all relative paths will have an extra slash
                # at the beginning
                fp = from_dir_relpath + '/' + f

                # absolute path
                fap = from_dir_abspath + '/' + f

                f_ie = inv.get_child(from_dir_id, f)
                if f_ie:
                    c = 'V'
                elif self.is_ignored(fp[1:]):
                    c = 'I'
                else:
                    # we may not have found this file, because of a unicode issue
                    f_norm, can_access = osutils.normalized_filename(f)
                    if f == f_norm or not can_access:
                        # No change, so treat this file normally
                        c = '?'
                    else:
                        # this file can be accessed by a normalized path
                        # check again if it is versioned
                        # these lines are repeated here for performance
                        f = f_norm
                        fp = from_dir_relpath + '/' + f
                        fap = from_dir_abspath + '/' + f
                        f_ie = inv.get_child(from_dir_id, f)
                        if f_ie:
                            c = 'V'
                        elif self.is_ignored(fp[1:]):
                            c = 'I'
                        else:
                            c = '?'

                fk = file_kind(fap)

                # make a last minute entry
                if f_ie:
                    yield fp[1:], c, fk, f_ie.file_id, f_ie
                else:
                    try:
                        yield fp[1:], c, fk, None, fk_entries[fk]()
                    except KeyError:
                        yield fp[1:], c, fk, None, TreeEntry()
                    continue

                if fk != 'directory':
                    continue

                # But do this child first if recursing down
                if recursive:
                    new_children = os.listdir(fap)
                    new_children.sort()
                    new_children = collections.deque(new_children)
                    stack.append((f_ie.file_id, fp, fap, 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()

    @needs_tree_write_lock
    def move(self, from_paths, to_dir=None, after=False, **kwargs):
        """Rename files.

        to_dir must exist in the inventory.

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

        Note that to_dir is only the last component of the new name;
        this doesn't change the directory.

        For each entry in from_paths the move mode will be determined
        independently.

        The first mode moves the file in the filesystem and updates the
        inventory. The second mode only updates the inventory without
        touching the file on the filesystem. This is the new mode introduced
        in version 0.15.

        move uses the second mode if 'after == True' and the target is not
        versioned but present in the working tree.

        move uses the second mode if 'after == False' and the source is
        versioned but no longer in the working tree, and the target is not
        versioned but present in the working tree.

        move uses the first mode if 'after == False' and the source is
        versioned and present in the working tree, and the target is not
        versioned and not present in the working tree.

        Everything else results in an error.

        This returns a list of (from_path, to_path) pairs for each
        entry that is moved.
        """
        rename_entries = []
        rename_tuples = []

        # check for deprecated use of signature
        if to_dir is None:
            to_dir = kwargs.get('to_name', None)
            if to_dir is None:
                raise TypeError('You must supply a target directory')
            else:
                symbol_versioning.warn('The parameter to_name was deprecated'
                                       ' in version 0.13. Use to_dir instead',
                                       DeprecationWarning)

        # check destination directory
        if isinstance(from_paths, basestring):
            raise ValueError()
        inv = self.inventory
        to_abs = self.abspath(to_dir)
        if not isdir(to_abs):
            raise errors.BzrMoveFailedError('',to_dir,
                errors.NotADirectory(to_abs))
        if not self.has_filename(to_dir):
            raise errors.BzrMoveFailedError('',to_dir,
                errors.NotInWorkingDirectory(to_dir))
        to_dir_id = inv.path2id(to_dir)
        if to_dir_id is None:
            raise errors.BzrMoveFailedError('',to_dir,
                errors.NotVersionedError(path=str(to_dir)))

        to_dir_ie = inv[to_dir_id]
        if to_dir_ie.kind != 'directory':
            raise errors.BzrMoveFailedError('',to_dir,
                errors.NotADirectory(to_abs))

        # create rename entries and tuples
        for from_rel in from_paths:
            from_tail = splitpath(from_rel)[-1]
            from_id = inv.path2id(from_rel)
            if from_id is None:
                raise errors.BzrMoveFailedError(from_rel,to_dir,
                    errors.NotVersionedError(path=str(from_rel)))

            from_entry = inv[from_id]
            from_parent_id = from_entry.parent_id
            to_rel = pathjoin(to_dir, from_tail)
            rename_entry = WorkingTree._RenameEntry(from_rel=from_rel,
                                         from_id=from_id,
                                         from_tail=from_tail,
                                         from_parent_id=from_parent_id,
                                         to_rel=to_rel, to_tail=from_tail,
                                         to_parent_id=to_dir_id)
            rename_entries.append(rename_entry)
            rename_tuples.append((from_rel, to_rel))

        # determine which move mode to use. checks also for movability
        rename_entries = self._determine_mv_mode(rename_entries, after)

        original_modified = self._inventory_is_modified
        try:
            if len(from_paths):
                self._inventory_is_modified = True
            self._move(rename_entries)
        except:
            # restore the inventory on error
            self._inventory_is_modified = original_modified
            raise
        self._write_inventory(inv)
        return rename_tuples

    def _determine_mv_mode(self, rename_entries, after=False):
        """Determines for each from-to pair if both inventory and working tree
        or only the inventory has to be changed.

        Also does basic plausability tests.
        """
        inv = self.inventory

        for rename_entry in rename_entries:
            # store to local variables for easier reference
            from_rel = rename_entry.from_rel
            from_id = rename_entry.from_id
            to_rel = rename_entry.to_rel
            to_id = inv.path2id(to_rel)
            only_change_inv = False

            # check the inventory for source and destination
            if from_id is None:
                raise errors.BzrMoveFailedError(from_rel,to_rel,
                    errors.NotVersionedError(path=str(from_rel)))
            if to_id is not None:
                raise errors.BzrMoveFailedError(from_rel,to_rel,
                    errors.AlreadyVersionedError(path=str(to_rel)))

            # try to determine the mode for rename (only change inv or change
            # inv and file system)
            if after:
                if not self.has_filename(to_rel):
                    raise errors.BzrMoveFailedError(from_id,to_rel,
                        errors.NoSuchFile(path=str(to_rel),
                        extra="New file has not been created yet"))
                only_change_inv = True
            elif not self.has_filename(from_rel) and self.has_filename(to_rel):
                only_change_inv = True
            elif self.has_filename(from_rel) and not self.has_filename(to_rel):
                only_change_inv = False
            elif (not self.case_sensitive
                  and from_rel.lower() == to_rel.lower()
                  and self.has_filename(from_rel)):
                only_change_inv = False
            else:
                # something is wrong, so lets determine what exactly
                if not self.has_filename(from_rel) and \
                   not self.has_filename(to_rel):
                    raise errors.BzrRenameFailedError(from_rel,to_rel,
                        errors.PathsDoNotExist(paths=(str(from_rel),
                        str(to_rel))))
                else:
                    raise errors.RenameFailedFilesExist(from_rel, to_rel)
            rename_entry.only_change_inv = only_change_inv
        return rename_entries

    def _move(self, rename_entries):
        """Moves a list of files.

        Depending on the value of the flag 'only_change_inv', the
        file will be moved on the file system or not.
        """
        inv = self.inventory
        moved = []

        for entry in rename_entries:
            try:
                self._move_entry(entry)
            except:
                self._rollback_move(moved)
                raise
            moved.append(entry)

    def _rollback_move(self, moved):
        """Try to rollback a previous move in case of an filesystem error."""
        inv = self.inventory
        for entry in moved:
            try:
                self._move_entry(WorkingTree._RenameEntry(
                    entry.to_rel, entry.from_id,
                    entry.to_tail, entry.to_parent_id, entry.from_rel,
                    entry.from_tail, entry.from_parent_id,
                    entry.only_change_inv))
            except errors.BzrMoveFailedError, e:
                raise errors.BzrMoveFailedError( '', '', "Rollback failed."
                        " The working tree is in an inconsistent state."
                        " Please consider doing a 'bzr revert'."
                        " Error message is: %s" % e)

    def _move_entry(self, entry):
        inv = self.inventory
        from_rel_abs = self.abspath(entry.from_rel)
        to_rel_abs = self.abspath(entry.to_rel)
        if from_rel_abs == to_rel_abs:
            raise errors.BzrMoveFailedError(entry.from_rel, entry.to_rel,
                "Source and target are identical.")

        if not entry.only_change_inv:
            try:
                osutils.rename(from_rel_abs, to_rel_abs)
            except OSError, e:
                raise errors.BzrMoveFailedError(entry.from_rel,
                    entry.to_rel, e[1])
        inv.rename(entry.from_id, entry.to_parent_id, entry.to_tail)

    @needs_tree_write_lock
    def rename_one(self, from_rel, to_rel, after=False):
        """Rename one file.

        This can change the directory or the filename or both.

        rename_one has several 'modes' to work. First, it can rename a physical
        file and change the file_id. That is the normal mode. Second, it can
        only change the file_id without touching any physical file. This is
        the new mode introduced in version 0.15.

        rename_one uses the second mode if 'after == True' and 'to_rel' is not
        versioned but present in the working tree.

        rename_one uses the second mode if 'after == False' and 'from_rel' is
        versioned but no longer in the working tree, and 'to_rel' is not
        versioned but present in the working tree.

        rename_one uses the first mode if 'after == False' and 'from_rel' is
        versioned and present in the working tree, and 'to_rel' is not
        versioned and not present in the working tree.

        Everything else results in an error.
        """
        inv = self.inventory
        rename_entries = []

        # create rename entries and tuples
        from_tail = splitpath(from_rel)[-1]
        from_id = inv.path2id(from_rel)
        if from_id is None:
            # if file is missing in the inventory maybe it's in the basis_tree
            basis_tree = self.branch.basis_tree()
            from_id = basis_tree.path2id(from_rel)
            if from_id is None:
                raise errors.BzrRenameFailedError(from_rel,to_rel,
                    errors.NotVersionedError(path=str(from_rel)))
            # put entry back in the inventory so we can rename it
            from_entry = basis_tree.inventory[from_id].copy()
            inv.add(from_entry)
        else:
            from_entry = inv[from_id]
        from_parent_id = from_entry.parent_id
        to_dir, to_tail = os.path.split(to_rel)
        to_dir_id = inv.path2id(to_dir)
        rename_entry = WorkingTree._RenameEntry(from_rel=from_rel,
                                     from_id=from_id,
                                     from_tail=from_tail,
                                     from_parent_id=from_parent_id,
                                     to_rel=to_rel, to_tail=to_tail,
                                     to_parent_id=to_dir_id)
        rename_entries.append(rename_entry)

        # determine which move mode to use. checks also for movability
        rename_entries = self._determine_mv_mode(rename_entries, after)

        # check if the target changed directory and if the target directory is
        # versioned
        if to_dir_id is None:
            raise errors.BzrMoveFailedError(from_rel,to_rel,
                errors.NotVersionedError(path=str(to_dir)))

        # all checks done. now we can continue with our actual work
        mutter('rename_one:\n'
               '  from_id   {%s}\n'
               '  from_rel: %r\n'
               '  to_rel:   %r\n'
               '  to_dir    %r\n'
               '  to_dir_id {%s}\n',
               from_id, from_rel, to_rel, to_dir, to_dir_id)

        self._move(rename_entries)
        self._write_inventory(inv)

    class _RenameEntry(object):
        def __init__(self, from_rel, from_id, from_tail, from_parent_id,
                     to_rel, to_tail, to_parent_id, only_change_inv=False):
            self.from_rel = from_rel
            self.from_id = from_id
            self.from_tail = from_tail
            self.from_parent_id = from_parent_id
            self.to_rel = to_rel
            self.to_tail = to_tail
            self.to_parent_id = to_parent_id
            self.only_change_inv = only_change_inv

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

        These are files in the working directory that are not versioned or
        control files or ignored.
        """
        # force the extras method to be fully executed before returning, to
        # prevent race conditions with the lock
        return iter(
            [subp for subp in self.extras() if not self.is_ignored(subp)])

    @needs_tree_write_lock
    def unversion(self, file_ids):
        """Remove the file ids in file_ids from the current versioned set.

        When a file_id is unversioned, all of its children are automatically
        unversioned.

        :param file_ids: The file ids to stop versioning.
        :raises: NoSuchId if any fileid is not currently versioned.
        """
        for file_id in file_ids:
            if file_id not in self._inventory:
                raise errors.NoSuchId(self, file_id)
        for file_id in file_ids:
            if self._inventory.has_id(file_id):
                self._inventory.remove_recursive_id(file_id)
        if len(file_ids):
            # in the future this should just set a dirty bit to wait for the
            # final unlock. However, until all methods of workingtree start
            # with the current in -memory inventory rather than triggering
            # a read, it is more complex - we need to teach read_inventory
            # to know when to read, and when to not read first... and possibly
            # to save first when the in memory one may be corrupted.
            # so for now, we just only write it if it is indeed dirty.
            # - RBC 20060907
            self._write_inventory(self._inventory)

    def _iter_conflicts(self):
        conflicted = set()
        for info in self.list_files():
            path = info[0]
            stem = get_conflicted_stem(path)
            if stem is None:
                continue
            if stem not in conflicted:
                conflicted.add(stem)
                yield stem

    @needs_write_lock
    def pull(self, source, overwrite=False, stop_revision=None,
             change_reporter=None, possible_transports=None, local=False):
        top_pb = ui.ui_factory.nested_progress_bar()
        source.lock_read()
        try:
            pp = ProgressPhase("Pull phase", 2, top_pb)
            pp.next_phase()
            old_revision_info = self.branch.last_revision_info()
            basis_tree = self.basis_tree()
            count = self.branch.pull(source, overwrite, stop_revision,
                                     possible_transports=possible_transports,
                                     local=local)
            new_revision_info = self.branch.last_revision_info()
            if new_revision_info != old_revision_info:
                pp.next_phase()
                repository = self.branch.repository
                pb = ui.ui_factory.nested_progress_bar()
                basis_tree.lock_read()
                try:
                    new_basis_tree = self.branch.basis_tree()
                    merge.merge_inner(
                                self.branch,
                                new_basis_tree,
                                basis_tree,
                                this_tree=self,
                                pb=pb,
                                change_reporter=change_reporter)
                    if (basis_tree.inventory.root is None and
                        new_basis_tree.inventory.root is not None):
                        self.set_root_id(new_basis_tree.get_root_id())
                finally:
                    pb.finished()
                    basis_tree.unlock()
                # TODO - dedup parents list with things merged by pull ?
                # reuse the revisiontree we merged against to set the new
                # tree data.
                parent_trees = [(self.branch.last_revision(), new_basis_tree)]
                # we have to pull the merge trees out again, because
                # merge_inner has set the ids. - this corner is not yet
                # layered well enough to prevent double handling.
                # XXX TODO: Fix the double handling: telling the tree about
                # the already known parent data is wasteful.
                merges = self.get_parent_ids()[1:]
                parent_trees.extend([
                    (parent, repository.revision_tree(parent)) for
                     parent in merges])
                self.set_parent_trees(parent_trees)
            return count
        finally:
            source.unlock()
            top_pb.finished()

    @needs_write_lock
    def put_file_bytes_non_atomic(self, file_id, bytes):
        """See MutableTree.put_file_bytes_non_atomic."""
        stream = file(self.id2abspath(file_id), 'wb')
        try:
            stream.write(bytes)
        finally:
            stream.close()
        # TODO: update the hashcache here ?

    def extras(self):
        """Yield all unversioned files in this WorkingTree.

        If there are any unversioned directories then only the directory is
        returned, not all its children.  But if there are unversioned files
        under a versioned subdirectory, they are returned.

        Currently returned depth-first, sorted by name within directories.
        This is the same order used by 'osutils.walkdirs'.
        """
        ## TODO: Work from given directory downwards
        for path, dir_entry in self.inventory.directories():
            # mutter("search for unknowns in %r", path)
            dirabs = self.abspath(path)
            if not isdir(dirabs):
                # e.g. directory deleted
                continue

            fl = []
            for subf in os.listdir(dirabs):
                if self.bzrdir.is_control_filename(subf):
                    continue
                if subf not in dir_entry.children:
                    try:
                        (subf_norm,
                         can_access) = osutils.normalized_filename(subf)
                    except UnicodeDecodeError:
                        path_os_enc = path.encode(osutils._fs_enc)
                        relpath = path_os_enc + '/' + subf
                        raise errors.BadFilenameEncoding(relpath,
                                                         osutils._fs_enc)
                    if subf_norm != subf and can_access:
                        if subf_norm not in dir_entry.children:
                            fl.append(subf_norm)
                    else:
                        fl.append(subf)

            fl.sort()
            for subf in fl:
                subp = pathjoin(path, subf)
                yield subp

    def ignored_files(self):
        """Yield list of PATH, IGNORE_PATTERN"""
        for subp in self.extras():
            pat = self.is_ignored(subp)
            if pat is not None:
                yield subp, pat

    def get_ignore_list(self):
        """Return list of ignore patterns.

        Cached in the Tree object after the first call.
        """
        ignoreset = getattr(self, '_ignoreset', None)
        if ignoreset is not None:
            return ignoreset

        ignore_globs = set()
        ignore_globs.update(ignores.get_runtime_ignores())
        ignore_globs.update(ignores.get_user_ignores())
        if self.has_filename(bzrlib.IGNORE_FILENAME):
            f = self.get_file_byname(bzrlib.IGNORE_FILENAME)
            try:
                ignore_globs.update(ignores.parse_ignore_file(f))
            finally:
                f.close()
        self._ignoreset = ignore_globs
        return ignore_globs

    def _flush_ignore_list_cache(self):
        """Resets the cached ignore list to force a cache rebuild."""
        self._ignoreset = None
        self._ignoreglobster = None

    def is_ignored(self, filename):
        r"""Check whether the filename matches an ignore pattern.

        Patterns containing '/' or '\' need to match the whole path;
        others match against only the last component.

        If the file is ignored, returns the pattern which caused it to
        be ignored, otherwise None.  So this can simply be used as a
        boolean if desired."""
        if getattr(self, '_ignoreglobster', None) is None:
            self._ignoreglobster = globbing.Globster(self.get_ignore_list())
        return self._ignoreglobster.match(filename)

    def kind(self, file_id):
        return file_kind(self.id2abspath(file_id))

    def stored_kind(self, file_id):
        """See Tree.stored_kind"""
        return self.inventory[file_id].kind

    def _comparison_data(self, entry, path):
        abspath = self.abspath(path)
        try:
            stat_value = os.lstat(abspath)
        except OSError, e:
            if getattr(e, 'errno', None) == errno.ENOENT:
                stat_value = None
                kind = None
                executable = False
            else:
                raise
        else:
            mode = stat_value.st_mode
            kind = osutils.file_kind_from_stat_mode(mode)
            if not supports_executable():
                executable = entry is not None and entry.executable
            else:
                executable = bool(stat.S_ISREG(mode) and stat.S_IEXEC & mode)
        return kind, executable, stat_value

    def _file_size(self, entry, stat_value):
        return stat_value.st_size

    def last_revision(self):
        """Return the last revision of the branch for this tree.

        This format tree does not support a separate marker for last-revision
        compared to the branch.

        See MutableTree.last_revision
        """
        return self._last_revision()

    @needs_read_lock
    def _last_revision(self):
        """helper for get_parent_ids."""
        return _mod_revision.ensure_null(self.branch.last_revision())

    def is_locked(self):
        return self._control_files.is_locked()

    def _must_be_locked(self):
        if not self.is_locked():
            raise errors.ObjectNotLocked(self)

    def lock_read(self):
        """See Branch.lock_read, and WorkingTree.unlock."""
        if not self.is_locked():
            self._reset_data()
        self.branch.lock_read()
        try:
            return self._control_files.lock_read()
        except:
            self.branch.unlock()
            raise

    def lock_tree_write(self):
        """See MutableTree.lock_tree_write, and WorkingTree.unlock."""
        if not self.is_locked():
            self._reset_data()
        self.branch.lock_read()
        try:
            return self._control_files.lock_write()
        except:
            self.branch.unlock()
            raise

    def lock_write(self):
        """See MutableTree.lock_write, and WorkingTree.unlock."""
        if not self.is_locked():
            self._reset_data()
        self.branch.lock_write()
        try:
            return self._control_files.lock_write()
        except:
            self.branch.unlock()
            raise

    def get_physical_lock_status(self):
        return self._control_files.get_physical_lock_status()

    def _basis_inventory_name(self):
        return 'basis-inventory-cache'

    def _reset_data(self):
        """Reset transient data that cannot be revalidated."""
        self._inventory_is_modified = False
        f = self._transport.get('inventory')
        try:
            result = self._deserialize(f)
        finally:
            f.close()
        self._set_inventory(result, dirty=False)

    @needs_tree_write_lock
    def set_last_revision(self, new_revision):
        """Change the last revision in the working tree."""
        if self._change_last_revision(new_revision):
            self._cache_basis_inventory(new_revision)

    def _change_last_revision(self, new_revision):
        """Template method part of set_last_revision to perform the change.

        This is used to allow WorkingTree3 instances to not affect branch
        when their last revision is set.
        """
        if _mod_revision.is_null(new_revision):
            self.branch.set_revision_history([])
            return False
        try:
            self.branch.generate_revision_history(new_revision)
        except errors.NoSuchRevision:
            # not present in the repo - dont try to set it deeper than the tip
            self.branch.set_revision_history([new_revision])
        return True

    def _write_basis_inventory(self, xml):
        """Write the basis inventory XML to the basis-inventory file"""
        path = self._basis_inventory_name()
        sio = StringIO(xml)
        self._transport.put_file(path, sio,
            mode=self.bzrdir._get_file_mode())

    def _create_basis_xml_from_inventory(self, revision_id, inventory):
        """Create the text that will be saved in basis-inventory"""
        inventory.revision_id = revision_id
        return xml7.serializer_v7.write_inventory_to_string(inventory)

    def _cache_basis_inventory(self, new_revision):
        """Cache new_revision as the basis inventory."""
        # TODO: this should allow the ready-to-use inventory to be passed in,
        # as commit already has that ready-to-use [while the format is the
        # same, that is].
        try:
            # this double handles the inventory - unpack and repack -
            # but is easier to understand. We can/should put a conditional
            # in here based on whether the inventory is in the latest format
            # - perhaps we should repack all inventories on a repository
            # upgrade ?
            # the fast path is to copy the raw xml from the repository. If the
            # xml contains 'revision_id="', then we assume the right
            # revision_id is set. We must check for this full string, because a
            # root node id can legitimately look like 'revision_id' but cannot
            # contain a '"'.
            xml = self.branch.repository.get_inventory_xml(new_revision)
            firstline = xml.split('\n', 1)[0]
            if (not 'revision_id="' in firstline or
                'format="7"' not in firstline):
                inv = self.branch.repository._serializer.read_inventory_from_string(
                    xml, new_revision)
                xml = self._create_basis_xml_from_inventory(new_revision, inv)
            self._write_basis_inventory(xml)
        except (errors.NoSuchRevision, errors.RevisionNotPresent):
            pass

    def read_basis_inventory(self):
        """Read the cached basis inventory."""
        path = self._basis_inventory_name()
        return self._transport.get_bytes(path)

    @needs_read_lock
    def read_working_inventory(self):
        """Read the working inventory.

        :raises errors.InventoryModified: read_working_inventory will fail
            when the current in memory inventory has been modified.
        """
        # conceptually this should be an implementation detail of the tree.
        # XXX: Deprecate this.
        # ElementTree does its own conversion from UTF-8, so open in
        # binary.
        if self._inventory_is_modified:
            raise errors.InventoryModified(self)
        f = self._transport.get('inventory')
        try:
            result = self._deserialize(f)
        finally:
            f.close()
        self._set_inventory(result, dirty=False)
        return result

    @needs_tree_write_lock
    def remove(self, files, verbose=False, to_file=None, keep_files=True,
        force=False):
        """Remove nominated files from the working inventory.

        :files: File paths relative to the basedir.
        :keep_files: If true, the files will also be kept.
        :force: Delete files and directories, even if they are changed and
            even if the directories are not empty.
        """
        if isinstance(files, basestring):
            files = [files]

        inv_delta = []

        new_files=set()
        unknown_nested_files=set()
        if to_file is None:
            to_file = sys.stdout

        def recurse_directory_to_add_files(directory):
            # Recurse directory and add all files
            # so we can check if they have changed.
            for parent_info, file_infos in\
                self.walkdirs(directory):
                for relpath, basename, kind, lstat, fileid, kind in file_infos:
                    # Is it versioned or ignored?
                    if self.path2id(relpath) or self.is_ignored(relpath):
                        # Add nested content for deletion.
                        new_files.add(relpath)
                    else:
                        # Files which are not versioned and not ignored
                        # should be treated as unknown.
                        unknown_nested_files.add((relpath, None, kind))

        for filename in files:
            # Get file name into canonical form.
            abspath = self.abspath(filename)
            filename = self.relpath(abspath)
            if len(filename) > 0:
                new_files.add(filename)
                recurse_directory_to_add_files(filename)

        files = list(new_files)

        if len(files) == 0:
            return # nothing to do

        # Sort needed to first handle directory content before the directory
        files.sort(reverse=True)

        # Bail out if we are going to delete files we shouldn't
        if not keep_files and not force:
            has_changed_files = len(unknown_nested_files) > 0
            if not has_changed_files:
                for (file_id, path, content_change, versioned, parent_id, name,
                     kind, executable) in self.iter_changes(self.basis_tree(),
                         include_unchanged=True, require_versioned=False,
                         want_unversioned=True, specific_files=files):
                    if versioned == (False, False):
                        # The record is unknown ...
                        if not self.is_ignored(path[1]):
                            # ... but not ignored
                            has_changed_files = True
                            break
                    elif content_change and (kind[1] is not None):
                        # Versioned and changed, but not deleted
                        has_changed_files = True
                        break

            if has_changed_files:
                # Make delta show ALL applicable changes in error message.
                tree_delta = self.changes_from(self.basis_tree(),
                    require_versioned=False, want_unversioned=True,
                    specific_files=files)
                for unknown_file in unknown_nested_files:
                    if unknown_file not in tree_delta.unversioned:
                        tree_delta.unversioned.extend((unknown_file,))
                raise errors.BzrRemoveChangedFilesError(tree_delta)

        # Build inv_delta and delete files where applicable,
        # do this before any modifications to inventory.
        for f in files:
            fid = self.path2id(f)
            message = None
            if not fid:
                message = "%s is not versioned." % (f,)
            else:
                if verbose:
                    # having removed it, it must be either ignored or unknown
                    if self.is_ignored(f):
                        new_status = 'I'
                    else:
                        new_status = '?'
                    # XXX: Really should be a more abstract reporter interface
                    kind_ch = osutils.kind_marker(self.kind(fid))
                    to_file.write(new_status + '       ' + f + kind_ch + '\n')
                # Unversion file
                inv_delta.append((f, None, fid, None))
                message = "removed %s" % (f,)

            if not keep_files:
                abs_path = self.abspath(f)
                if osutils.lexists(abs_path):
                    if (osutils.isdir(abs_path) and
                        len(os.listdir(abs_path)) > 0):
                        if force:
                            osutils.rmtree(abs_path)
                        else:
                            message = "%s is not an empty directory "\
                                "and won't be deleted." % (f,)
                    else:
                        osutils.delete_any(abs_path)
                        message = "deleted %s" % (f,)
                elif message is not None:
                    # Only care if we haven't done anything yet.
                    message = "%s does not exist." % (f,)

            # Print only one message (if any) per file.
            if message is not None:
                note(message)
        self.apply_inventory_delta(inv_delta)

    @needs_tree_write_lock
    def revert(self, filenames=None, old_tree=None, backups=True,
               pb=DummyProgress(), report_changes=False):
        from bzrlib.conflicts import resolve
        if filenames == []:
            filenames = None
            symbol_versioning.warn('Using [] to revert all files is deprecated'
                ' as of bzr 0.91.  Please use None (the default) instead.',
                DeprecationWarning, stacklevel=2)
        if old_tree is None:
            basis_tree = self.basis_tree()
            basis_tree.lock_read()
            old_tree = basis_tree
        else:
            basis_tree = None
        try:
            conflicts = transform.revert(self, old_tree, filenames, backups, pb,
                                         report_changes)
            if filenames is None and len(self.get_parent_ids()) > 1:
                parent_trees = []
                last_revision = self.last_revision()
                if last_revision != _mod_revision.NULL_REVISION:
                    if basis_tree is None:
                        basis_tree = self.basis_tree()
                        basis_tree.lock_read()
                    parent_trees.append((last_revision, basis_tree))
                self.set_parent_trees(parent_trees)
                resolve(self)
            else:
                resolve(self, filenames, ignore_misses=True, recursive=True)
        finally:
            if basis_tree is not None:
                basis_tree.unlock()
        return conflicts

    def revision_tree(self, revision_id):
        """See Tree.revision_tree.

        WorkingTree can supply revision_trees for the basis revision only
        because there is only one cached inventory in the bzr directory.
        """
        if revision_id == self.last_revision():
            try:
                xml = self.read_basis_inventory()
            except errors.NoSuchFile:
                pass
            else:
                try:
                    inv = xml7.serializer_v7.read_inventory_from_string(xml)
                    # dont use the repository revision_tree api because we want
                    # to supply the inventory.
                    if inv.revision_id == revision_id:
                        return revisiontree.RevisionTree(self.branch.repository,
                            inv, revision_id)
                except errors.BadInventoryFormat:
                    pass
        # raise if there was no inventory, or if we read the wrong inventory.
        raise errors.NoSuchRevisionInTree(self, revision_id)

    # XXX: This method should be deprecated in favour of taking in a proper
    # new Inventory object.
    @needs_tree_write_lock
    def set_inventory(self, new_inventory_list):
        from bzrlib.inventory import (Inventory,
                                      InventoryDirectory,
                                      InventoryFile,
                                      InventoryLink)
        inv = Inventory(self.get_root_id())
        for path, file_id, parent, kind in new_inventory_list:
            name = os.path.basename(path)
            if name == "":
                continue
            # fixme, there should be a factory function inv,add_??
            if kind == 'directory':
                inv.add(InventoryDirectory(file_id, name, parent))
            elif kind == 'file':
                inv.add(InventoryFile(file_id, name, parent))
            elif kind == 'symlink':
                inv.add(InventoryLink(file_id, name, parent))
            else:
                raise errors.BzrError("unknown kind %r" % kind)
        self._write_inventory(inv)

    @needs_tree_write_lock
    def set_root_id(self, file_id):
        """Set the root id for this tree."""
        # for compatability
        if file_id is None:
            raise ValueError(
                'WorkingTree.set_root_id with fileid=None')
        file_id = osutils.safe_file_id(file_id)
        self._set_root_id(file_id)

    def _set_root_id(self, file_id):
        """Set the root id for this tree, in a format specific manner.

        :param file_id: The file id to assign to the root. It must not be
            present in the current inventory or an error will occur. It must
            not be None, but rather a valid file id.
        """
        inv = self._inventory
        orig_root_id = inv.root.file_id
        # TODO: it might be nice to exit early if there was nothing
        # to do, saving us from trigger a sync on unlock.
        self._inventory_is_modified = True
        # we preserve the root inventory entry object, but
        # unlinkit from the byid index
        del inv._byid[inv.root.file_id]
        inv.root.file_id = file_id
        # and link it into the index with the new changed id.
        inv._byid[inv.root.file_id] = inv.root
        # and finally update all children to reference the new id.
        # XXX: this should be safe to just look at the root.children
        # list, not the WHOLE INVENTORY.
        for fid in inv:
            entry = inv[fid]
            if entry.parent_id == orig_root_id:
                entry.parent_id = inv.root.file_id

    def unlock(self):
        """See Branch.unlock.

        WorkingTree locking just uses the Branch locking facilities.
        This is current because all working trees have an embedded branch
        within them. IF in the future, we were to make branch data shareable
        between multiple working trees, i.e. via shared storage, then we
        would probably want to lock both the local tree, and the branch.
        """
        raise NotImplementedError(self.unlock)

    def update(self, change_reporter=None, possible_transports=None):
        """Update a working tree along its branch.

        This will update the branch if its bound too, which means we have
        multiple trees involved:

        - The new basis tree of the master.
        - The old basis tree of the branch.
        - The old basis tree of the working tree.
        - The current working tree state.

        Pathologically, all three may be different, and non-ancestors of each
        other.  Conceptually we want to:

        - Preserve the wt.basis->wt.state changes
        - Transform the wt.basis to the new master basis.
        - Apply a merge of the old branch basis to get any 'local' changes from
          it into the tree.
        - Restore the wt.basis->wt.state changes.

        There isn't a single operation at the moment to do that, so we:
        - Merge current state -> basis tree of the master w.r.t. the old tree
          basis.
        - Do a 'normal' merge of the old branch basis if it is relevant.
        """
        if self.branch.get_bound_location() is not None:
            self.lock_write()
            update_branch = True
        else:
            self.lock_tree_write()
            update_branch = False
        try:
            if update_branch:
                old_tip = self.branch.update(possible_transports)
            else:
                old_tip = None
            return self._update_tree(old_tip, change_reporter)
        finally:
            self.unlock()

    @needs_tree_write_lock
    def _update_tree(self, old_tip=None, change_reporter=None):
        """Update a tree to the master branch.

        :param old_tip: if supplied, the previous tip revision the branch,
            before it was changed to the master branch's tip.
        """
        # here if old_tip is not None, it is the old tip of the branch before
        # it was updated from the master branch. This should become a pending
        # merge in the working tree to preserve the user existing work.  we
        # cant set that until we update the working trees last revision to be
        # one from the new branch, because it will just get absorbed by the
        # parent de-duplication logic.
        #
        # We MUST save it even if an error occurs, because otherwise the users
        # local work is unreferenced and will appear to have been lost.
        #
        result = 0
        try:
            last_rev = self.get_parent_ids()[0]
        except IndexError:
            last_rev = _mod_revision.NULL_REVISION
        if last_rev != _mod_revision.ensure_null(self.branch.last_revision()):
            # merge tree state up to new branch tip.
            basis = self.basis_tree()
            basis.lock_read()
            try:
                to_tree = self.branch.basis_tree()
                if basis.inventory.root is None:
                    self.set_root_id(to_tree.get_root_id())
                    self.flush()
                result += merge.merge_inner(
                                      self.branch,
                                      to_tree,
                                      basis,
                                      this_tree=self,
                                      change_reporter=change_reporter)
            finally:
                basis.unlock()
            # TODO - dedup parents list with things merged by pull ?
            # reuse the tree we've updated to to set the basis:
            parent_trees = [(self.branch.last_revision(), to_tree)]
            merges = self.get_parent_ids()[1:]
            # Ideally we ask the tree for the trees here, that way the working
            # tree can decide whether to give us the entire tree or give us a
            # lazy initialised tree. dirstate for instance will have the trees
            # in ram already, whereas a last-revision + basis-inventory tree
            # will not, but also does not need them when setting parents.
            for parent in merges:
                parent_trees.append(
                    (parent, self.branch.repository.revision_tree(parent)))
            if (old_tip is not None and not _mod_revision.is_null(old_tip)):
                parent_trees.append(
                    (old_tip, self.branch.repository.revision_tree(old_tip)))
            self.set_parent_trees(parent_trees)
            last_rev = parent_trees[0][0]
        else:
            # the working tree had the same last-revision as the master
            # branch did. We may still have pivot local work from the local
            # branch into old_tip:
            if (old_tip is not None and not _mod_revision.is_null(old_tip)):
                self.add_parent_tree_id(old_tip)
        if (old_tip is not None and not _mod_revision.is_null(old_tip)
            and old_tip != last_rev):
            # our last revision was not the prior branch last revision
            # and we have converted that last revision to a pending merge.
            # base is somewhere between the branch tip now
            # and the now pending merge

            # Since we just modified the working tree and inventory, flush out
            # the current state, before we modify it again.
            # TODO: jam 20070214 WorkingTree3 doesn't require this, dirstate
            #       requires it only because TreeTransform directly munges the
            #       inventory and calls tree._write_inventory(). Ultimately we
            #       should be able to remove this extra flush.
            self.flush()
            graph = self.branch.repository.get_graph()
            base_rev_id = graph.find_unique_lca(self.branch.last_revision(),
                                                old_tip)
            base_tree = self.branch.repository.revision_tree(base_rev_id)
            other_tree = self.branch.repository.revision_tree(old_tip)
            result += merge.merge_inner(
                                  self.branch,
                                  other_tree,
                                  base_tree,
                                  this_tree=self,
                                  change_reporter=change_reporter)
        return result

    def _write_hashcache_if_dirty(self):
        """Write out the hashcache if it is dirty."""
        if self._hashcache.needs_write:
            try:
                self._hashcache.write()
            except OSError, e:
                if e.errno not in (errno.EPERM, errno.EACCES):
                    raise
                # TODO: jam 20061219 Should this be a warning? A single line
                #       warning might be sufficient to let the user know what
                #       is going on.
                mutter('Could not write hashcache for %s\nError: %s',
                       self._hashcache.cache_file_name(), e)

    @needs_tree_write_lock
    def _write_inventory(self, inv):
        """Write inventory as the current inventory."""
        self._set_inventory(inv, dirty=True)
        self.flush()

    def set_conflicts(self, arg):
        raise errors.UnsupportedOperation(self.set_conflicts, self)

    def add_conflicts(self, arg):
        raise errors.UnsupportedOperation(self.add_conflicts, self)

    @needs_read_lock
    def conflicts(self):
        conflicts = _mod_conflicts.ConflictList()
        for conflicted in self._iter_conflicts():
            text = True
            try:
                if file_kind(self.abspath(conflicted)) != "file":
                    text = False
            except errors.NoSuchFile:
                text = False
            if text is True:
                for suffix in ('.THIS', '.OTHER'):
                    try:
                        kind = file_kind(self.abspath(conflicted+suffix))
                        if kind != "file":
                            text = False
                    except errors.NoSuchFile:
                        text = False
                    if text == False:
                        break
            ctype = {True: 'text conflict', False: 'contents conflict'}[text]
            conflicts.append(_mod_conflicts.Conflict.factory(ctype,
                             path=conflicted,
                             file_id=self.path2id(conflicted)))
        return conflicts

    def walkdirs(self, prefix=""):
        """Walk the directories of this tree.

        returns a generator which yields items in the form:
                ((curren_directory_path, fileid),
                 [(file1_path, file1_name, file1_kind, (lstat), file1_id,
                   file1_kind), ... ])

        This API returns a generator, which is only valid during the current
        tree transaction - within a single lock_read or lock_write duration.

        If the tree is not locked, it may cause an error to be raised,
        depending on the tree implementation.
        """
        disk_top = self.abspath(prefix)
        if disk_top.endswith('/'):
            disk_top = disk_top[:-1]
        top_strip_len = len(disk_top) + 1
        inventory_iterator = self._walkdirs(prefix)
        disk_iterator = osutils.walkdirs(disk_top, prefix)
        try:
            current_disk = disk_iterator.next()
            disk_finished = False
        except OSError, e:
            if not (e.errno == errno.ENOENT or
                (sys.platform == 'win32' and e.errno == ERROR_PATH_NOT_FOUND)):
                raise
            current_disk = None
            disk_finished = True
        try:
            current_inv = inventory_iterator.next()
            inv_finished = False
        except StopIteration:
            current_inv = None
            inv_finished = True
        while not inv_finished or not disk_finished:
            if current_disk:
                ((cur_disk_dir_relpath, cur_disk_dir_path_from_top),
                    cur_disk_dir_content) = current_disk
            else:
                ((cur_disk_dir_relpath, cur_disk_dir_path_from_top),
                    cur_disk_dir_content) = ((None, None), None)
            if not disk_finished:
                # strip out .bzr dirs
                if (cur_disk_dir_path_from_top[top_strip_len:] == '' and
                    len(cur_disk_dir_content) > 0):
                    # osutils.walkdirs can be made nicer -
                    # yield the path-from-prefix rather than the pathjoined
                    # value.
                    bzrdir_loc = bisect_left(cur_disk_dir_content,
                        ('.bzr', '.bzr'))
                    if (bzrdir_loc < len(cur_disk_dir_content)
                        and self.bzrdir.is_control_filename(
                            cur_disk_dir_content[bzrdir_loc][0])):
                        # we dont yield the contents of, or, .bzr itself.
                        del cur_disk_dir_content[bzrdir_loc]
            if inv_finished:
                # everything is unknown
                direction = 1
            elif disk_finished:
                # everything is missing
                direction = -1
            else:
                direction = cmp(current_inv[0][0], cur_disk_dir_relpath)
            if direction > 0:
                # disk is before inventory - unknown
                dirblock = [(relpath, basename, kind, stat, None, None) for
                    relpath, basename, kind, stat, top_path in
                    cur_disk_dir_content]
                yield (cur_disk_dir_relpath, None), dirblock
                try:
                    current_disk = disk_iterator.next()
                except StopIteration:
                    disk_finished = True
            elif direction < 0:
                # inventory is before disk - missing.
                dirblock = [(relpath, basename, 'unknown', None, fileid, kind)
                    for relpath, basename, dkind, stat, fileid, kind in
                    current_inv[1]]
                yield (current_inv[0][0], current_inv[0][1]), dirblock
                try:
                    current_inv = inventory_iterator.next()
                except StopIteration:
                    inv_finished = True
            else:
                # versioned present directory
                # merge the inventory and disk data together
                dirblock = []
                for relpath, subiterator in itertools.groupby(sorted(
                    current_inv[1] + cur_disk_dir_content,
                    key=operator.itemgetter(0)), operator.itemgetter(1)):
                    path_elements = list(subiterator)
                    if len(path_elements) == 2:
                        inv_row, disk_row = path_elements
                        # versioned, present file
                        dirblock.append((inv_row[0],
                            inv_row[1], disk_row[2],
                            disk_row[3], inv_row[4],
                            inv_row[5]))
                    elif len(path_elements[0]) == 5:
                        # unknown disk file
                        dirblock.append((path_elements[0][0],
                            path_elements[0][1], path_elements[0][2],
                            path_elements[0][3], None, None))
                    elif len(path_elements[0]) == 6:
                        # versioned, absent file.
                        dirblock.append((path_elements[0][0],
                            path_elements[0][1], 'unknown', None,
                            path_elements[0][4], path_elements[0][5]))
                    else:
                        raise NotImplementedError('unreachable code')
                yield current_inv[0], dirblock
                try:
                    current_inv = inventory_iterator.next()
                except StopIteration:
                    inv_finished = True
                try:
                    current_disk = disk_iterator.next()
                except StopIteration:
                    disk_finished = True

    def _walkdirs(self, prefix=""):
        """Walk the directories of this tree.

           :prefix: is used as the directrory to start with.
           returns a generator which yields items in the form:
                ((curren_directory_path, fileid),
                 [(file1_path, file1_name, file1_kind, None, file1_id,
                   file1_kind), ... ])
        """
        _directory = 'directory'
        # get the root in the inventory
        inv = self.inventory
        top_id = inv.path2id(prefix)
        if top_id is None:
            pending = []
        else:
            pending = [(prefix, '', _directory, None, top_id, None)]
        while pending:
            dirblock = []
            currentdir = pending.pop()
            # 0 - relpath, 1- basename, 2- kind, 3- stat, 4-id, 5-kind
            top_id = currentdir[4]
            if currentdir[0]:
                relroot = currentdir[0] + '/'
            else:
                relroot = ""
            # FIXME: stash the node in pending
            entry = inv[top_id]
            if entry.kind == 'directory':
                for name, child in entry.sorted_children():
                    dirblock.append((relroot + name, name, child.kind, None,
                        child.file_id, child.kind
                        ))
            yield (currentdir[0], entry.file_id), dirblock
            # push the user specified dirs from dirblock
            for dir in reversed(dirblock):
                if dir[2] == _directory:
                    pending.append(dir)

    @needs_tree_write_lock
    def auto_resolve(self):
        """Automatically resolve text conflicts according to contents.

        Only text conflicts are auto_resolvable. Files with no conflict markers
        are considered 'resolved', because bzr always puts conflict markers
        into files that have text conflicts.  The corresponding .THIS .BASE and
        .OTHER files are deleted, as per 'resolve'.
        :return: a tuple of ConflictLists: (un_resolved, resolved).
        """
        un_resolved = _mod_conflicts.ConflictList()
        resolved = _mod_conflicts.ConflictList()
        conflict_re = re.compile('^(<{7}|={7}|>{7})')
        for conflict in self.conflicts():
            if (conflict.typestring != 'text conflict' or
                self.kind(conflict.file_id) != 'file'):
                un_resolved.append(conflict)
                continue
            my_file = open(self.id2abspath(conflict.file_id), 'rb')
            try:
                for line in my_file:
                    if conflict_re.search(line):
                        un_resolved.append(conflict)
                        break
                else:
                    resolved.append(conflict)
            finally:
                my_file.close()
        resolved.remove_files(self)
        self.set_conflicts(un_resolved)
        return un_resolved, resolved

    @needs_read_lock
    def _check(self, references):
        """Check the tree for consistency.

        :param references: A dict with keys matching the items returned by
            self._get_check_refs(), and values from looking those keys up in
            the repository.
        """
        tree_basis = self.basis_tree()
        tree_basis.lock_read()
        try:
            repo_basis = references[('trees', self.last_revision())]
            if len(list(repo_basis.iter_changes(tree_basis))) > 0:
                raise errors.BzrCheckError(
                    "Mismatched basis inventory content.")
            self._validate()
        finally:
            tree_basis.unlock()

    def _validate(self):
        """Validate internal structures.

        This is meant mostly for the test suite. To give it a chance to detect
        corruption after actions have occurred. The default implementation is a
        just a no-op.

        :return: None. An exception should be raised if there is an error.
        """
        return

    def _get_rules_searcher(self, default_searcher):
        """See Tree._get_rules_searcher."""
        if self._rules_searcher is None:
            self._rules_searcher = super(WorkingTree,
                self)._get_rules_searcher(default_searcher)
        return self._rules_searcher

    def get_shelf_manager(self):
        """Return the ShelfManager for this WorkingTree."""
        from bzrlib.shelf import ShelfManager
        return ShelfManager(self, self._transport)


class WorkingTree2(WorkingTree):
    """This is the Format 2 working tree.

    This was the first weave based working tree.
     - uses os locks for locking.
     - uses the branch last-revision.
    """

    def __init__(self, *args, **kwargs):
        super(WorkingTree2, self).__init__(*args, **kwargs)
        # WorkingTree2 has more of a constraint that self._inventory must
        # exist. Because this is an older format, we don't mind the overhead
        # caused by the extra computation here.

        # Newer WorkingTree's should only have self._inventory set when they
        # have a read lock.
        if self._inventory is None:
            self.read_working_inventory()

    def _get_check_refs(self):
        """Return the references needed to perform a check of this tree."""
        return [('trees', self.last_revision())]

    def lock_tree_write(self):
        """See WorkingTree.lock_tree_write().

        In Format2 WorkingTrees we have a single lock for the branch and tree
        so lock_tree_write() degrades to lock_write().
        """
        self.branch.lock_write()
        try:
            return self._control_files.lock_write()
        except:
            self.branch.unlock()
            raise

    def unlock(self):
        # do non-implementation specific cleanup
        self._cleanup()

        # we share control files:
        if self._control_files._lock_count == 3:
            # _inventory_is_modified is always False during a read lock.
            if self._inventory_is_modified:
                self.flush()
            self._write_hashcache_if_dirty()

        # reverse order of locking.
        try:
            return self._control_files.unlock()
        finally:
            self.branch.unlock()


class WorkingTree3(WorkingTree):
    """This is the Format 3 working tree.

    This differs from the base WorkingTree by:
     - having its own file lock
     - having its own last-revision property.

    This is new in bzr 0.8
    """

    @needs_read_lock
    def _last_revision(self):
        """See Mutable.last_revision."""
        try:
            return self._transport.get_bytes('last-revision')
        except errors.NoSuchFile:
            return _mod_revision.NULL_REVISION

    def _change_last_revision(self, revision_id):
        """See WorkingTree._change_last_revision."""
        if revision_id is None or revision_id == _mod_revision.NULL_REVISION:
            try:
                self._transport.delete('last-revision')
            except errors.NoSuchFile:
                pass
            return False
        else:
            self._transport.put_bytes('last-revision', revision_id,
                mode=self.bzrdir._get_file_mode())
            return True

    def _get_check_refs(self):
        """Return the references needed to perform a check of this tree."""
        return [('trees', self.last_revision())]

    @needs_tree_write_lock
    def set_conflicts(self, conflicts):
        self._put_rio('conflicts', conflicts.to_stanzas(),
                      CONFLICT_HEADER_1)

    @needs_tree_write_lock
    def add_conflicts(self, new_conflicts):
        conflict_set = set(self.conflicts())
        conflict_set.update(set(list(new_conflicts)))
        self.set_conflicts(_mod_conflicts.ConflictList(sorted(conflict_set,
                                       key=_mod_conflicts.Conflict.sort_key)))

    @needs_read_lock
    def conflicts(self):
        try:
            confile = self._transport.get('conflicts')
        except errors.NoSuchFile:
            return _mod_conflicts.ConflictList()
        try:
            try:
                if confile.next() != CONFLICT_HEADER_1 + '\n':
                    raise errors.ConflictFormatError()
            except StopIteration:
                raise errors.ConflictFormatError()
            return _mod_conflicts.ConflictList.from_stanzas(RioReader(confile))
        finally:
            confile.close()

    def unlock(self):
        # do non-implementation specific cleanup
        self._cleanup()
        if self._control_files._lock_count == 1:
            # _inventory_is_modified is always False during a read lock.
            if self._inventory_is_modified:
                self.flush()
            self._write_hashcache_if_dirty()
        # reverse order of locking.
        try:
            return self._control_files.unlock()
        finally:
            self.branch.unlock()


def get_conflicted_stem(path):
    for suffix in _mod_conflicts.CONFLICT_SUFFIXES:
        if path.endswith(suffix):
            return path[:-len(suffix)]


class WorkingTreeFormat(object):
    """An encapsulation of the initialization and open routines for a format.

    Formats provide three things:
     * An initialization routine,
     * a format string,
     * an open routine.

    Formats are placed in an dict by their format string for reference
    during workingtree opening. Its not required that these be instances, they
    can be classes themselves with class methods - it simply depends on
    whether state is needed for a given format or not.

    Once a format is deprecated, just deprecate the initialize and open
    methods on the format class. Do not deprecate the object, as the
    object will be created every time regardless.
    """

    _default_format = None
    """The default format used for new trees."""

    _formats = {}
    """The known formats."""

    requires_rich_root = False

    upgrade_recommended = False

    @classmethod
    def find_format(klass, a_bzrdir):
        """Return the format for the working tree object in a_bzrdir."""
        try:
            transport = a_bzrdir.get_workingtree_transport(None)
            format_string = transport.get_bytes("format")
            return klass._formats[format_string]
        except errors.NoSuchFile:
            raise errors.NoWorkingTree(base=transport.base)
        except KeyError:
            raise errors.UnknownFormatError(format=format_string,
                                            kind="working tree")

    def __eq__(self, other):
        return self.__class__ is other.__class__

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

    @classmethod
    def get_default_format(klass):
        """Return the current default format."""
        return klass._default_format

    def get_format_string(self):
        """Return the ASCII format string that identifies this format."""
        raise NotImplementedError(self.get_format_string)

    def get_format_description(self):
        """Return the short description for this format."""
        raise NotImplementedError(self.get_format_description)

    def is_supported(self):
        """Is this format supported?

        Supported formats can be initialized and opened.
        Unsupported formats may not support initialization or committing or
        some other features depending on the reason for not being supported.
        """
        return True

    def supports_content_filtering(self):
        """True if this format supports content filtering."""
        return False

    def supports_views(self):
        """True if this format supports stored views."""
        return False

    @classmethod
    def register_format(klass, format):
        klass._formats[format.get_format_string()] = format

    @classmethod
    def set_default_format(klass, format):
        klass._default_format = format

    @classmethod
    def unregister_format(klass, format):
        del klass._formats[format.get_format_string()]


class WorkingTreeFormat2(WorkingTreeFormat):
    """The second working tree format.

    This format modified the hash cache from the format 1 hash cache.
    """

    upgrade_recommended = True

    def get_format_description(self):
        """See WorkingTreeFormat.get_format_description()."""
        return "Working tree format 2"

    def _stub_initialize_on_transport(self, transport, file_mode):
        """Workaround: create control files for a remote working tree.

        This ensures that it can later be updated and dealt with locally,
        since BzrDirFormat6 and BzrDirFormat5 cannot represent dirs with
        no working tree.  (See bug #43064).
        """
        sio = StringIO()
        inv = inventory.Inventory()
        xml5.serializer_v5.write_inventory(inv, sio, working=True)
        sio.seek(0)
        transport.put_file('inventory', sio, file_mode)
        transport.put_bytes('pending-merges', '', file_mode)

    def initialize(self, a_bzrdir, revision_id=None, from_branch=None,
                   accelerator_tree=None, hardlink=False):
        """See WorkingTreeFormat.initialize()."""
        if not isinstance(a_bzrdir.transport, LocalTransport):
            raise errors.NotLocalUrl(a_bzrdir.transport.base)
        if from_branch is not None:
            branch = from_branch
        else:
            branch = a_bzrdir.open_branch()
        if revision_id is None:
            revision_id = _mod_revision.ensure_null(branch.last_revision())
        branch.lock_write()
        try:
            branch.generate_revision_history(revision_id)
        finally:
            branch.unlock()
        inv = inventory.Inventory()
        wt = WorkingTree2(a_bzrdir.root_transport.local_abspath('.'),
                         branch,
                         inv,
                         _internal=True,
                         _format=self,
                         _bzrdir=a_bzrdir)
        basis_tree = branch.repository.revision_tree(revision_id)
        if basis_tree.inventory.root is not None:
            wt.set_root_id(basis_tree.get_root_id())
        # set the parent list and cache the basis tree.
        if _mod_revision.is_null(revision_id):
            parent_trees = []
        else:
            parent_trees = [(revision_id, basis_tree)]
        wt.set_parent_trees(parent_trees)
        transform.build_tree(basis_tree, wt)
        return wt

    def __init__(self):
        super(WorkingTreeFormat2, self).__init__()
        self._matchingbzrdir = bzrdir.BzrDirFormat6()

    def open(self, a_bzrdir, _found=False):
        """Return the WorkingTree object for a_bzrdir

        _found is a private parameter, do not use it. It is used to indicate
               if format probing has already been done.
        """
        if not _found:
            # we are being called directly and must probe.
            raise NotImplementedError
        if not isinstance(a_bzrdir.transport, LocalTransport):
            raise errors.NotLocalUrl(a_bzrdir.transport.base)
        wt = WorkingTree2(a_bzrdir.root_transport.local_abspath('.'),
                           _internal=True,
                           _format=self,
                           _bzrdir=a_bzrdir)
        return wt

class WorkingTreeFormat3(WorkingTreeFormat):
    """The second working tree format updated to record a format marker.

    This format:
        - exists within a metadir controlling .bzr
        - includes an explicit version marker for the workingtree control
          files, separate from the BzrDir format
        - modifies the hash cache format
        - is new in bzr 0.8
        - uses a LockDir to guard access for writes.
    """

    upgrade_recommended = True

    def get_format_string(self):
        """See WorkingTreeFormat.get_format_string()."""
        return "Bazaar-NG Working Tree format 3"

    def get_format_description(self):
        """See WorkingTreeFormat.get_format_description()."""
        return "Working tree format 3"

    _lock_file_name = 'lock'
    _lock_class = LockDir

    _tree_class = WorkingTree3

    def __get_matchingbzrdir(self):
        return bzrdir.BzrDirMetaFormat1()

    _matchingbzrdir = property(__get_matchingbzrdir)

    def _open_control_files(self, a_bzrdir):
        transport = a_bzrdir.get_workingtree_transport(None)
        return LockableFiles(transport, self._lock_file_name,
                             self._lock_class)

    def initialize(self, a_bzrdir, revision_id=None, from_branch=None,
                   accelerator_tree=None, hardlink=False):
        """See WorkingTreeFormat.initialize().

        :param revision_id: if supplied, create a working tree at a different
            revision than the branch is at.
        :param accelerator_tree: A tree which can be used for retrieving file
            contents more quickly than the revision tree, i.e. a workingtree.
            The revision tree will be used for cases where accelerator_tree's
            content is different.
        :param hardlink: If true, hard-link files from accelerator_tree,
            where possible.
        """
        if not isinstance(a_bzrdir.transport, LocalTransport):
            raise errors.NotLocalUrl(a_bzrdir.transport.base)
        transport = a_bzrdir.get_workingtree_transport(self)
        control_files = self._open_control_files(a_bzrdir)
        control_files.create_lock()
        control_files.lock_write()
        transport.put_bytes('format', self.get_format_string(),
            mode=a_bzrdir._get_file_mode())
        if from_branch is not None:
            branch = from_branch
        else:
            branch = a_bzrdir.open_branch()
        if revision_id is None:
            revision_id = _mod_revision.ensure_null(branch.last_revision())
        # WorkingTree3 can handle an inventory which has a unique root id.
        # as of bzr 0.12. However, bzr 0.11 and earlier fail to handle
        # those trees. And because there isn't a format bump inbetween, we
        # are maintaining compatibility with older clients.
        # inv = Inventory(root_id=gen_root_id())
        inv = self._initial_inventory()
        wt = self._tree_class(a_bzrdir.root_transport.local_abspath('.'),
                         branch,
                         inv,
                         _internal=True,
                         _format=self,
                         _bzrdir=a_bzrdir,
                         _control_files=control_files)
        wt.lock_tree_write()
        try:
            basis_tree = branch.repository.revision_tree(revision_id)
            # only set an explicit root id if there is one to set.
            if basis_tree.inventory.root is not None:
                wt.set_root_id(basis_tree.get_root_id())
            if revision_id == _mod_revision.NULL_REVISION:
                wt.set_parent_trees([])
            else:
                wt.set_parent_trees([(revision_id, basis_tree)])
            transform.build_tree(basis_tree, wt)
        finally:
            # Unlock in this order so that the unlock-triggers-flush in
            # WorkingTree is given a chance to fire.
            control_files.unlock()
            wt.unlock()
        return wt

    def _initial_inventory(self):
        return inventory.Inventory()

    def __init__(self):
        super(WorkingTreeFormat3, self).__init__()

    def open(self, a_bzrdir, _found=False):
        """Return the WorkingTree object for a_bzrdir

        _found is a private parameter, do not use it. It is used to indicate
               if format probing has already been done.
        """
        if not _found:
            # we are being called directly and must probe.
            raise NotImplementedError
        if not isinstance(a_bzrdir.transport, LocalTransport):
            raise errors.NotLocalUrl(a_bzrdir.transport.base)
        wt = self._open(a_bzrdir, self._open_control_files(a_bzrdir))
        return wt

    def _open(self, a_bzrdir, control_files):
        """Open the tree itself.

        :param a_bzrdir: the dir for the tree.
        :param control_files: the control files for the tree.
        """
        return self._tree_class(a_bzrdir.root_transport.local_abspath('.'),
                                _internal=True,
                                _format=self,
                                _bzrdir=a_bzrdir,
                                _control_files=control_files)

    def __str__(self):
        return self.get_format_string()


__default_format = WorkingTreeFormat6()
WorkingTreeFormat.register_format(__default_format)
WorkingTreeFormat.register_format(WorkingTreeFormat5())
WorkingTreeFormat.register_format(WorkingTreeFormat4())
WorkingTreeFormat.register_format(WorkingTreeFormat3())
WorkingTreeFormat.set_default_format(__default_format)
# formats which have no format string are not discoverable
# and not independently creatable, so are not registered.
_legacy_formats = [WorkingTreeFormat2(),
                   ]