~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
# Copyright (C) 2006-2012, 2016 Canonical Ltd
#
# Authors:
#   Johan Rydberg <jrydberg@gnu.org>
#
# 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


# TODO: might be nice to create a versionedfile with some type of corruption
# considered typical and check that it can be detected/corrected.

from gzip import GzipFile
from itertools import chain, izip
from StringIO import StringIO

from bzrlib import (
    errors,
    graph as _mod_graph,
    groupcompress,
    knit as _mod_knit,
    osutils,
    progress,
    transport,
    ui,
    )
from bzrlib.errors import (
                           RevisionNotPresent,
                           RevisionAlreadyPresent,
                           )
from bzrlib.knit import (
    cleanup_pack_knit,
    make_file_factory,
    make_pack_factory,
    )
from bzrlib.tests import (
    TestCase,
    TestCaseWithMemoryTransport,
    TestNotApplicable,
    TestSkipped,
    )
from bzrlib.tests.http_utils import TestCaseWithWebserver
from bzrlib.transport.memory import MemoryTransport
import bzrlib.versionedfile as versionedfile
from bzrlib.versionedfile import (
    ConstantMapper,
    HashEscapedPrefixMapper,
    PrefixMapper,
    VirtualVersionedFiles,
    make_versioned_files_factory,
    )
from bzrlib.weave import WeaveFile
from bzrlib.weavefile import write_weave
from bzrlib.tests.scenarios import load_tests_apply_scenarios


load_tests = load_tests_apply_scenarios


def get_diamond_vf(f, trailing_eol=True, left_only=False):
    """Get a diamond graph to exercise deltas and merges.

    :param trailing_eol: If True end the last line with \n.
    """
    parents = {
        'origin': (),
        'base': (('origin',),),
        'left': (('base',),),
        'right': (('base',),),
        'merged': (('left',), ('right',)),
        }
    # insert a diamond graph to exercise deltas and merges.
    if trailing_eol:
        last_char = '\n'
    else:
        last_char = ''
    f.add_lines('origin', [], ['origin' + last_char])
    f.add_lines('base', ['origin'], ['base' + last_char])
    f.add_lines('left', ['base'], ['base\n', 'left' + last_char])
    if not left_only:
        f.add_lines('right', ['base'],
            ['base\n', 'right' + last_char])
        f.add_lines('merged', ['left', 'right'],
            ['base\n', 'left\n', 'right\n', 'merged' + last_char])
    return f, parents


def get_diamond_files(files, key_length, trailing_eol=True, left_only=False,
    nograph=False, nokeys=False):
    """Get a diamond graph to exercise deltas and merges.

    This creates a 5-node graph in files. If files supports 2-length keys two
    graphs are made to exercise the support for multiple ids.

    :param trailing_eol: If True end the last line with \n.
    :param key_length: The length of keys in files. Currently supports length 1
        and 2 keys.
    :param left_only: If True do not add the right and merged nodes.
    :param nograph: If True, do not provide parents to the add_lines calls;
        this is useful for tests that need inserted data but have graphless
        stores.
    :param nokeys: If True, pass None is as the key for all insertions.
        Currently implies nograph.
    :return: The results of the add_lines calls.
    """
    if nokeys:
        nograph = True
    if key_length == 1:
        prefixes = [()]
    else:
        prefixes = [('FileA',), ('FileB',)]
    # insert a diamond graph to exercise deltas and merges.
    if trailing_eol:
        last_char = '\n'
    else:
        last_char = ''
    result = []
    def get_parents(suffix_list):
        if nograph:
            return ()
        else:
            result = [prefix + suffix for suffix in suffix_list]
            return result
    def get_key(suffix):
        if nokeys:
            return (None, )
        else:
            return (suffix,)
    # we loop over each key because that spreads the inserts across prefixes,
    # which is how commit operates.
    for prefix in prefixes:
        result.append(files.add_lines(prefix + get_key('origin'), (),
            ['origin' + last_char]))
    for prefix in prefixes:
        result.append(files.add_lines(prefix + get_key('base'),
            get_parents([('origin',)]), ['base' + last_char]))
    for prefix in prefixes:
        result.append(files.add_lines(prefix + get_key('left'),
            get_parents([('base',)]),
            ['base\n', 'left' + last_char]))
    if not left_only:
        for prefix in prefixes:
            result.append(files.add_lines(prefix + get_key('right'),
                get_parents([('base',)]),
                ['base\n', 'right' + last_char]))
        for prefix in prefixes:
            result.append(files.add_lines(prefix + get_key('merged'),
                get_parents([('left',), ('right',)]),
                ['base\n', 'left\n', 'right\n', 'merged' + last_char]))
    return result


class VersionedFileTestMixIn(object):
    """A mixin test class for testing VersionedFiles.

    This is not an adaptor-style test at this point because
    theres no dynamic substitution of versioned file implementations,
    they are strictly controlled by their owning repositories.
    """

    def get_transaction(self):
        if not hasattr(self, '_transaction'):
            self._transaction = None
        return self._transaction

    def test_add(self):
        f = self.get_file()
        f.add_lines('r0', [], ['a\n', 'b\n'])
        f.add_lines('r1', ['r0'], ['b\n', 'c\n'])
        def verify_file(f):
            versions = f.versions()
            self.assertTrue('r0' in versions)
            self.assertTrue('r1' in versions)
            self.assertEqual(f.get_lines('r0'), ['a\n', 'b\n'])
            self.assertEqual(f.get_text('r0'), 'a\nb\n')
            self.assertEqual(f.get_lines('r1'), ['b\n', 'c\n'])
            self.assertEqual(2, len(f))
            self.assertEqual(2, f.num_versions())

            self.assertRaises(RevisionNotPresent,
                f.add_lines, 'r2', ['foo'], [])
            self.assertRaises(RevisionAlreadyPresent,
                f.add_lines, 'r1', [], [])
        verify_file(f)
        # this checks that reopen with create=True does not break anything.
        f = self.reopen_file(create=True)
        verify_file(f)

    def test_adds_with_parent_texts(self):
        f = self.get_file()
        parent_texts = {}
        _, _, parent_texts['r0'] = f.add_lines('r0', [], ['a\n', 'b\n'])
        try:
            _, _, parent_texts['r1'] = f.add_lines_with_ghosts('r1',
                ['r0', 'ghost'], ['b\n', 'c\n'], parent_texts=parent_texts)
        except NotImplementedError:
            # if the format doesn't support ghosts, just add normally.
            _, _, parent_texts['r1'] = f.add_lines('r1',
                ['r0'], ['b\n', 'c\n'], parent_texts=parent_texts)
        f.add_lines('r2', ['r1'], ['c\n', 'd\n'], parent_texts=parent_texts)
        self.assertNotEqual(None, parent_texts['r0'])
        self.assertNotEqual(None, parent_texts['r1'])
        def verify_file(f):
            versions = f.versions()
            self.assertTrue('r0' in versions)
            self.assertTrue('r1' in versions)
            self.assertTrue('r2' in versions)
            self.assertEqual(f.get_lines('r0'), ['a\n', 'b\n'])
            self.assertEqual(f.get_lines('r1'), ['b\n', 'c\n'])
            self.assertEqual(f.get_lines('r2'), ['c\n', 'd\n'])
            self.assertEqual(3, f.num_versions())
            origins = f.annotate('r1')
            self.assertEqual(origins[0][0], 'r0')
            self.assertEqual(origins[1][0], 'r1')
            origins = f.annotate('r2')
            self.assertEqual(origins[0][0], 'r1')
            self.assertEqual(origins[1][0], 'r2')

        verify_file(f)
        f = self.reopen_file()
        verify_file(f)

    def test_add_unicode_content(self):
        # unicode content is not permitted in versioned files.
        # versioned files version sequences of bytes only.
        vf = self.get_file()
        self.assertRaises(errors.BzrBadParameterUnicode,
            vf.add_lines, 'a', [], ['a\n', u'b\n', 'c\n'])
        self.assertRaises(
            (errors.BzrBadParameterUnicode, NotImplementedError),
            vf.add_lines_with_ghosts, 'a', [], ['a\n', u'b\n', 'c\n'])

    def test_add_follows_left_matching_blocks(self):
        """If we change left_matching_blocks, delta changes

        Note: There are multiple correct deltas in this case, because
        we start with 1 "a" and we get 3.
        """
        vf = self.get_file()
        if isinstance(vf, WeaveFile):
            raise TestSkipped("WeaveFile ignores left_matching_blocks")
        vf.add_lines('1', [], ['a\n'])
        vf.add_lines('2', ['1'], ['a\n', 'a\n', 'a\n'],
                     left_matching_blocks=[(0, 0, 1), (1, 3, 0)])
        self.assertEqual(['a\n', 'a\n', 'a\n'], vf.get_lines('2'))
        vf.add_lines('3', ['1'], ['a\n', 'a\n', 'a\n'],
                     left_matching_blocks=[(0, 2, 1), (1, 3, 0)])
        self.assertEqual(['a\n', 'a\n', 'a\n'], vf.get_lines('3'))

    def test_inline_newline_throws(self):
        # \r characters are not permitted in lines being added
        vf = self.get_file()
        self.assertRaises(errors.BzrBadParameterContainsNewline,
            vf.add_lines, 'a', [], ['a\n\n'])
        self.assertRaises(
            (errors.BzrBadParameterContainsNewline, NotImplementedError),
            vf.add_lines_with_ghosts, 'a', [], ['a\n\n'])
        # but inline CR's are allowed
        vf.add_lines('a', [], ['a\r\n'])
        try:
            vf.add_lines_with_ghosts('b', [], ['a\r\n'])
        except NotImplementedError:
            pass

    def test_add_reserved(self):
        vf = self.get_file()
        self.assertRaises(errors.ReservedId,
            vf.add_lines, 'a:', [], ['a\n', 'b\n', 'c\n'])

    def test_add_lines_nostoresha(self):
        """When nostore_sha is supplied using old content raises."""
        vf = self.get_file()
        empty_text = ('a', [])
        sample_text_nl = ('b', ["foo\n", "bar\n"])
        sample_text_no_nl = ('c', ["foo\n", "bar"])
        shas = []
        for version, lines in (empty_text, sample_text_nl, sample_text_no_nl):
            sha, _, _ = vf.add_lines(version, [], lines)
            shas.append(sha)
        # we now have a copy of all the lines in the vf.
        for sha, (version, lines) in zip(
            shas, (empty_text, sample_text_nl, sample_text_no_nl)):
            self.assertRaises(errors.ExistingContent,
                vf.add_lines, version + "2", [], lines,
                nostore_sha=sha)
            # and no new version should have been added.
            self.assertRaises(errors.RevisionNotPresent, vf.get_lines,
                version + "2")

    def test_add_lines_with_ghosts_nostoresha(self):
        """When nostore_sha is supplied using old content raises."""
        vf = self.get_file()
        empty_text = ('a', [])
        sample_text_nl = ('b', ["foo\n", "bar\n"])
        sample_text_no_nl = ('c', ["foo\n", "bar"])
        shas = []
        for version, lines in (empty_text, sample_text_nl, sample_text_no_nl):
            sha, _, _ = vf.add_lines(version, [], lines)
            shas.append(sha)
        # we now have a copy of all the lines in the vf.
        # is the test applicable to this vf implementation?
        try:
            vf.add_lines_with_ghosts('d', [], [])
        except NotImplementedError:
            raise TestSkipped("add_lines_with_ghosts is optional")
        for sha, (version, lines) in zip(
            shas, (empty_text, sample_text_nl, sample_text_no_nl)):
            self.assertRaises(errors.ExistingContent,
                vf.add_lines_with_ghosts, version + "2", [], lines,
                nostore_sha=sha)
            # and no new version should have been added.
            self.assertRaises(errors.RevisionNotPresent, vf.get_lines,
                version + "2")

    def test_add_lines_return_value(self):
        # add_lines should return the sha1 and the text size.
        vf = self.get_file()
        empty_text = ('a', [])
        sample_text_nl = ('b', ["foo\n", "bar\n"])
        sample_text_no_nl = ('c', ["foo\n", "bar"])
        # check results for the three cases:
        for version, lines in (empty_text, sample_text_nl, sample_text_no_nl):
            # the first two elements are the same for all versioned files:
            # - the digest and the size of the text. For some versioned files
            #   additional data is returned in additional tuple elements.
            result = vf.add_lines(version, [], lines)
            self.assertEqual(3, len(result))
            self.assertEqual((osutils.sha_strings(lines), sum(map(len, lines))),
                result[0:2])
        # parents should not affect the result:
        lines = sample_text_nl[1]
        self.assertEqual((osutils.sha_strings(lines), sum(map(len, lines))),
            vf.add_lines('d', ['b', 'c'], lines)[0:2])

    def test_get_reserved(self):
        vf = self.get_file()
        self.assertRaises(errors.ReservedId, vf.get_texts, ['b:'])
        self.assertRaises(errors.ReservedId, vf.get_lines, 'b:')
        self.assertRaises(errors.ReservedId, vf.get_text, 'b:')

    def test_add_unchanged_last_line_noeol_snapshot(self):
        """Add a text with an unchanged last line with no eol should work."""
        # Test adding this in a number of chain lengths; because the interface
        # for VersionedFile does not allow forcing a specific chain length, we
        # just use a small base to get the first snapshot, then a much longer
        # first line for the next add (which will make the third add snapshot)
        # and so on. 20 has been chosen as an aribtrary figure - knits use 200
        # as a capped delta length, but ideally we would have some way of
        # tuning the test to the store (e.g. keep going until a snapshot
        # happens).
        for length in range(20):
            version_lines = {}
            vf = self.get_file('case-%d' % length)
            prefix = 'step-%d'
            parents = []
            for step in range(length):
                version = prefix % step
                lines = (['prelude \n'] * step) + ['line']
                vf.add_lines(version, parents, lines)
                version_lines[version] = lines
                parents = [version]
            vf.add_lines('no-eol', parents, ['line'])
            vf.get_texts(version_lines.keys())
            self.assertEqualDiff('line', vf.get_text('no-eol'))

    def test_get_texts_eol_variation(self):
        # similar to the failure in <http://bugs.launchpad.net/234748>
        vf = self.get_file()
        sample_text_nl = ["line\n"]
        sample_text_no_nl = ["line"]
        versions = []
        version_lines = {}
        parents = []
        for i in range(4):
            version = 'v%d' % i
            if i % 2:
                lines = sample_text_nl
            else:
                lines = sample_text_no_nl
            # left_matching blocks is an internal api; it operates on the
            # *internal* representation for a knit, which is with *all* lines
            # being normalised to end with \n - even the final line in a no_nl
            # file. Using it here ensures that a broken internal implementation
            # (which is what this test tests) will generate a correct line
            # delta (which is to say, an empty delta).
            vf.add_lines(version, parents, lines,
                left_matching_blocks=[(0, 0, 1)])
            parents = [version]
            versions.append(version)
            version_lines[version] = lines
        vf.check()
        vf.get_texts(versions)
        vf.get_texts(reversed(versions))

    def test_add_lines_with_matching_blocks_noeol_last_line(self):
        """Add a text with an unchanged last line with no eol should work."""
        from bzrlib import multiparent
        # Hand verified sha1 of the text we're adding.
        sha1 = '6a1d115ec7b60afb664dc14890b5af5ce3c827a4'
        # Create a mpdiff which adds a new line before the trailing line, and
        # reuse the last line unaltered (which can cause annotation reuse).
        # Test adding this in two situations:
        # On top of a new insertion
        vf = self.get_file('fulltext')
        vf.add_lines('noeol', [], ['line'])
        vf.add_lines('noeol2', ['noeol'], ['newline\n', 'line'],
            left_matching_blocks=[(0, 1, 1)])
        self.assertEqualDiff('newline\nline', vf.get_text('noeol2'))
        # On top of a delta
        vf = self.get_file('delta')
        vf.add_lines('base', [], ['line'])
        vf.add_lines('noeol', ['base'], ['prelude\n', 'line'])
        vf.add_lines('noeol2', ['noeol'], ['newline\n', 'line'],
            left_matching_blocks=[(1, 1, 1)])
        self.assertEqualDiff('newline\nline', vf.get_text('noeol2'))

    def test_make_mpdiffs(self):
        from bzrlib import multiparent
        vf = self.get_file('foo')
        sha1s = self._setup_for_deltas(vf)
        new_vf = self.get_file('bar')
        for version in multiparent.topo_iter(vf):
            mpdiff = vf.make_mpdiffs([version])[0]
            new_vf.add_mpdiffs([(version, vf.get_parent_map([version])[version],
                                 vf.get_sha1s([version])[version], mpdiff)])
            self.assertEqualDiff(vf.get_text(version),
                                 new_vf.get_text(version))

    def test_make_mpdiffs_with_ghosts(self):
        vf = self.get_file('foo')
        try:
            vf.add_lines_with_ghosts('text', ['ghost'], ['line\n'])
        except NotImplementedError:
            # old Weave formats do not allow ghosts
            return
        self.assertRaises(errors.RevisionNotPresent, vf.make_mpdiffs, ['ghost'])

    def _setup_for_deltas(self, f):
        self.assertFalse(f.has_version('base'))
        # add texts that should trip the knit maximum delta chain threshold
        # as well as doing parallel chains of data in knits.
        # this is done by two chains of 25 insertions
        f.add_lines('base', [], ['line\n'])
        f.add_lines('noeol', ['base'], ['line'])
        # detailed eol tests:
        # shared last line with parent no-eol
        f.add_lines('noeolsecond', ['noeol'], ['line\n', 'line'])
        # differing last line with parent, both no-eol
        f.add_lines('noeolnotshared', ['noeolsecond'], ['line\n', 'phone'])
        # add eol following a noneol parent, change content
        f.add_lines('eol', ['noeol'], ['phone\n'])
        # add eol following a noneol parent, no change content
        f.add_lines('eolline', ['noeol'], ['line\n'])
        # noeol with no parents:
        f.add_lines('noeolbase', [], ['line'])
        # noeol preceeding its leftmost parent in the output:
        # this is done by making it a merge of two parents with no common
        # anestry: noeolbase and noeol with the
        # later-inserted parent the leftmost.
        f.add_lines('eolbeforefirstparent', ['noeolbase', 'noeol'], ['line'])
        # two identical eol texts
        f.add_lines('noeoldup', ['noeol'], ['line'])
        next_parent = 'base'
        text_name = 'chain1-'
        text = ['line\n']
        sha1s = {0 :'da6d3141cb4a5e6f464bf6e0518042ddc7bfd079',
                 1 :'45e21ea146a81ea44a821737acdb4f9791c8abe7',
                 2 :'e1f11570edf3e2a070052366c582837a4fe4e9fa',
                 3 :'26b4b8626da827088c514b8f9bbe4ebf181edda1',
                 4 :'e28a5510be25ba84d31121cff00956f9970ae6f6',
                 5 :'d63ec0ce22e11dcf65a931b69255d3ac747a318d',
                 6 :'2c2888d288cb5e1d98009d822fedfe6019c6a4ea',
                 7 :'95c14da9cafbf828e3e74a6f016d87926ba234ab',
                 8 :'779e9a0b28f9f832528d4b21e17e168c67697272',
                 9 :'1f8ff4e5c6ff78ac106fcfe6b1e8cb8740ff9a8f',
                 10:'131a2ae712cf51ed62f143e3fbac3d4206c25a05',
                 11:'c5a9d6f520d2515e1ec401a8f8a67e6c3c89f199',
                 12:'31a2286267f24d8bedaa43355f8ad7129509ea85',
                 13:'dc2a7fe80e8ec5cae920973973a8ee28b2da5e0a',
                 14:'2c4b1736566b8ca6051e668de68650686a3922f2',
                 15:'5912e4ecd9b0c07be4d013e7e2bdcf9323276cde',
                 16:'b0d2e18d3559a00580f6b49804c23fea500feab3',
                 17:'8e1d43ad72f7562d7cb8f57ee584e20eb1a69fc7',
                 18:'5cf64a3459ae28efa60239e44b20312d25b253f3',
                 19:'1ebed371807ba5935958ad0884595126e8c4e823',
                 20:'2aa62a8b06fb3b3b892a3292a068ade69d5ee0d3',
                 21:'01edc447978004f6e4e962b417a4ae1955b6fe5d',
                 22:'d8d8dc49c4bf0bab401e0298bb5ad827768618bb',
                 23:'c21f62b1c482862983a8ffb2b0c64b3451876e3f',
                 24:'c0593fe795e00dff6b3c0fe857a074364d5f04fc',
                 25:'dd1a1cf2ba9cc225c3aff729953e6364bf1d1855',
                 }
        for depth in range(26):
            new_version = text_name + '%s' % depth
            text = text + ['line\n']
            f.add_lines(new_version, [next_parent], text)
            next_parent = new_version
        next_parent = 'base'
        text_name = 'chain2-'
        text = ['line\n']
        for depth in range(26):
            new_version = text_name + '%s' % depth
            text = text + ['line\n']
            f.add_lines(new_version, [next_parent], text)
            next_parent = new_version
        return sha1s

    def test_ancestry(self):
        f = self.get_file()
        self.assertEqual([], f.get_ancestry([]))
        f.add_lines('r0', [], ['a\n', 'b\n'])
        f.add_lines('r1', ['r0'], ['b\n', 'c\n'])
        f.add_lines('r2', ['r0'], ['b\n', 'c\n'])
        f.add_lines('r3', ['r2'], ['b\n', 'c\n'])
        f.add_lines('rM', ['r1', 'r2'], ['b\n', 'c\n'])
        self.assertEqual([], f.get_ancestry([]))
        versions = f.get_ancestry(['rM'])
        # there are some possibilities:
        # r0 r1 r2 rM r3
        # r0 r1 r2 r3 rM
        # etc
        # so we check indexes
        r0 = versions.index('r0')
        r1 = versions.index('r1')
        r2 = versions.index('r2')
        self.assertFalse('r3' in versions)
        rM = versions.index('rM')
        self.assertTrue(r0 < r1)
        self.assertTrue(r0 < r2)
        self.assertTrue(r1 < rM)
        self.assertTrue(r2 < rM)

        self.assertRaises(RevisionNotPresent,
            f.get_ancestry, ['rM', 'rX'])

        self.assertEqual(set(f.get_ancestry('rM')),
            set(f.get_ancestry('rM', topo_sorted=False)))

    def test_mutate_after_finish(self):
        self._transaction = 'before'
        f = self.get_file()
        self._transaction = 'after'
        self.assertRaises(errors.OutSideTransaction, f.add_lines, '', [], [])
        self.assertRaises(errors.OutSideTransaction, f.add_lines_with_ghosts, '', [], [])

    def test_copy_to(self):
        f = self.get_file()
        f.add_lines('0', [], ['a\n'])
        t = MemoryTransport()
        f.copy_to('foo', t)
        for suffix in self.get_factory().get_suffixes():
            self.assertTrue(t.has('foo' + suffix))

    def test_get_suffixes(self):
        f = self.get_file()
        # and should be a list
        self.assertTrue(isinstance(self.get_factory().get_suffixes(), list))

    def test_get_parent_map(self):
        f = self.get_file()
        f.add_lines('r0', [], ['a\n', 'b\n'])
        self.assertEqual(
            {'r0':()}, f.get_parent_map(['r0']))
        f.add_lines('r1', ['r0'], ['a\n', 'b\n'])
        self.assertEqual(
            {'r1':('r0',)}, f.get_parent_map(['r1']))
        self.assertEqual(
            {'r0':(),
             'r1':('r0',)},
            f.get_parent_map(['r0', 'r1']))
        f.add_lines('r2', [], ['a\n', 'b\n'])
        f.add_lines('r3', [], ['a\n', 'b\n'])
        f.add_lines('m', ['r0', 'r1', 'r2', 'r3'], ['a\n', 'b\n'])
        self.assertEqual(
            {'m':('r0', 'r1', 'r2', 'r3')}, f.get_parent_map(['m']))
        self.assertEqual({}, f.get_parent_map('y'))
        self.assertEqual(
            {'r0':(),
             'r1':('r0',)},
            f.get_parent_map(['r0', 'y', 'r1']))

    def test_annotate(self):
        f = self.get_file()
        f.add_lines('r0', [], ['a\n', 'b\n'])
        f.add_lines('r1', ['r0'], ['c\n', 'b\n'])
        origins = f.annotate('r1')
        self.assertEqual(origins[0][0], 'r1')
        self.assertEqual(origins[1][0], 'r0')

        self.assertRaises(RevisionNotPresent,
            f.annotate, 'foo')

    def test_detection(self):
        # Test weaves detect corruption.
        #
        # Weaves contain a checksum of their texts.
        # When a text is extracted, this checksum should be
        # verified.

        w = self.get_file_corrupted_text()

        self.assertEqual('hello\n', w.get_text('v1'))
        self.assertRaises(errors.WeaveInvalidChecksum, w.get_text, 'v2')
        self.assertRaises(errors.WeaveInvalidChecksum, w.get_lines, 'v2')
        self.assertRaises(errors.WeaveInvalidChecksum, w.check)

        w = self.get_file_corrupted_checksum()

        self.assertEqual('hello\n', w.get_text('v1'))
        self.assertRaises(errors.WeaveInvalidChecksum, w.get_text, 'v2')
        self.assertRaises(errors.WeaveInvalidChecksum, w.get_lines, 'v2')
        self.assertRaises(errors.WeaveInvalidChecksum, w.check)

    def get_file_corrupted_text(self):
        """Return a versioned file with corrupt text but valid metadata."""
        raise NotImplementedError(self.get_file_corrupted_text)

    def reopen_file(self, name='foo'):
        """Open the versioned file from disk again."""
        raise NotImplementedError(self.reopen_file)

    def test_iter_lines_added_or_present_in_versions(self):
        # test that we get at least an equalset of the lines added by
        # versions in the weave
        # the ordering here is to make a tree so that dumb searches have
        # more changes to muck up.

        class InstrumentedProgress(progress.ProgressTask):

            def __init__(self):
                progress.ProgressTask.__init__(self)
                self.updates = []

            def update(self, msg=None, current=None, total=None):
                self.updates.append((msg, current, total))

        vf = self.get_file()
        # add a base to get included
        vf.add_lines('base', [], ['base\n'])
        # add a ancestor to be included on one side
        vf.add_lines('lancestor', [], ['lancestor\n'])
        # add a ancestor to be included on the other side
        vf.add_lines('rancestor', ['base'], ['rancestor\n'])
        # add a child of rancestor with no eofile-nl
        vf.add_lines('child', ['rancestor'], ['base\n', 'child\n'])
        # add a child of lancestor and base to join the two roots
        vf.add_lines('otherchild',
                     ['lancestor', 'base'],
                     ['base\n', 'lancestor\n', 'otherchild\n'])
        def iter_with_versions(versions, expected):
            # now we need to see what lines are returned, and how often.
            lines = {}
            progress = InstrumentedProgress()
            # iterate over the lines
            for line in vf.iter_lines_added_or_present_in_versions(versions,
                pb=progress):
                lines.setdefault(line, 0)
                lines[line] += 1
            if []!= progress.updates:
                self.assertEqual(expected, progress.updates)
            return lines
        lines = iter_with_versions(['child', 'otherchild'],
                                   [('Walking content', 0, 2),
                                    ('Walking content', 1, 2),
                                    ('Walking content', 2, 2)])
        # we must see child and otherchild
        self.assertTrue(lines[('child\n', 'child')] > 0)
        self.assertTrue(lines[('otherchild\n', 'otherchild')] > 0)
        # we dont care if we got more than that.

        # test all lines
        lines = iter_with_versions(None, [('Walking content', 0, 5),
                                          ('Walking content', 1, 5),
                                          ('Walking content', 2, 5),
                                          ('Walking content', 3, 5),
                                          ('Walking content', 4, 5),
                                          ('Walking content', 5, 5)])
        # all lines must be seen at least once
        self.assertTrue(lines[('base\n', 'base')] > 0)
        self.assertTrue(lines[('lancestor\n', 'lancestor')] > 0)
        self.assertTrue(lines[('rancestor\n', 'rancestor')] > 0)
        self.assertTrue(lines[('child\n', 'child')] > 0)
        self.assertTrue(lines[('otherchild\n', 'otherchild')] > 0)

    def test_add_lines_with_ghosts(self):
        # some versioned file formats allow lines to be added with parent
        # information that is > than that in the format. Formats that do
        # not support this need to raise NotImplementedError on the
        # add_lines_with_ghosts api.
        vf = self.get_file()
        # add a revision with ghost parents
        # The preferred form is utf8, but we should translate when needed
        parent_id_unicode = u'b\xbfse'
        parent_id_utf8 = parent_id_unicode.encode('utf8')
        try:
            vf.add_lines_with_ghosts('notbxbfse', [parent_id_utf8], [])
        except NotImplementedError:
            # check the other ghost apis are also not implemented
            self.assertRaises(NotImplementedError, vf.get_ancestry_with_ghosts, ['foo'])
            self.assertRaises(NotImplementedError, vf.get_parents_with_ghosts, 'foo')
            return
        vf = self.reopen_file()
        # test key graph related apis: getncestry, _graph, get_parents
        # has_version
        # - these are ghost unaware and must not be reflect ghosts
        self.assertEqual(['notbxbfse'], vf.get_ancestry('notbxbfse'))
        self.assertFalse(vf.has_version(parent_id_utf8))
        # we have _with_ghost apis to give us ghost information.
        self.assertEqual([parent_id_utf8, 'notbxbfse'], vf.get_ancestry_with_ghosts(['notbxbfse']))
        self.assertEqual([parent_id_utf8], vf.get_parents_with_ghosts('notbxbfse'))
        # if we add something that is a ghost of another, it should correct the
        # results of the prior apis
        vf.add_lines(parent_id_utf8, [], [])
        self.assertEqual([parent_id_utf8, 'notbxbfse'], vf.get_ancestry(['notbxbfse']))
        self.assertEqual({'notbxbfse':(parent_id_utf8,)},
            vf.get_parent_map(['notbxbfse']))
        self.assertTrue(vf.has_version(parent_id_utf8))
        # we have _with_ghost apis to give us ghost information.
        self.assertEqual([parent_id_utf8, 'notbxbfse'],
            vf.get_ancestry_with_ghosts(['notbxbfse']))
        self.assertEqual([parent_id_utf8], vf.get_parents_with_ghosts('notbxbfse'))

    def test_add_lines_with_ghosts_after_normal_revs(self):
        # some versioned file formats allow lines to be added with parent
        # information that is > than that in the format. Formats that do
        # not support this need to raise NotImplementedError on the
        # add_lines_with_ghosts api.
        vf = self.get_file()
        # probe for ghost support
        try:
            vf.add_lines_with_ghosts('base', [], ['line\n', 'line_b\n'])
        except NotImplementedError:
            return
        vf.add_lines_with_ghosts('references_ghost',
                                 ['base', 'a_ghost'],
                                 ['line\n', 'line_b\n', 'line_c\n'])
        origins = vf.annotate('references_ghost')
        self.assertEqual(('base', 'line\n'), origins[0])
        self.assertEqual(('base', 'line_b\n'), origins[1])
        self.assertEqual(('references_ghost', 'line_c\n'), origins[2])

    def test_readonly_mode(self):
        t = self.get_transport()
        factory = self.get_factory()
        vf = factory('id', t, 0777, create=True, access_mode='w')
        vf = factory('id', t, access_mode='r')
        self.assertRaises(errors.ReadOnlyError, vf.add_lines, 'base', [], [])
        self.assertRaises(errors.ReadOnlyError,
                          vf.add_lines_with_ghosts,
                          'base',
                          [],
                          [])

    def test_get_sha1s(self):
        # check the sha1 data is available
        vf = self.get_file()
        # a simple file
        vf.add_lines('a', [], ['a\n'])
        # the same file, different metadata
        vf.add_lines('b', ['a'], ['a\n'])
        # a file differing only in last newline.
        vf.add_lines('c', [], ['a'])
        self.assertEqual({
            'a': '3f786850e387550fdab836ed7e6dc881de23001b',
            'c': '86f7e437faa5a7fce15d1ddcb9eaeaea377667b8',
            'b': '3f786850e387550fdab836ed7e6dc881de23001b',
            },
            vf.get_sha1s(['a', 'c', 'b']))


class TestWeave(TestCaseWithMemoryTransport, VersionedFileTestMixIn):

    def get_file(self, name='foo'):
        return WeaveFile(name, self.get_transport(),
                         create=True,
                         get_scope=self.get_transaction)

    def get_file_corrupted_text(self):
        w = WeaveFile('foo', self.get_transport(),
                      create=True,
                      get_scope=self.get_transaction)
        w.add_lines('v1', [], ['hello\n'])
        w.add_lines('v2', ['v1'], ['hello\n', 'there\n'])

        # We are going to invasively corrupt the text
        # Make sure the internals of weave are the same
        self.assertEqual([('{', 0)
                        , 'hello\n'
                        , ('}', None)
                        , ('{', 1)
                        , 'there\n'
                        , ('}', None)
                        ], w._weave)

        self.assertEqual(['f572d396fae9206628714fb2ce00f72e94f2258f'
                        , '90f265c6e75f1c8f9ab76dcf85528352c5f215ef'
                        ], w._sha1s)
        w.check()

        # Corrupted
        w._weave[4] = 'There\n'
        return w

    def get_file_corrupted_checksum(self):
        w = self.get_file_corrupted_text()
        # Corrected
        w._weave[4] = 'there\n'
        self.assertEqual('hello\nthere\n', w.get_text('v2'))

        #Invalid checksum, first digit changed
        w._sha1s[1] =  'f0f265c6e75f1c8f9ab76dcf85528352c5f215ef'
        return w

    def reopen_file(self, name='foo', create=False):
        return WeaveFile(name, self.get_transport(),
                         create=create,
                         get_scope=self.get_transaction)

    def test_no_implicit_create(self):
        self.assertRaises(errors.NoSuchFile,
                          WeaveFile,
                          'foo',
                          self.get_transport(),
                          get_scope=self.get_transaction)

    def get_factory(self):
        return WeaveFile


class TestPlanMergeVersionedFile(TestCaseWithMemoryTransport):

    def setUp(self):
        super(TestPlanMergeVersionedFile, self).setUp()
        mapper = PrefixMapper()
        factory = make_file_factory(True, mapper)
        self.vf1 = factory(self.get_transport('root-1'))
        self.vf2 = factory(self.get_transport('root-2'))
        self.plan_merge_vf = versionedfile._PlanMergeVersionedFile('root')
        self.plan_merge_vf.fallback_versionedfiles.extend([self.vf1, self.vf2])

    def test_add_lines(self):
        self.plan_merge_vf.add_lines(('root', 'a:'), [], [])
        self.assertRaises(ValueError, self.plan_merge_vf.add_lines,
            ('root', 'a'), [], [])
        self.assertRaises(ValueError, self.plan_merge_vf.add_lines,
            ('root', 'a:'), None, [])
        self.assertRaises(ValueError, self.plan_merge_vf.add_lines,
            ('root', 'a:'), [], None)

    def setup_abcde(self):
        self.vf1.add_lines(('root', 'A'), [], ['a'])
        self.vf1.add_lines(('root', 'B'), [('root', 'A')], ['b'])
        self.vf2.add_lines(('root', 'C'), [], ['c'])
        self.vf2.add_lines(('root', 'D'), [('root', 'C')], ['d'])
        self.plan_merge_vf.add_lines(('root', 'E:'),
            [('root', 'B'), ('root', 'D')], ['e'])

    def test_get_parents(self):
        self.setup_abcde()
        self.assertEqual({('root', 'B'):(('root', 'A'),)},
            self.plan_merge_vf.get_parent_map([('root', 'B')]))
        self.assertEqual({('root', 'D'):(('root', 'C'),)},
            self.plan_merge_vf.get_parent_map([('root', 'D')]))
        self.assertEqual({('root', 'E:'):(('root', 'B'),('root', 'D'))},
            self.plan_merge_vf.get_parent_map([('root', 'E:')]))
        self.assertEqual({},
            self.plan_merge_vf.get_parent_map([('root', 'F')]))
        self.assertEqual({
                ('root', 'B'):(('root', 'A'),),
                ('root', 'D'):(('root', 'C'),),
                ('root', 'E:'):(('root', 'B'),('root', 'D')),
                },
            self.plan_merge_vf.get_parent_map(
                [('root', 'B'), ('root', 'D'), ('root', 'E:'), ('root', 'F')]))

    def test_get_record_stream(self):
        self.setup_abcde()
        def get_record(suffix):
            return self.plan_merge_vf.get_record_stream(
                [('root', suffix)], 'unordered', True).next()
        self.assertEqual('a', get_record('A').get_bytes_as('fulltext'))
        self.assertEqual('c', get_record('C').get_bytes_as('fulltext'))
        self.assertEqual('e', get_record('E:').get_bytes_as('fulltext'))
        self.assertEqual('absent', get_record('F').storage_kind)


class TestReadonlyHttpMixin(object):

    def get_transaction(self):
        return 1

    def test_readonly_http_works(self):
        # we should be able to read from http with a versioned file.
        vf = self.get_file()
        # try an empty file access
        readonly_vf = self.get_factory()('foo',
            transport.get_transport_from_url(self.get_readonly_url('.')))
        self.assertEqual([], readonly_vf.versions())

    def test_readonly_http_works_with_feeling(self):
        # we should be able to read from http with a versioned file.
        vf = self.get_file()
        # now with feeling.
        vf.add_lines('1', [], ['a\n'])
        vf.add_lines('2', ['1'], ['b\n', 'a\n'])
        readonly_vf = self.get_factory()('foo',
            transport.get_transport_from_url(self.get_readonly_url('.')))
        self.assertEqual(['1', '2'], vf.versions())
        self.assertEqual(['1', '2'], readonly_vf.versions())
        for version in readonly_vf.versions():
            readonly_vf.get_lines(version)


class TestWeaveHTTP(TestCaseWithWebserver, TestReadonlyHttpMixin):

    def get_file(self):
        return WeaveFile('foo', self.get_transport(),
                         create=True,
                         get_scope=self.get_transaction)

    def get_factory(self):
        return WeaveFile


class MergeCasesMixin(object):

    def doMerge(self, base, a, b, mp):
        from cStringIO import StringIO
        from textwrap import dedent

        def addcrlf(x):
            return x + '\n'

        w = self.get_file()
        w.add_lines('text0', [], map(addcrlf, base))
        w.add_lines('text1', ['text0'], map(addcrlf, a))
        w.add_lines('text2', ['text0'], map(addcrlf, b))

        self.log_contents(w)

        self.log('merge plan:')
        p = list(w.plan_merge('text1', 'text2'))
        for state, line in p:
            if line:
                self.log('%12s | %s' % (state, line[:-1]))

        self.log('merge:')
        mt = StringIO()
        mt.writelines(w.weave_merge(p))
        mt.seek(0)
        self.log(mt.getvalue())

        mp = map(addcrlf, mp)
        self.assertEqual(mt.readlines(), mp)


    def testOneInsert(self):
        self.doMerge([],
                     ['aa'],
                     [],
                     ['aa'])

    def testSeparateInserts(self):
        self.doMerge(['aaa', 'bbb', 'ccc'],
                     ['aaa', 'xxx', 'bbb', 'ccc'],
                     ['aaa', 'bbb', 'yyy', 'ccc'],
                     ['aaa', 'xxx', 'bbb', 'yyy', 'ccc'])

    def testSameInsert(self):
        self.doMerge(['aaa', 'bbb', 'ccc'],
                     ['aaa', 'xxx', 'bbb', 'ccc'],
                     ['aaa', 'xxx', 'bbb', 'yyy', 'ccc'],
                     ['aaa', 'xxx', 'bbb', 'yyy', 'ccc'])
    overlappedInsertExpected = ['aaa', 'xxx', 'yyy', 'bbb']
    def testOverlappedInsert(self):
        self.doMerge(['aaa', 'bbb'],
                     ['aaa', 'xxx', 'yyy', 'bbb'],
                     ['aaa', 'xxx', 'bbb'], self.overlappedInsertExpected)

        # really it ought to reduce this to
        # ['aaa', 'xxx', 'yyy', 'bbb']


    def testClashReplace(self):
        self.doMerge(['aaa'],
                     ['xxx'],
                     ['yyy', 'zzz'],
                     ['<<<<<<< ', 'xxx', '=======', 'yyy', 'zzz',
                      '>>>>>>> '])

    def testNonClashInsert1(self):
        self.doMerge(['aaa'],
                     ['xxx', 'aaa'],
                     ['yyy', 'zzz'],
                     ['<<<<<<< ', 'xxx', 'aaa', '=======', 'yyy', 'zzz',
                      '>>>>>>> '])

    def testNonClashInsert2(self):
        self.doMerge(['aaa'],
                     ['aaa'],
                     ['yyy', 'zzz'],
                     ['yyy', 'zzz'])


    def testDeleteAndModify(self):
        """Clashing delete and modification.

        If one side modifies a region and the other deletes it then
        there should be a conflict with one side blank.
        """

        #######################################
        # skippd, not working yet
        return

        self.doMerge(['aaa', 'bbb', 'ccc'],
                     ['aaa', 'ddd', 'ccc'],
                     ['aaa', 'ccc'],
                     ['<<<<<<<< ', 'aaa', '=======', '>>>>>>> ', 'ccc'])

    def _test_merge_from_strings(self, base, a, b, expected):
        w = self.get_file()
        w.add_lines('text0', [], base.splitlines(True))
        w.add_lines('text1', ['text0'], a.splitlines(True))
        w.add_lines('text2', ['text0'], b.splitlines(True))
        self.log('merge plan:')
        p = list(w.plan_merge('text1', 'text2'))
        for state, line in p:
            if line:
                self.log('%12s | %s' % (state, line[:-1]))
        self.log('merge result:')
        result_text = ''.join(w.weave_merge(p))
        self.log(result_text)
        self.assertEqualDiff(result_text, expected)

    def test_weave_merge_conflicts(self):
        # does weave merge properly handle plans that end with unchanged?
        result = ''.join(self.get_file().weave_merge([('new-a', 'hello\n')]))
        self.assertEqual(result, 'hello\n')

    def test_deletion_extended(self):
        """One side deletes, the other deletes more.
        """
        base = """\
            line 1
            line 2
            line 3
            """
        a = """\
            line 1
            line 2
            """
        b = """\
            line 1
            """
        result = """\
            line 1
<<<<<<<\x20
            line 2
=======
>>>>>>>\x20
            """
        self._test_merge_from_strings(base, a, b, result)

    def test_deletion_overlap(self):
        """Delete overlapping regions with no other conflict.

        Arguably it'd be better to treat these as agreement, rather than
        conflict, but for now conflict is safer.
        """
        base = """\
            start context
            int a() {}
            int b() {}
            int c() {}
            end context
            """
        a = """\
            start context
            int a() {}
            end context
            """
        b = """\
            start context
            int c() {}
            end context
            """
        result = """\
            start context
<<<<<<<\x20
            int a() {}
=======
            int c() {}
>>>>>>>\x20
            end context
            """
        self._test_merge_from_strings(base, a, b, result)

    def test_agreement_deletion(self):
        """Agree to delete some lines, without conflicts."""
        base = """\
            start context
            base line 1
            base line 2
            end context
            """
        a = """\
            start context
            base line 1
            end context
            """
        b = """\
            start context
            base line 1
            end context
            """
        result = """\
            start context
            base line 1
            end context
            """
        self._test_merge_from_strings(base, a, b, result)

    def test_sync_on_deletion(self):
        """Specific case of merge where we can synchronize incorrectly.

        A previous version of the weave merge concluded that the two versions
        agreed on deleting line 2, and this could be a synchronization point.
        Line 1 was then considered in isolation, and thought to be deleted on
        both sides.

        It's better to consider the whole thing as a disagreement region.
        """
        base = """\
            start context
            base line 1
            base line 2
            end context
            """
        a = """\
            start context
            base line 1
            a's replacement line 2
            end context
            """
        b = """\
            start context
            b replaces
            both lines
            end context
            """
        result = """\
            start context
<<<<<<<\x20
            base line 1
            a's replacement line 2
=======
            b replaces
            both lines
>>>>>>>\x20
            end context
            """
        self._test_merge_from_strings(base, a, b, result)


class TestWeaveMerge(TestCaseWithMemoryTransport, MergeCasesMixin):

    def get_file(self, name='foo'):
        return WeaveFile(name, self.get_transport(),
                         create=True)

    def log_contents(self, w):
        self.log('weave is:')
        tmpf = StringIO()
        write_weave(w, tmpf)
        self.log(tmpf.getvalue())

    overlappedInsertExpected = ['aaa', '<<<<<<< ', 'xxx', 'yyy', '=======',
                                'xxx', '>>>>>>> ', 'bbb']


class TestContentFactoryAdaption(TestCaseWithMemoryTransport):

    def test_select_adaptor(self):
        """Test expected adapters exist."""
        # One scenario for each lookup combination we expect to use.
        # Each is source_kind, requested_kind, adapter class
        scenarios = [
            ('knit-delta-gz', 'fulltext', _mod_knit.DeltaPlainToFullText),
            ('knit-ft-gz', 'fulltext', _mod_knit.FTPlainToFullText),
            ('knit-annotated-delta-gz', 'knit-delta-gz',
                _mod_knit.DeltaAnnotatedToUnannotated),
            ('knit-annotated-delta-gz', 'fulltext',
                _mod_knit.DeltaAnnotatedToFullText),
            ('knit-annotated-ft-gz', 'knit-ft-gz',
                _mod_knit.FTAnnotatedToUnannotated),
            ('knit-annotated-ft-gz', 'fulltext',
                _mod_knit.FTAnnotatedToFullText),
            ]
        for source, requested, klass in scenarios:
            adapter_factory = versionedfile.adapter_registry.get(
                (source, requested))
            adapter = adapter_factory(None)
            self.assertIsInstance(adapter, klass)

    def get_knit(self, annotated=True):
        mapper = ConstantMapper('knit')
        transport = self.get_transport()
        return make_file_factory(annotated, mapper)(transport)

    def helpGetBytes(self, f, ft_adapter, delta_adapter):
        """Grab the interested adapted texts for tests."""
        # origin is a fulltext
        entries = f.get_record_stream([('origin',)], 'unordered', False)
        base = entries.next()
        ft_data = ft_adapter.get_bytes(base)
        # merged is both a delta and multiple parents.
        entries = f.get_record_stream([('merged',)], 'unordered', False)
        merged = entries.next()
        delta_data = delta_adapter.get_bytes(merged)
        return ft_data, delta_data

    def test_deannotation_noeol(self):
        """Test converting annotated knits to unannotated knits."""
        # we need a full text, and a delta
        f = self.get_knit()
        get_diamond_files(f, 1, trailing_eol=False)
        ft_data, delta_data = self.helpGetBytes(f,
            _mod_knit.FTAnnotatedToUnannotated(None),
            _mod_knit.DeltaAnnotatedToUnannotated(None))
        self.assertEqual(
            'version origin 1 b284f94827db1fa2970d9e2014f080413b547a7e\n'
            'origin\n'
            'end origin\n',
            GzipFile(mode='rb', fileobj=StringIO(ft_data)).read())
        self.assertEqual(
            'version merged 4 32c2e79763b3f90e8ccde37f9710b6629c25a796\n'
            '1,2,3\nleft\nright\nmerged\nend merged\n',
            GzipFile(mode='rb', fileobj=StringIO(delta_data)).read())

    def test_deannotation(self):
        """Test converting annotated knits to unannotated knits."""
        # we need a full text, and a delta
        f = self.get_knit()
        get_diamond_files(f, 1)
        ft_data, delta_data = self.helpGetBytes(f,
            _mod_knit.FTAnnotatedToUnannotated(None),
            _mod_knit.DeltaAnnotatedToUnannotated(None))
        self.assertEqual(
            'version origin 1 00e364d235126be43292ab09cb4686cf703ddc17\n'
            'origin\n'
            'end origin\n',
            GzipFile(mode='rb', fileobj=StringIO(ft_data)).read())
        self.assertEqual(
            'version merged 3 ed8bce375198ea62444dc71952b22cfc2b09226d\n'
            '2,2,2\nright\nmerged\nend merged\n',
            GzipFile(mode='rb', fileobj=StringIO(delta_data)).read())

    def test_annotated_to_fulltext_no_eol(self):
        """Test adapting annotated knits to full texts (for -> weaves)."""
        # we need a full text, and a delta
        f = self.get_knit()
        get_diamond_files(f, 1, trailing_eol=False)
        # Reconstructing a full text requires a backing versioned file, and it
        # must have the base lines requested from it.
        logged_vf = versionedfile.RecordingVersionedFilesDecorator(f)
        ft_data, delta_data = self.helpGetBytes(f,
            _mod_knit.FTAnnotatedToFullText(None),
            _mod_knit.DeltaAnnotatedToFullText(logged_vf))
        self.assertEqual('origin', ft_data)
        self.assertEqual('base\nleft\nright\nmerged', delta_data)
        self.assertEqual([('get_record_stream', [('left',)], 'unordered',
            True)], logged_vf.calls)

    def test_annotated_to_fulltext(self):
        """Test adapting annotated knits to full texts (for -> weaves)."""
        # we need a full text, and a delta
        f = self.get_knit()
        get_diamond_files(f, 1)
        # Reconstructing a full text requires a backing versioned file, and it
        # must have the base lines requested from it.
        logged_vf = versionedfile.RecordingVersionedFilesDecorator(f)
        ft_data, delta_data = self.helpGetBytes(f,
            _mod_knit.FTAnnotatedToFullText(None),
            _mod_knit.DeltaAnnotatedToFullText(logged_vf))
        self.assertEqual('origin\n', ft_data)
        self.assertEqual('base\nleft\nright\nmerged\n', delta_data)
        self.assertEqual([('get_record_stream', [('left',)], 'unordered',
            True)], logged_vf.calls)

    def test_unannotated_to_fulltext(self):
        """Test adapting unannotated knits to full texts.

        This is used for -> weaves, and for -> annotated knits.
        """
        # we need a full text, and a delta
        f = self.get_knit(annotated=False)
        get_diamond_files(f, 1)
        # Reconstructing a full text requires a backing versioned file, and it
        # must have the base lines requested from it.
        logged_vf = versionedfile.RecordingVersionedFilesDecorator(f)
        ft_data, delta_data = self.helpGetBytes(f,
            _mod_knit.FTPlainToFullText(None),
            _mod_knit.DeltaPlainToFullText(logged_vf))
        self.assertEqual('origin\n', ft_data)
        self.assertEqual('base\nleft\nright\nmerged\n', delta_data)
        self.assertEqual([('get_record_stream', [('left',)], 'unordered',
            True)], logged_vf.calls)

    def test_unannotated_to_fulltext_no_eol(self):
        """Test adapting unannotated knits to full texts.

        This is used for -> weaves, and for -> annotated knits.
        """
        # we need a full text, and a delta
        f = self.get_knit(annotated=False)
        get_diamond_files(f, 1, trailing_eol=False)
        # Reconstructing a full text requires a backing versioned file, and it
        # must have the base lines requested from it.
        logged_vf = versionedfile.RecordingVersionedFilesDecorator(f)
        ft_data, delta_data = self.helpGetBytes(f,
            _mod_knit.FTPlainToFullText(None),
            _mod_knit.DeltaPlainToFullText(logged_vf))
        self.assertEqual('origin', ft_data)
        self.assertEqual('base\nleft\nright\nmerged', delta_data)
        self.assertEqual([('get_record_stream', [('left',)], 'unordered',
            True)], logged_vf.calls)


class TestKeyMapper(TestCaseWithMemoryTransport):
    """Tests for various key mapping logic."""

    def test_identity_mapper(self):
        mapper = versionedfile.ConstantMapper("inventory")
        self.assertEqual("inventory", mapper.map(('foo@ar',)))
        self.assertEqual("inventory", mapper.map(('quux',)))

    def test_prefix_mapper(self):
        #format5: plain
        mapper = versionedfile.PrefixMapper()
        self.assertEqual("file-id", mapper.map(("file-id", "revision-id")))
        self.assertEqual("new-id", mapper.map(("new-id", "revision-id")))
        self.assertEqual(('file-id',), mapper.unmap("file-id"))
        self.assertEqual(('new-id',), mapper.unmap("new-id"))

    def test_hash_prefix_mapper(self):
        #format6: hash + plain
        mapper = versionedfile.HashPrefixMapper()
        self.assertEqual("9b/file-id", mapper.map(("file-id", "revision-id")))
        self.assertEqual("45/new-id", mapper.map(("new-id", "revision-id")))
        self.assertEqual(('file-id',), mapper.unmap("9b/file-id"))
        self.assertEqual(('new-id',), mapper.unmap("45/new-id"))

    def test_hash_escaped_mapper(self):
        #knit1: hash + escaped
        mapper = versionedfile.HashEscapedPrefixMapper()
        self.assertEqual("88/%2520", mapper.map((" ", "revision-id")))
        self.assertEqual("ed/fil%2545-%2549d", mapper.map(("filE-Id",
            "revision-id")))
        self.assertEqual("88/ne%2557-%2549d", mapper.map(("neW-Id",
            "revision-id")))
        self.assertEqual(('filE-Id',), mapper.unmap("ed/fil%2545-%2549d"))
        self.assertEqual(('neW-Id',), mapper.unmap("88/ne%2557-%2549d"))


class TestVersionedFiles(TestCaseWithMemoryTransport):
    """Tests for the multiple-file variant of VersionedFile."""

    # We want to be sure of behaviour for:
    # weaves prefix layout (weave texts)
    # individually named weaves (weave inventories)
    # annotated knits - prefix|hash|hash-escape layout, we test the third only
    #                   as it is the most complex mapper.
    # individually named knits
    # individual no-graph knits in packs (signatures)
    # individual graph knits in packs (inventories)
    # individual graph nocompression knits in packs (revisions)
    # plain text knits in packs (texts)
    len_one_scenarios = [
        ('weave-named', {
            'cleanup':None,
            'factory':make_versioned_files_factory(WeaveFile,
                ConstantMapper('inventory')),
            'graph':True,
            'key_length':1,
            'support_partial_insertion': False,
            }),
        ('named-knit', {
            'cleanup':None,
            'factory':make_file_factory(False, ConstantMapper('revisions')),
            'graph':True,
            'key_length':1,
            'support_partial_insertion': False,
            }),
        ('named-nograph-nodelta-knit-pack', {
            'cleanup':cleanup_pack_knit,
            'factory':make_pack_factory(False, False, 1),
            'graph':False,
            'key_length':1,
            'support_partial_insertion': False,
            }),
        ('named-graph-knit-pack', {
            'cleanup':cleanup_pack_knit,
            'factory':make_pack_factory(True, True, 1),
            'graph':True,
            'key_length':1,
            'support_partial_insertion': True,
            }),
        ('named-graph-nodelta-knit-pack', {
            'cleanup':cleanup_pack_knit,
            'factory':make_pack_factory(True, False, 1),
            'graph':True,
            'key_length':1,
            'support_partial_insertion': False,
            }),
        ('groupcompress-nograph', {
            'cleanup':groupcompress.cleanup_pack_group,
            'factory':groupcompress.make_pack_factory(False, False, 1),
            'graph': False,
            'key_length':1,
            'support_partial_insertion':False,
            }),
        ]
    len_two_scenarios = [
        ('weave-prefix', {
            'cleanup':None,
            'factory':make_versioned_files_factory(WeaveFile,
                PrefixMapper()),
            'graph':True,
            'key_length':2,
            'support_partial_insertion': False,
            }),
        ('annotated-knit-escape', {
            'cleanup':None,
            'factory':make_file_factory(True, HashEscapedPrefixMapper()),
            'graph':True,
            'key_length':2,
            'support_partial_insertion': False,
            }),
        ('plain-knit-pack', {
            'cleanup':cleanup_pack_knit,
            'factory':make_pack_factory(True, True, 2),
            'graph':True,
            'key_length':2,
            'support_partial_insertion': True,
            }),
        ('groupcompress', {
            'cleanup':groupcompress.cleanup_pack_group,
            'factory':groupcompress.make_pack_factory(True, False, 1),
            'graph': True,
            'key_length':1,
            'support_partial_insertion':False,
            }),
        ]

    scenarios = len_one_scenarios + len_two_scenarios

    def get_versionedfiles(self, relpath='files'):
        transport = self.get_transport(relpath)
        if relpath != '.':
            transport.mkdir('.')
        files = self.factory(transport)
        if self.cleanup is not None:
            self.addCleanup(self.cleanup, files)
        return files

    def get_simple_key(self, suffix):
        """Return a key for the object under test."""
        if self.key_length == 1:
            return (suffix,)
        else:
            return ('FileA',) + (suffix,)

    def test_add_fallback_implies_without_fallbacks(self):
        f = self.get_versionedfiles('files')
        if getattr(f, 'add_fallback_versioned_files', None) is None:
            raise TestNotApplicable("%s doesn't support fallbacks"
                                    % (f.__class__.__name__,))
        g = self.get_versionedfiles('fallback')
        key_a = self.get_simple_key('a')
        g.add_lines(key_a, [], ['\n'])
        f.add_fallback_versioned_files(g)
        self.assertTrue(key_a in f.get_parent_map([key_a]))
        self.assertFalse(key_a in f.without_fallbacks().get_parent_map([key_a]))

    def test_add_lines(self):
        f = self.get_versionedfiles()
        key0 = self.get_simple_key('r0')
        key1 = self.get_simple_key('r1')
        key2 = self.get_simple_key('r2')
        keyf = self.get_simple_key('foo')
        f.add_lines(key0, [], ['a\n', 'b\n'])
        if self.graph:
            f.add_lines(key1, [key0], ['b\n', 'c\n'])
        else:
            f.add_lines(key1, [], ['b\n', 'c\n'])
        keys = f.keys()
        self.assertTrue(key0 in keys)
        self.assertTrue(key1 in keys)
        records = []
        for record in f.get_record_stream([key0, key1], 'unordered', True):
            records.append((record.key, record.get_bytes_as('fulltext')))
        records.sort()
        self.assertEqual([(key0, 'a\nb\n'), (key1, 'b\nc\n')], records)

    def test__add_text(self):
        f = self.get_versionedfiles()
        key0 = self.get_simple_key('r0')
        key1 = self.get_simple_key('r1')
        key2 = self.get_simple_key('r2')
        keyf = self.get_simple_key('foo')
        f._add_text(key0, [], 'a\nb\n')
        if self.graph:
            f._add_text(key1, [key0], 'b\nc\n')
        else:
            f._add_text(key1, [], 'b\nc\n')
        keys = f.keys()
        self.assertTrue(key0 in keys)
        self.assertTrue(key1 in keys)
        records = []
        for record in f.get_record_stream([key0, key1], 'unordered', True):
            records.append((record.key, record.get_bytes_as('fulltext')))
        records.sort()
        self.assertEqual([(key0, 'a\nb\n'), (key1, 'b\nc\n')], records)

    def test_annotate(self):
        files = self.get_versionedfiles()
        self.get_diamond_files(files)
        if self.key_length == 1:
            prefix = ()
        else:
            prefix = ('FileA',)
        # introduced full text
        origins = files.annotate(prefix + ('origin',))
        self.assertEqual([
            (prefix + ('origin',), 'origin\n')],
            origins)
        # a delta
        origins = files.annotate(prefix + ('base',))
        self.assertEqual([
            (prefix + ('base',), 'base\n')],
            origins)
        # a merge
        origins = files.annotate(prefix + ('merged',))
        if self.graph:
            self.assertEqual([
                (prefix + ('base',), 'base\n'),
                (prefix + ('left',), 'left\n'),
                (prefix + ('right',), 'right\n'),
                (prefix + ('merged',), 'merged\n')
                ],
                origins)
        else:
            # Without a graph everything is new.
            self.assertEqual([
                (prefix + ('merged',), 'base\n'),
                (prefix + ('merged',), 'left\n'),
                (prefix + ('merged',), 'right\n'),
                (prefix + ('merged',), 'merged\n')
                ],
                origins)
        self.assertRaises(RevisionNotPresent,
            files.annotate, prefix + ('missing-key',))

    def test_check_no_parameters(self):
        files = self.get_versionedfiles()

    def test_check_progressbar_parameter(self):
        """A progress bar can be supplied because check can be a generator."""
        pb = ui.ui_factory.nested_progress_bar()
        self.addCleanup(pb.finished)
        files = self.get_versionedfiles()
        files.check(progress_bar=pb)

    def test_check_with_keys_becomes_generator(self):
        files = self.get_versionedfiles()
        self.get_diamond_files(files)
        keys = files.keys()
        entries = files.check(keys=keys)
        seen = set()
        # Texts output should be fulltexts.
        self.capture_stream(files, entries, seen.add,
            files.get_parent_map(keys), require_fulltext=True)
        # All texts should be output.
        self.assertEqual(set(keys), seen)

    def test_clear_cache(self):
        files = self.get_versionedfiles()
        files.clear_cache()

    def test_construct(self):
        """Each parameterised test can be constructed on a transport."""
        files = self.get_versionedfiles()

    def get_diamond_files(self, files, trailing_eol=True, left_only=False,
        nokeys=False):
        return get_diamond_files(files, self.key_length,
            trailing_eol=trailing_eol, nograph=not self.graph,
            left_only=left_only, nokeys=nokeys)

    def _add_content_nostoresha(self, add_lines):
        """When nostore_sha is supplied using old content raises."""
        vf = self.get_versionedfiles()
        empty_text = ('a', [])
        sample_text_nl = ('b', ["foo\n", "bar\n"])
        sample_text_no_nl = ('c', ["foo\n", "bar"])
        shas = []
        for version, lines in (empty_text, sample_text_nl, sample_text_no_nl):
            if add_lines:
                sha, _, _ = vf.add_lines(self.get_simple_key(version), [],
                                         lines)
            else:
                sha, _, _ = vf._add_text(self.get_simple_key(version), [],
                                         ''.join(lines))
            shas.append(sha)
        # we now have a copy of all the lines in the vf.
        for sha, (version, lines) in zip(
            shas, (empty_text, sample_text_nl, sample_text_no_nl)):
            new_key = self.get_simple_key(version + "2")
            self.assertRaises(errors.ExistingContent,
                vf.add_lines, new_key, [], lines,
                nostore_sha=sha)
            self.assertRaises(errors.ExistingContent,
                vf._add_text, new_key, [], ''.join(lines),
                nostore_sha=sha)
            # and no new version should have been added.
            record = vf.get_record_stream([new_key], 'unordered', True).next()
            self.assertEqual('absent', record.storage_kind)

    def test_add_lines_nostoresha(self):
        self._add_content_nostoresha(add_lines=True)

    def test__add_text_nostoresha(self):
        self._add_content_nostoresha(add_lines=False)

    def test_add_lines_return(self):
        files = self.get_versionedfiles()
        # save code by using the stock data insertion helper.
        adds = self.get_diamond_files(files)
        results = []
        # We can only validate the first 2 elements returned from add_lines.
        for add in adds:
            self.assertEqual(3, len(add))
            results.append(add[:2])
        if self.key_length == 1:
            self.assertEqual([
                ('00e364d235126be43292ab09cb4686cf703ddc17', 7),
                ('51c64a6f4fc375daf0d24aafbabe4d91b6f4bb44', 5),
                ('a8478686da38e370e32e42e8a0c220e33ee9132f', 10),
                ('9ef09dfa9d86780bdec9219a22560c6ece8e0ef1', 11),
                ('ed8bce375198ea62444dc71952b22cfc2b09226d', 23)],
                results)
        elif self.key_length == 2:
            self.assertEqual([
                ('00e364d235126be43292ab09cb4686cf703ddc17', 7),
                ('00e364d235126be43292ab09cb4686cf703ddc17', 7),
                ('51c64a6f4fc375daf0d24aafbabe4d91b6f4bb44', 5),
                ('51c64a6f4fc375daf0d24aafbabe4d91b6f4bb44', 5),
                ('a8478686da38e370e32e42e8a0c220e33ee9132f', 10),
                ('a8478686da38e370e32e42e8a0c220e33ee9132f', 10),
                ('9ef09dfa9d86780bdec9219a22560c6ece8e0ef1', 11),
                ('9ef09dfa9d86780bdec9219a22560c6ece8e0ef1', 11),
                ('ed8bce375198ea62444dc71952b22cfc2b09226d', 23),
                ('ed8bce375198ea62444dc71952b22cfc2b09226d', 23)],
                results)

    def test_add_lines_no_key_generates_chk_key(self):
        files = self.get_versionedfiles()
        # save code by using the stock data insertion helper.
        adds = self.get_diamond_files(files, nokeys=True)
        results = []
        # We can only validate the first 2 elements returned from add_lines.
        for add in adds:
            self.assertEqual(3, len(add))
            results.append(add[:2])
        if self.key_length == 1:
            self.assertEqual([
                ('00e364d235126be43292ab09cb4686cf703ddc17', 7),
                ('51c64a6f4fc375daf0d24aafbabe4d91b6f4bb44', 5),
                ('a8478686da38e370e32e42e8a0c220e33ee9132f', 10),
                ('9ef09dfa9d86780bdec9219a22560c6ece8e0ef1', 11),
                ('ed8bce375198ea62444dc71952b22cfc2b09226d', 23)],
                results)
            # Check the added items got CHK keys.
            self.assertEqual(set([
                ('sha1:00e364d235126be43292ab09cb4686cf703ddc17',),
                ('sha1:51c64a6f4fc375daf0d24aafbabe4d91b6f4bb44',),
                ('sha1:9ef09dfa9d86780bdec9219a22560c6ece8e0ef1',),
                ('sha1:a8478686da38e370e32e42e8a0c220e33ee9132f',),
                ('sha1:ed8bce375198ea62444dc71952b22cfc2b09226d',),
                ]),
                files.keys())
        elif self.key_length == 2:
            self.assertEqual([
                ('00e364d235126be43292ab09cb4686cf703ddc17', 7),
                ('00e364d235126be43292ab09cb4686cf703ddc17', 7),
                ('51c64a6f4fc375daf0d24aafbabe4d91b6f4bb44', 5),
                ('51c64a6f4fc375daf0d24aafbabe4d91b6f4bb44', 5),
                ('a8478686da38e370e32e42e8a0c220e33ee9132f', 10),
                ('a8478686da38e370e32e42e8a0c220e33ee9132f', 10),
                ('9ef09dfa9d86780bdec9219a22560c6ece8e0ef1', 11),
                ('9ef09dfa9d86780bdec9219a22560c6ece8e0ef1', 11),
                ('ed8bce375198ea62444dc71952b22cfc2b09226d', 23),
                ('ed8bce375198ea62444dc71952b22cfc2b09226d', 23)],
                results)
            # Check the added items got CHK keys.
            self.assertEqual(set([
                ('FileA', 'sha1:00e364d235126be43292ab09cb4686cf703ddc17'),
                ('FileA', 'sha1:51c64a6f4fc375daf0d24aafbabe4d91b6f4bb44'),
                ('FileA', 'sha1:9ef09dfa9d86780bdec9219a22560c6ece8e0ef1'),
                ('FileA', 'sha1:a8478686da38e370e32e42e8a0c220e33ee9132f'),
                ('FileA', 'sha1:ed8bce375198ea62444dc71952b22cfc2b09226d'),
                ('FileB', 'sha1:00e364d235126be43292ab09cb4686cf703ddc17'),
                ('FileB', 'sha1:51c64a6f4fc375daf0d24aafbabe4d91b6f4bb44'),
                ('FileB', 'sha1:9ef09dfa9d86780bdec9219a22560c6ece8e0ef1'),
                ('FileB', 'sha1:a8478686da38e370e32e42e8a0c220e33ee9132f'),
                ('FileB', 'sha1:ed8bce375198ea62444dc71952b22cfc2b09226d'),
                ]),
                files.keys())

    def test_empty_lines(self):
        """Empty files can be stored."""
        f = self.get_versionedfiles()
        key_a = self.get_simple_key('a')
        f.add_lines(key_a, [], [])
        self.assertEqual('',
            f.get_record_stream([key_a], 'unordered', True
                ).next().get_bytes_as('fulltext'))
        key_b = self.get_simple_key('b')
        f.add_lines(key_b, self.get_parents([key_a]), [])
        self.assertEqual('',
            f.get_record_stream([key_b], 'unordered', True
                ).next().get_bytes_as('fulltext'))

    def test_newline_only(self):
        f = self.get_versionedfiles()
        key_a = self.get_simple_key('a')
        f.add_lines(key_a, [], ['\n'])
        self.assertEqual('\n',
            f.get_record_stream([key_a], 'unordered', True
                ).next().get_bytes_as('fulltext'))
        key_b = self.get_simple_key('b')
        f.add_lines(key_b, self.get_parents([key_a]), ['\n'])
        self.assertEqual('\n',
            f.get_record_stream([key_b], 'unordered', True
                ).next().get_bytes_as('fulltext'))

    def test_get_known_graph_ancestry(self):
        f = self.get_versionedfiles()
        if not self.graph:
            raise TestNotApplicable('ancestry info only relevant with graph.')
        key_a = self.get_simple_key('a')
        key_b = self.get_simple_key('b')
        key_c = self.get_simple_key('c')
        # A
        # |\
        # | B
        # |/
        # C
        f.add_lines(key_a, [], ['\n'])
        f.add_lines(key_b, [key_a], ['\n'])
        f.add_lines(key_c, [key_a, key_b], ['\n'])
        kg = f.get_known_graph_ancestry([key_c])
        self.assertIsInstance(kg, _mod_graph.KnownGraph)
        self.assertEqual([key_a, key_b, key_c], list(kg.topo_sort()))

    def test_known_graph_with_fallbacks(self):
        f = self.get_versionedfiles('files')
        if not self.graph:
            raise TestNotApplicable('ancestry info only relevant with graph.')
        if getattr(f, 'add_fallback_versioned_files', None) is None:
            raise TestNotApplicable("%s doesn't support fallbacks"
                                    % (f.__class__.__name__,))
        key_a = self.get_simple_key('a')
        key_b = self.get_simple_key('b')
        key_c = self.get_simple_key('c')
        # A     only in fallback
        # |\
        # | B
        # |/
        # C
        g = self.get_versionedfiles('fallback')
        g.add_lines(key_a, [], ['\n'])
        f.add_fallback_versioned_files(g)
        f.add_lines(key_b, [key_a], ['\n'])
        f.add_lines(key_c, [key_a, key_b], ['\n'])
        kg = f.get_known_graph_ancestry([key_c])
        self.assertEqual([key_a, key_b, key_c], list(kg.topo_sort()))

    def test_get_record_stream_empty(self):
        """An empty stream can be requested without error."""
        f = self.get_versionedfiles()
        entries = f.get_record_stream([], 'unordered', False)
        self.assertEqual([], list(entries))

    def assertValidStorageKind(self, storage_kind):
        """Assert that storage_kind is a valid storage_kind."""
        self.assertSubset([storage_kind],
            ['mpdiff', 'knit-annotated-ft', 'knit-annotated-delta',
             'knit-ft', 'knit-delta', 'chunked', 'fulltext',
             'knit-annotated-ft-gz', 'knit-annotated-delta-gz', 'knit-ft-gz',
             'knit-delta-gz',
             'knit-delta-closure', 'knit-delta-closure-ref',
             'groupcompress-block', 'groupcompress-block-ref'])

    def capture_stream(self, f, entries, on_seen, parents,
        require_fulltext=False):
        """Capture a stream for testing."""
        for factory in entries:
            on_seen(factory.key)
            self.assertValidStorageKind(factory.storage_kind)
            if factory.sha1 is not None:
                self.assertEqual(f.get_sha1s([factory.key])[factory.key],
                    factory.sha1)
            self.assertEqual(parents[factory.key], factory.parents)
            self.assertIsInstance(factory.get_bytes_as(factory.storage_kind),
                str)
            if require_fulltext:
                factory.get_bytes_as('fulltext')

    def test_get_record_stream_interface(self):
        """each item in a stream has to provide a regular interface."""
        files = self.get_versionedfiles()
        self.get_diamond_files(files)
        keys, _ = self.get_keys_and_sort_order()
        parent_map = files.get_parent_map(keys)
        entries = files.get_record_stream(keys, 'unordered', False)
        seen = set()
        self.capture_stream(files, entries, seen.add, parent_map)
        self.assertEqual(set(keys), seen)

    def get_keys_and_sort_order(self):
        """Get diamond test keys list, and their sort ordering."""
        if self.key_length == 1:
            keys = [('merged',), ('left',), ('right',), ('base',)]
            sort_order = {('merged',):2, ('left',):1, ('right',):1, ('base',):0}
        else:
            keys = [
                ('FileA', 'merged'), ('FileA', 'left'), ('FileA', 'right'),
                ('FileA', 'base'),
                ('FileB', 'merged'), ('FileB', 'left'), ('FileB', 'right'),
                ('FileB', 'base'),
                ]
            sort_order = {
                ('FileA', 'merged'):2, ('FileA', 'left'):1, ('FileA', 'right'):1,
                ('FileA', 'base'):0,
                ('FileB', 'merged'):2, ('FileB', 'left'):1, ('FileB', 'right'):1,
                ('FileB', 'base'):0,
                }
        return keys, sort_order

    def get_keys_and_groupcompress_sort_order(self):
        """Get diamond test keys list, and their groupcompress sort ordering."""
        if self.key_length == 1:
            keys = [('merged',), ('left',), ('right',), ('base',)]
            sort_order = {('merged',):0, ('left',):1, ('right',):1, ('base',):2}
        else:
            keys = [
                ('FileA', 'merged'), ('FileA', 'left'), ('FileA', 'right'),
                ('FileA', 'base'),
                ('FileB', 'merged'), ('FileB', 'left'), ('FileB', 'right'),
                ('FileB', 'base'),
                ]
            sort_order = {
                ('FileA', 'merged'):0, ('FileA', 'left'):1, ('FileA', 'right'):1,
                ('FileA', 'base'):2,
                ('FileB', 'merged'):3, ('FileB', 'left'):4, ('FileB', 'right'):4,
                ('FileB', 'base'):5,
                }
        return keys, sort_order

    def test_get_record_stream_interface_ordered(self):
        """each item in a stream has to provide a regular interface."""
        files = self.get_versionedfiles()
        self.get_diamond_files(files)
        keys, sort_order = self.get_keys_and_sort_order()
        parent_map = files.get_parent_map(keys)
        entries = files.get_record_stream(keys, 'topological', False)
        seen = []
        self.capture_stream(files, entries, seen.append, parent_map)
        self.assertStreamOrder(sort_order, seen, keys)

    def test_get_record_stream_interface_ordered_with_delta_closure(self):
        """each item must be accessible as a fulltext."""
        files = self.get_versionedfiles()
        self.get_diamond_files(files)
        keys, sort_order = self.get_keys_and_sort_order()
        parent_map = files.get_parent_map(keys)
        entries = files.get_record_stream(keys, 'topological', True)
        seen = []
        for factory in entries:
            seen.append(factory.key)
            self.assertValidStorageKind(factory.storage_kind)
            self.assertSubset([factory.sha1],
                [None, files.get_sha1s([factory.key])[factory.key]])
            self.assertEqual(parent_map[factory.key], factory.parents)
            # self.assertEqual(files.get_text(factory.key),
            ft_bytes = factory.get_bytes_as('fulltext')
            self.assertIsInstance(ft_bytes, str)
            chunked_bytes = factory.get_bytes_as('chunked')
            self.assertEqualDiff(ft_bytes, ''.join(chunked_bytes))

        self.assertStreamOrder(sort_order, seen, keys)

    def test_get_record_stream_interface_groupcompress(self):
        """each item in a stream has to provide a regular interface."""
        files = self.get_versionedfiles()
        self.get_diamond_files(files)
        keys, sort_order = self.get_keys_and_groupcompress_sort_order()
        parent_map = files.get_parent_map(keys)
        entries = files.get_record_stream(keys, 'groupcompress', False)
        seen = []
        self.capture_stream(files, entries, seen.append, parent_map)
        self.assertStreamOrder(sort_order, seen, keys)

    def assertStreamOrder(self, sort_order, seen, keys):
        self.assertEqual(len(set(seen)), len(keys))
        if self.key_length == 1:
            lows = {():0}
        else:
            lows = {('FileA',):0, ('FileB',):0}
        if not self.graph:
            self.assertEqual(set(keys), set(seen))
        else:
            for key in seen:
                sort_pos = sort_order[key]
                self.assertTrue(sort_pos >= lows[key[:-1]],
                    "Out of order in sorted stream: %r, %r" % (key, seen))
                lows[key[:-1]] = sort_pos

    def test_get_record_stream_unknown_storage_kind_raises(self):
        """Asking for a storage kind that the stream cannot supply raises."""
        files = self.get_versionedfiles()
        self.get_diamond_files(files)
        if self.key_length == 1:
            keys = [('merged',), ('left',), ('right',), ('base',)]
        else:
            keys = [
                ('FileA', 'merged'), ('FileA', 'left'), ('FileA', 'right'),
                ('FileA', 'base'),
                ('FileB', 'merged'), ('FileB', 'left'), ('FileB', 'right'),
                ('FileB', 'base'),
                ]
        parent_map = files.get_parent_map(keys)
        entries = files.get_record_stream(keys, 'unordered', False)
        # We track the contents because we should be able to try, fail a
        # particular kind and then ask for one that works and continue.
        seen = set()
        for factory in entries:
            seen.add(factory.key)
            self.assertValidStorageKind(factory.storage_kind)
            if factory.sha1 is not None:
                self.assertEqual(files.get_sha1s([factory.key])[factory.key],
                                 factory.sha1)
            self.assertEqual(parent_map[factory.key], factory.parents)
            # currently no stream emits mpdiff
            self.assertRaises(errors.UnavailableRepresentation,
                factory.get_bytes_as, 'mpdiff')
            self.assertIsInstance(factory.get_bytes_as(factory.storage_kind),
                str)
        self.assertEqual(set(keys), seen)

    def test_get_record_stream_missing_records_are_absent(self):
        files = self.get_versionedfiles()
        self.get_diamond_files(files)
        if self.key_length == 1:
            keys = [('merged',), ('left',), ('right',), ('absent',), ('base',)]
        else:
            keys = [
                ('FileA', 'merged'), ('FileA', 'left'), ('FileA', 'right'),
                ('FileA', 'absent'), ('FileA', 'base'),
                ('FileB', 'merged'), ('FileB', 'left'), ('FileB', 'right'),
                ('FileB', 'absent'), ('FileB', 'base'),
                ('absent', 'absent'),
                ]
        parent_map = files.get_parent_map(keys)
        entries = files.get_record_stream(keys, 'unordered', False)
        self.assertAbsentRecord(files, keys, parent_map, entries)
        entries = files.get_record_stream(keys, 'topological', False)
        self.assertAbsentRecord(files, keys, parent_map, entries)

    def assertRecordHasContent(self, record, bytes):
        """Assert that record has the bytes bytes."""
        self.assertEqual(bytes, record.get_bytes_as('fulltext'))
        self.assertEqual(bytes, ''.join(record.get_bytes_as('chunked')))

    def test_get_record_stream_native_formats_are_wire_ready_one_ft(self):
        files = self.get_versionedfiles()
        key = self.get_simple_key('foo')
        files.add_lines(key, (), ['my text\n', 'content'])
        stream = files.get_record_stream([key], 'unordered', False)
        record = stream.next()
        if record.storage_kind in ('chunked', 'fulltext'):
            # chunked and fulltext representations are for direct use not wire
            # serialisation: check they are able to be used directly. To send
            # such records over the wire translation will be needed.
            self.assertRecordHasContent(record, "my text\ncontent")
        else:
            bytes = [record.get_bytes_as(record.storage_kind)]
            network_stream = versionedfile.NetworkRecordStream(bytes).read()
            source_record = record
            records = []
            for record in network_stream:
                records.append(record)
                self.assertEqual(source_record.storage_kind,
                    record.storage_kind)
                self.assertEqual(source_record.parents, record.parents)
                self.assertEqual(
                    source_record.get_bytes_as(source_record.storage_kind),
                    record.get_bytes_as(record.storage_kind))
            self.assertEqual(1, len(records))

    def assertStreamMetaEqual(self, records, expected, stream):
        """Assert that streams expected and stream have the same records.

        :param records: A list to collect the seen records.
        :return: A generator of the records in stream.
        """
        # We make assertions during copying to catch things early for
        # easier debugging.
        for record, ref_record in izip(stream, expected):
            records.append(record)
            self.assertEqual(ref_record.key, record.key)
            self.assertEqual(ref_record.storage_kind, record.storage_kind)
            self.assertEqual(ref_record.parents, record.parents)
            yield record

    def stream_to_bytes_or_skip_counter(self, skipped_records, full_texts,
        stream):
        """Convert a stream to a bytes iterator.

        :param skipped_records: A list with one element to increment when a
            record is skipped.
        :param full_texts: A dict from key->fulltext representation, for
            checking chunked or fulltext stored records.
        :param stream: A record_stream.
        :return: An iterator over the bytes of each record.
        """
        for record in stream:
            if record.storage_kind in ('chunked', 'fulltext'):
                skipped_records[0] += 1
                # check the content is correct for direct use.
                self.assertRecordHasContent(record, full_texts[record.key])
            else:
                yield record.get_bytes_as(record.storage_kind)

    def test_get_record_stream_native_formats_are_wire_ready_ft_delta(self):
        files = self.get_versionedfiles()
        target_files = self.get_versionedfiles('target')
        key = self.get_simple_key('ft')
        key_delta = self.get_simple_key('delta')
        files.add_lines(key, (), ['my text\n', 'content'])
        if self.graph:
            delta_parents = (key,)
        else:
            delta_parents = ()
        files.add_lines(key_delta, delta_parents, ['different\n', 'content\n'])
        local = files.get_record_stream([key, key_delta], 'unordered', False)
        ref = files.get_record_stream([key, key_delta], 'unordered', False)
        skipped_records = [0]
        full_texts = {
            key: "my text\ncontent",
            key_delta: "different\ncontent\n",
            }
        byte_stream = self.stream_to_bytes_or_skip_counter(
            skipped_records, full_texts, local)
        network_stream = versionedfile.NetworkRecordStream(byte_stream).read()
        records = []
        # insert the stream from the network into a versioned files object so we can
        # check the content was carried across correctly without doing delta
        # inspection.
        target_files.insert_record_stream(
            self.assertStreamMetaEqual(records, ref, network_stream))
        # No duplicates on the wire thank you!
        self.assertEqual(2, len(records) + skipped_records[0])
        if len(records):
            # if any content was copied it all must have all been.
            self.assertIdenticalVersionedFile(files, target_files)

    def test_get_record_stream_native_formats_are_wire_ready_delta(self):
        # copy a delta over the wire
        files = self.get_versionedfiles()
        target_files = self.get_versionedfiles('target')
        key = self.get_simple_key('ft')
        key_delta = self.get_simple_key('delta')
        files.add_lines(key, (), ['my text\n', 'content'])
        if self.graph:
            delta_parents = (key,)
        else:
            delta_parents = ()
        files.add_lines(key_delta, delta_parents, ['different\n', 'content\n'])
        # Copy the basis text across so we can reconstruct the delta during
        # insertion into target.
        target_files.insert_record_stream(files.get_record_stream([key],
            'unordered', False))
        local = files.get_record_stream([key_delta], 'unordered', False)
        ref = files.get_record_stream([key_delta], 'unordered', False)
        skipped_records = [0]
        full_texts = {
            key_delta: "different\ncontent\n",
            }
        byte_stream = self.stream_to_bytes_or_skip_counter(
            skipped_records, full_texts, local)
        network_stream = versionedfile.NetworkRecordStream(byte_stream).read()
        records = []
        # insert the stream from the network into a versioned files object so we can
        # check the content was carried across correctly without doing delta
        # inspection during check_stream.
        target_files.insert_record_stream(
            self.assertStreamMetaEqual(records, ref, network_stream))
        # No duplicates on the wire thank you!
        self.assertEqual(1, len(records) + skipped_records[0])
        if len(records):
            # if any content was copied it all must have all been
            self.assertIdenticalVersionedFile(files, target_files)

    def test_get_record_stream_wire_ready_delta_closure_included(self):
        # copy a delta over the wire with the ability to get its full text.
        files = self.get_versionedfiles()
        key = self.get_simple_key('ft')
        key_delta = self.get_simple_key('delta')
        files.add_lines(key, (), ['my text\n', 'content'])
        if self.graph:
            delta_parents = (key,)
        else:
            delta_parents = ()
        files.add_lines(key_delta, delta_parents, ['different\n', 'content\n'])
        local = files.get_record_stream([key_delta], 'unordered', True)
        ref = files.get_record_stream([key_delta], 'unordered', True)
        skipped_records = [0]
        full_texts = {
            key_delta: "different\ncontent\n",
            }
        byte_stream = self.stream_to_bytes_or_skip_counter(
            skipped_records, full_texts, local)
        network_stream = versionedfile.NetworkRecordStream(byte_stream).read()
        records = []
        # insert the stream from the network into a versioned files object so we can
        # check the content was carried across correctly without doing delta
        # inspection during check_stream.
        for record in self.assertStreamMetaEqual(records, ref, network_stream):
            # we have to be able to get the full text out:
            self.assertRecordHasContent(record, full_texts[record.key])
        # No duplicates on the wire thank you!
        self.assertEqual(1, len(records) + skipped_records[0])

    def assertAbsentRecord(self, files, keys, parents, entries):
        """Helper for test_get_record_stream_missing_records_are_absent."""
        seen = set()
        for factory in entries:
            seen.add(factory.key)
            if factory.key[-1] == 'absent':
                self.assertEqual('absent', factory.storage_kind)
                self.assertEqual(None, factory.sha1)
                self.assertEqual(None, factory.parents)
            else:
                self.assertValidStorageKind(factory.storage_kind)
                if factory.sha1 is not None:
                    sha1 = files.get_sha1s([factory.key])[factory.key]
                    self.assertEqual(sha1, factory.sha1)
                self.assertEqual(parents[factory.key], factory.parents)
                self.assertIsInstance(factory.get_bytes_as(factory.storage_kind),
                    str)
        self.assertEqual(set(keys), seen)

    def test_filter_absent_records(self):
        """Requested missing records can be filter trivially."""
        files = self.get_versionedfiles()
        self.get_diamond_files(files)
        keys, _ = self.get_keys_and_sort_order()
        parent_map = files.get_parent_map(keys)
        # Add an absent record in the middle of the present keys. (We don't ask
        # for just absent keys to ensure that content before and after the
        # absent keys is still delivered).
        present_keys = list(keys)
        if self.key_length == 1:
            keys.insert(2, ('extra',))
        else:
            keys.insert(2, ('extra', 'extra'))
        entries = files.get_record_stream(keys, 'unordered', False)
        seen = set()
        self.capture_stream(files, versionedfile.filter_absent(entries), seen.add,
            parent_map)
        self.assertEqual(set(present_keys), seen)

    def get_mapper(self):
        """Get a mapper suitable for the key length of the test interface."""
        if self.key_length == 1:
            return ConstantMapper('source')
        else:
            return HashEscapedPrefixMapper()

    def get_parents(self, parents):
        """Get parents, taking self.graph into consideration."""
        if self.graph:
            return parents
        else:
            return None

    def test_get_annotator(self):
        files = self.get_versionedfiles()
        self.get_diamond_files(files)
        origin_key = self.get_simple_key('origin')
        base_key = self.get_simple_key('base')
        left_key = self.get_simple_key('left')
        right_key = self.get_simple_key('right')
        merged_key = self.get_simple_key('merged')
        # annotator = files.get_annotator()
        # introduced full text
        origins, lines = files.get_annotator().annotate(origin_key)
        self.assertEqual([(origin_key,)], origins)
        self.assertEqual(['origin\n'], lines)
        # a delta
        origins, lines = files.get_annotator().annotate(base_key)
        self.assertEqual([(base_key,)], origins)
        # a merge
        origins, lines = files.get_annotator().annotate(merged_key)
        if self.graph:
            self.assertEqual([
                (base_key,),
                (left_key,),
                (right_key,),
                (merged_key,),
                ], origins)
        else:
            # Without a graph everything is new.
            self.assertEqual([
                (merged_key,),
                (merged_key,),
                (merged_key,),
                (merged_key,),
                ], origins)
        self.assertRaises(RevisionNotPresent,
            files.get_annotator().annotate, self.get_simple_key('missing-key'))

    def test_get_parent_map(self):
        files = self.get_versionedfiles()
        if self.key_length == 1:
            parent_details = [
                (('r0',), self.get_parents(())),
                (('r1',), self.get_parents((('r0',),))),
                (('r2',), self.get_parents(())),
                (('r3',), self.get_parents(())),
                (('m',), self.get_parents((('r0',),('r1',),('r2',),('r3',)))),
                ]
        else:
            parent_details = [
                (('FileA', 'r0'), self.get_parents(())),
                (('FileA', 'r1'), self.get_parents((('FileA', 'r0'),))),
                (('FileA', 'r2'), self.get_parents(())),
                (('FileA', 'r3'), self.get_parents(())),
                (('FileA', 'm'), self.get_parents((('FileA', 'r0'),
                    ('FileA', 'r1'), ('FileA', 'r2'), ('FileA', 'r3')))),
                ]
        for key, parents in parent_details:
            files.add_lines(key, parents, [])
            # immediately after adding it should be queryable.
            self.assertEqual({key:parents}, files.get_parent_map([key]))
        # We can ask for an empty set
        self.assertEqual({}, files.get_parent_map([]))
        # We can ask for many keys
        all_parents = dict(parent_details)
        self.assertEqual(all_parents, files.get_parent_map(all_parents.keys()))
        # Absent keys are just not included in the result.
        keys = all_parents.keys()
        if self.key_length == 1:
            keys.insert(1, ('missing',))
        else:
            keys.insert(1, ('missing', 'missing'))
        # Absent keys are just ignored
        self.assertEqual(all_parents, files.get_parent_map(keys))

    def test_get_sha1s(self):
        files = self.get_versionedfiles()
        self.get_diamond_files(files)
        if self.key_length == 1:
            keys = [('base',), ('origin',), ('left',), ('merged',), ('right',)]
        else:
            # ask for shas from different prefixes.
            keys = [
                ('FileA', 'base'), ('FileB', 'origin'), ('FileA', 'left'),
                ('FileA', 'merged'), ('FileB', 'right'),
                ]
        self.assertEqual({
            keys[0]: '51c64a6f4fc375daf0d24aafbabe4d91b6f4bb44',
            keys[1]: '00e364d235126be43292ab09cb4686cf703ddc17',
            keys[2]: 'a8478686da38e370e32e42e8a0c220e33ee9132f',
            keys[3]: 'ed8bce375198ea62444dc71952b22cfc2b09226d',
            keys[4]: '9ef09dfa9d86780bdec9219a22560c6ece8e0ef1',
            },
            files.get_sha1s(keys))

    def test_insert_record_stream_empty(self):
        """Inserting an empty record stream should work."""
        files = self.get_versionedfiles()
        files.insert_record_stream([])

    def assertIdenticalVersionedFile(self, expected, actual):
        """Assert that left and right have the same contents."""
        self.assertEqual(set(actual.keys()), set(expected.keys()))
        actual_parents = actual.get_parent_map(actual.keys())
        if self.graph:
            self.assertEqual(actual_parents, expected.get_parent_map(expected.keys()))
        else:
            for key, parents in actual_parents.items():
                self.assertEqual(None, parents)
        for key in actual.keys():
            actual_text = actual.get_record_stream(
                [key], 'unordered', True).next().get_bytes_as('fulltext')
            expected_text = expected.get_record_stream(
                [key], 'unordered', True).next().get_bytes_as('fulltext')
            self.assertEqual(actual_text, expected_text)

    def test_insert_record_stream_fulltexts(self):
        """Any file should accept a stream of fulltexts."""
        files = self.get_versionedfiles()
        mapper = self.get_mapper()
        source_transport = self.get_transport('source')
        source_transport.mkdir('.')
        # weaves always output fulltexts.
        source = make_versioned_files_factory(WeaveFile, mapper)(
            source_transport)
        self.get_diamond_files(source, trailing_eol=False)
        stream = source.get_record_stream(source.keys(), 'topological',
            False)
        files.insert_record_stream(stream)
        self.assertIdenticalVersionedFile(source, files)

    def test_insert_record_stream_fulltexts_noeol(self):
        """Any file should accept a stream of fulltexts."""
        files = self.get_versionedfiles()
        mapper = self.get_mapper()
        source_transport = self.get_transport('source')
        source_transport.mkdir('.')
        # weaves always output fulltexts.
        source = make_versioned_files_factory(WeaveFile, mapper)(
            source_transport)
        self.get_diamond_files(source, trailing_eol=False)
        stream = source.get_record_stream(source.keys(), 'topological',
            False)
        files.insert_record_stream(stream)
        self.assertIdenticalVersionedFile(source, files)

    def test_insert_record_stream_annotated_knits(self):
        """Any file should accept a stream from plain knits."""
        files = self.get_versionedfiles()
        mapper = self.get_mapper()
        source_transport = self.get_transport('source')
        source_transport.mkdir('.')
        source = make_file_factory(True, mapper)(source_transport)
        self.get_diamond_files(source)
        stream = source.get_record_stream(source.keys(), 'topological',
            False)
        files.insert_record_stream(stream)
        self.assertIdenticalVersionedFile(source, files)

    def test_insert_record_stream_annotated_knits_noeol(self):
        """Any file should accept a stream from plain knits."""
        files = self.get_versionedfiles()
        mapper = self.get_mapper()
        source_transport = self.get_transport('source')
        source_transport.mkdir('.')
        source = make_file_factory(True, mapper)(source_transport)
        self.get_diamond_files(source, trailing_eol=False)
        stream = source.get_record_stream(source.keys(), 'topological',
            False)
        files.insert_record_stream(stream)
        self.assertIdenticalVersionedFile(source, files)

    def test_insert_record_stream_plain_knits(self):
        """Any file should accept a stream from plain knits."""
        files = self.get_versionedfiles()
        mapper = self.get_mapper()
        source_transport = self.get_transport('source')
        source_transport.mkdir('.')
        source = make_file_factory(False, mapper)(source_transport)
        self.get_diamond_files(source)
        stream = source.get_record_stream(source.keys(), 'topological',
            False)
        files.insert_record_stream(stream)
        self.assertIdenticalVersionedFile(source, files)

    def test_insert_record_stream_plain_knits_noeol(self):
        """Any file should accept a stream from plain knits."""
        files = self.get_versionedfiles()
        mapper = self.get_mapper()
        source_transport = self.get_transport('source')
        source_transport.mkdir('.')
        source = make_file_factory(False, mapper)(source_transport)
        self.get_diamond_files(source, trailing_eol=False)
        stream = source.get_record_stream(source.keys(), 'topological',
            False)
        files.insert_record_stream(stream)
        self.assertIdenticalVersionedFile(source, files)

    def test_insert_record_stream_existing_keys(self):
        """Inserting keys already in a file should not error."""
        files = self.get_versionedfiles()
        source = self.get_versionedfiles('source')
        self.get_diamond_files(source)
        # insert some keys into f.
        self.get_diamond_files(files, left_only=True)
        stream = source.get_record_stream(source.keys(), 'topological',
            False)
        files.insert_record_stream(stream)
        self.assertIdenticalVersionedFile(source, files)

    def test_insert_record_stream_missing_keys(self):
        """Inserting a stream with absent keys should raise an error."""
        files = self.get_versionedfiles()
        source = self.get_versionedfiles('source')
        stream = source.get_record_stream([('missing',) * self.key_length],
            'topological', False)
        self.assertRaises(errors.RevisionNotPresent, files.insert_record_stream,
            stream)

    def test_insert_record_stream_out_of_order(self):
        """An out of order stream can either error or work."""
        files = self.get_versionedfiles()
        source = self.get_versionedfiles('source')
        self.get_diamond_files(source)
        if self.key_length == 1:
            origin_keys = [('origin',)]
            end_keys = [('merged',), ('left',)]
            start_keys = [('right',), ('base',)]
        else:
            origin_keys = [('FileA', 'origin'), ('FileB', 'origin')]
            end_keys = [('FileA', 'merged',), ('FileA', 'left',),
                ('FileB', 'merged',), ('FileB', 'left',)]
            start_keys = [('FileA', 'right',), ('FileA', 'base',),
                ('FileB', 'right',), ('FileB', 'base',)]
        origin_entries = source.get_record_stream(origin_keys, 'unordered', False)
        end_entries = source.get_record_stream(end_keys, 'topological', False)
        start_entries = source.get_record_stream(start_keys, 'topological', False)
        entries = chain(origin_entries, end_entries, start_entries)
        try:
            files.insert_record_stream(entries)
        except RevisionNotPresent:
            # Must not have corrupted the file.
            files.check()
        else:
            self.assertIdenticalVersionedFile(source, files)

    def test_insert_record_stream_long_parent_chain_out_of_order(self):
        """An out of order stream can either error or work."""
        if not self.graph:
            raise TestNotApplicable('ancestry info only relevant with graph.')
        # Create a reasonably long chain of records based on each other, where
        # most will be deltas.
        source = self.get_versionedfiles('source')
        parents = ()
        keys = []
        content = [('same same %d\n' % n) for n in range(500)]
        for letter in 'abcdefghijklmnopqrstuvwxyz':
            key = ('key-' + letter,)
            if self.key_length == 2:
                key = ('prefix',) + key
            content.append('content for ' + letter + '\n')
            source.add_lines(key, parents, content)
            keys.append(key)
            parents = (key,)
        # Create a stream of these records, excluding the first record that the
        # rest ultimately depend upon, and insert it into a new vf.
        streams = []
        for key in reversed(keys):
            streams.append(source.get_record_stream([key], 'unordered', False))
        deltas = chain(*streams[:-1])
        files = self.get_versionedfiles()
        try:
            files.insert_record_stream(deltas)
        except RevisionNotPresent:
            # Must not have corrupted the file.
            files.check()
        else:
            # Must only report either just the first key as a missing parent,
            # no key as missing (for nodelta scenarios).
            missing = set(files.get_missing_compression_parent_keys())
            missing.discard(keys[0])
            self.assertEqual(set(), missing)

    def get_knit_delta_source(self):
        """Get a source that can produce a stream with knit delta records,
        regardless of this test's scenario.
        """
        mapper = self.get_mapper()
        source_transport = self.get_transport('source')
        source_transport.mkdir('.')
        source = make_file_factory(False, mapper)(source_transport)
        get_diamond_files(source, self.key_length, trailing_eol=True,
            nograph=False, left_only=False)
        return source

    def test_insert_record_stream_delta_missing_basis_no_corruption(self):
        """Insertion where a needed basis is not included notifies the caller
        of the missing basis.  In the meantime a record missing its basis is
        not added.
        """
        source = self.get_knit_delta_source()
        keys = [self.get_simple_key('origin'), self.get_simple_key('merged')]
        entries = source.get_record_stream(keys, 'unordered', False)
        files = self.get_versionedfiles()
        if self.support_partial_insertion:
            self.assertEqual([],
                list(files.get_missing_compression_parent_keys()))
            files.insert_record_stream(entries)
            missing_bases = files.get_missing_compression_parent_keys()
            self.assertEqual(set([self.get_simple_key('left')]),
                set(missing_bases))
            self.assertEqual(set(keys), set(files.get_parent_map(keys)))
        else:
            self.assertRaises(
                errors.RevisionNotPresent, files.insert_record_stream, entries)
            files.check()

    def test_insert_record_stream_delta_missing_basis_can_be_added_later(self):
        """Insertion where a needed basis is not included notifies the caller
        of the missing basis.  That basis can be added in a second
        insert_record_stream call that does not need to repeat records present
        in the previous stream.  The record(s) that required that basis are
        fully inserted once their basis is no longer missing.
        """
        if not self.support_partial_insertion:
            raise TestNotApplicable(
                'versioned file scenario does not support partial insertion')
        source = self.get_knit_delta_source()
        entries = source.get_record_stream([self.get_simple_key('origin'),
            self.get_simple_key('merged')], 'unordered', False)
        files = self.get_versionedfiles()
        files.insert_record_stream(entries)
        missing_bases = files.get_missing_compression_parent_keys()
        self.assertEqual(set([self.get_simple_key('left')]),
            set(missing_bases))
        # 'merged' is inserted (although a commit of a write group involving
        # this versionedfiles would fail).
        merged_key = self.get_simple_key('merged')
        self.assertEqual(
            [merged_key], files.get_parent_map([merged_key]).keys())
        # Add the full delta closure of the missing records
        missing_entries = source.get_record_stream(
            missing_bases, 'unordered', True)
        files.insert_record_stream(missing_entries)
        # Now 'merged' is fully inserted (and a commit would succeed).
        self.assertEqual([], list(files.get_missing_compression_parent_keys()))
        self.assertEqual(
            [merged_key], files.get_parent_map([merged_key]).keys())
        files.check()

    def test_iter_lines_added_or_present_in_keys(self):
        # test that we get at least an equalset of the lines added by
        # versions in the store.
        # the ordering here is to make a tree so that dumb searches have
        # more changes to muck up.

        class InstrumentedProgress(progress.ProgressTask):

            def __init__(self):
                progress.ProgressTask.__init__(self)
                self.updates = []

            def update(self, msg=None, current=None, total=None):
                self.updates.append((msg, current, total))

        files = self.get_versionedfiles()
        # add a base to get included
        files.add_lines(self.get_simple_key('base'), (), ['base\n'])
        # add a ancestor to be included on one side
        files.add_lines(self.get_simple_key('lancestor'), (), ['lancestor\n'])
        # add a ancestor to be included on the other side
        files.add_lines(self.get_simple_key('rancestor'),
            self.get_parents([self.get_simple_key('base')]), ['rancestor\n'])
        # add a child of rancestor with no eofile-nl
        files.add_lines(self.get_simple_key('child'),
            self.get_parents([self.get_simple_key('rancestor')]),
            ['base\n', 'child\n'])
        # add a child of lancestor and base to join the two roots
        files.add_lines(self.get_simple_key('otherchild'),
            self.get_parents([self.get_simple_key('lancestor'),
                self.get_simple_key('base')]),
            ['base\n', 'lancestor\n', 'otherchild\n'])
        def iter_with_keys(keys, expected):
            # now we need to see what lines are returned, and how often.
            lines = {}
            progress = InstrumentedProgress()
            # iterate over the lines
            for line in files.iter_lines_added_or_present_in_keys(keys,
                pb=progress):
                lines.setdefault(line, 0)
                lines[line] += 1
            if []!= progress.updates:
                self.assertEqual(expected, progress.updates)
            return lines
        lines = iter_with_keys(
            [self.get_simple_key('child'), self.get_simple_key('otherchild')],
            [('Walking content', 0, 2),
             ('Walking content', 1, 2),
             ('Walking content', 2, 2)])
        # we must see child and otherchild
        self.assertTrue(lines[('child\n', self.get_simple_key('child'))] > 0)
        self.assertTrue(
            lines[('otherchild\n', self.get_simple_key('otherchild'))] > 0)
        # we dont care if we got more than that.

        # test all lines
        lines = iter_with_keys(files.keys(),
            [('Walking content', 0, 5),
             ('Walking content', 1, 5),
             ('Walking content', 2, 5),
             ('Walking content', 3, 5),
             ('Walking content', 4, 5),
             ('Walking content', 5, 5)])
        # all lines must be seen at least once
        self.assertTrue(lines[('base\n', self.get_simple_key('base'))] > 0)
        self.assertTrue(
            lines[('lancestor\n', self.get_simple_key('lancestor'))] > 0)
        self.assertTrue(
            lines[('rancestor\n', self.get_simple_key('rancestor'))] > 0)
        self.assertTrue(lines[('child\n', self.get_simple_key('child'))] > 0)
        self.assertTrue(
            lines[('otherchild\n', self.get_simple_key('otherchild'))] > 0)

    def test_make_mpdiffs(self):
        from bzrlib import multiparent
        files = self.get_versionedfiles('source')
        # add texts that should trip the knit maximum delta chain threshold
        # as well as doing parallel chains of data in knits.
        # this is done by two chains of 25 insertions
        files.add_lines(self.get_simple_key('base'), [], ['line\n'])
        files.add_lines(self.get_simple_key('noeol'),
            self.get_parents([self.get_simple_key('base')]), ['line'])
        # detailed eol tests:
        # shared last line with parent no-eol
        files.add_lines(self.get_simple_key('noeolsecond'),
            self.get_parents([self.get_simple_key('noeol')]),
                ['line\n', 'line'])
        # differing last line with parent, both no-eol
        files.add_lines(self.get_simple_key('noeolnotshared'),
            self.get_parents([self.get_simple_key('noeolsecond')]),
                ['line\n', 'phone'])
        # add eol following a noneol parent, change content
        files.add_lines(self.get_simple_key('eol'),
            self.get_parents([self.get_simple_key('noeol')]), ['phone\n'])
        # add eol following a noneol parent, no change content
        files.add_lines(self.get_simple_key('eolline'),
            self.get_parents([self.get_simple_key('noeol')]), ['line\n'])
        # noeol with no parents:
        files.add_lines(self.get_simple_key('noeolbase'), [], ['line'])
        # noeol preceeding its leftmost parent in the output:
        # this is done by making it a merge of two parents with no common
        # anestry: noeolbase and noeol with the
        # later-inserted parent the leftmost.
        files.add_lines(self.get_simple_key('eolbeforefirstparent'),
            self.get_parents([self.get_simple_key('noeolbase'),
                self.get_simple_key('noeol')]),
            ['line'])
        # two identical eol texts
        files.add_lines(self.get_simple_key('noeoldup'),
            self.get_parents([self.get_simple_key('noeol')]), ['line'])
        next_parent = self.get_simple_key('base')
        text_name = 'chain1-'
        text = ['line\n']
        sha1s = {0 :'da6d3141cb4a5e6f464bf6e0518042ddc7bfd079',
                 1 :'45e21ea146a81ea44a821737acdb4f9791c8abe7',
                 2 :'e1f11570edf3e2a070052366c582837a4fe4e9fa',
                 3 :'26b4b8626da827088c514b8f9bbe4ebf181edda1',
                 4 :'e28a5510be25ba84d31121cff00956f9970ae6f6',
                 5 :'d63ec0ce22e11dcf65a931b69255d3ac747a318d',
                 6 :'2c2888d288cb5e1d98009d822fedfe6019c6a4ea',
                 7 :'95c14da9cafbf828e3e74a6f016d87926ba234ab',
                 8 :'779e9a0b28f9f832528d4b21e17e168c67697272',
                 9 :'1f8ff4e5c6ff78ac106fcfe6b1e8cb8740ff9a8f',
                 10:'131a2ae712cf51ed62f143e3fbac3d4206c25a05',
                 11:'c5a9d6f520d2515e1ec401a8f8a67e6c3c89f199',
                 12:'31a2286267f24d8bedaa43355f8ad7129509ea85',
                 13:'dc2a7fe80e8ec5cae920973973a8ee28b2da5e0a',
                 14:'2c4b1736566b8ca6051e668de68650686a3922f2',
                 15:'5912e4ecd9b0c07be4d013e7e2bdcf9323276cde',
                 16:'b0d2e18d3559a00580f6b49804c23fea500feab3',
                 17:'8e1d43ad72f7562d7cb8f57ee584e20eb1a69fc7',
                 18:'5cf64a3459ae28efa60239e44b20312d25b253f3',
                 19:'1ebed371807ba5935958ad0884595126e8c4e823',
                 20:'2aa62a8b06fb3b3b892a3292a068ade69d5ee0d3',
                 21:'01edc447978004f6e4e962b417a4ae1955b6fe5d',
                 22:'d8d8dc49c4bf0bab401e0298bb5ad827768618bb',
                 23:'c21f62b1c482862983a8ffb2b0c64b3451876e3f',
                 24:'c0593fe795e00dff6b3c0fe857a074364d5f04fc',
                 25:'dd1a1cf2ba9cc225c3aff729953e6364bf1d1855',
                 }
        for depth in range(26):
            new_version = self.get_simple_key(text_name + '%s' % depth)
            text = text + ['line\n']
            files.add_lines(new_version, self.get_parents([next_parent]), text)
            next_parent = new_version
        next_parent = self.get_simple_key('base')
        text_name = 'chain2-'
        text = ['line\n']
        for depth in range(26):
            new_version = self.get_simple_key(text_name + '%s' % depth)
            text = text + ['line\n']
            files.add_lines(new_version, self.get_parents([next_parent]), text)
            next_parent = new_version
        target = self.get_versionedfiles('target')
        for key in multiparent.topo_iter_keys(files, files.keys()):
            mpdiff = files.make_mpdiffs([key])[0]
            parents = files.get_parent_map([key])[key] or []
            target.add_mpdiffs(
                [(key, parents, files.get_sha1s([key])[key], mpdiff)])
            self.assertEqualDiff(
                files.get_record_stream([key], 'unordered',
                    True).next().get_bytes_as('fulltext'),
                target.get_record_stream([key], 'unordered',
                    True).next().get_bytes_as('fulltext')
                )

    def test_keys(self):
        # While use is discouraged, versions() is still needed by aspects of
        # bzr.
        files = self.get_versionedfiles()
        self.assertEqual(set(), set(files.keys()))
        if self.key_length == 1:
            key = ('foo',)
        else:
            key = ('foo', 'bar',)
        files.add_lines(key, (), [])
        self.assertEqual(set([key]), set(files.keys()))


class VirtualVersionedFilesTests(TestCase):
    """Basic tests for the VirtualVersionedFiles implementations."""

    def _get_parent_map(self, keys):
        ret = {}
        for k in keys:
            if k in self._parent_map:
                ret[k] = self._parent_map[k]
        return ret

    def setUp(self):
        super(VirtualVersionedFilesTests, self).setUp()
        self._lines = {}
        self._parent_map = {}
        self.texts = VirtualVersionedFiles(self._get_parent_map,
                                           self._lines.get)

    def test_add_lines(self):
        self.assertRaises(NotImplementedError,
                self.texts.add_lines, "foo", [], [])

    def test_add_mpdiffs(self):
        self.assertRaises(NotImplementedError,
                self.texts.add_mpdiffs, [])

    def test_check_noerrors(self):
        self.texts.check()

    def test_insert_record_stream(self):
        self.assertRaises(NotImplementedError, self.texts.insert_record_stream,
                          [])

    def test_get_sha1s_nonexistent(self):
        self.assertEqual({}, self.texts.get_sha1s([("NONEXISTENT",)]))

    def test_get_sha1s(self):
        self._lines["key"] = ["dataline1", "dataline2"]
        self.assertEqual({("key",): osutils.sha_strings(self._lines["key"])},
                           self.texts.get_sha1s([("key",)]))

    def test_get_parent_map(self):
        self._parent_map = {"G": ("A", "B")}
        self.assertEqual({("G",): (("A",),("B",))},
                          self.texts.get_parent_map([("G",), ("L",)]))

    def test_get_record_stream(self):
        self._lines["A"] = ["FOO", "BAR"]
        it = self.texts.get_record_stream([("A",)], "unordered", True)
        record = it.next()
        self.assertEqual("chunked", record.storage_kind)
        self.assertEqual("FOOBAR", record.get_bytes_as("fulltext"))
        self.assertEqual(["FOO", "BAR"], record.get_bytes_as("chunked"))

    def test_get_record_stream_absent(self):
        it = self.texts.get_record_stream([("A",)], "unordered", True)
        record = it.next()
        self.assertEqual("absent", record.storage_kind)

    def test_iter_lines_added_or_present_in_keys(self):
        self._lines["A"] = ["FOO", "BAR"]
        self._lines["B"] = ["HEY"]
        self._lines["C"] = ["Alberta"]
        it = self.texts.iter_lines_added_or_present_in_keys([("A",), ("B",)])
        self.assertEqual(sorted([("FOO", "A"), ("BAR", "A"), ("HEY", "B")]),
            sorted(list(it)))


class TestOrderingVersionedFilesDecorator(TestCaseWithMemoryTransport):

    def get_ordering_vf(self, key_priority):
        builder = self.make_branch_builder('test')
        builder.start_series()
        builder.build_snapshot('A', None, [
            ('add', ('', 'TREE_ROOT', 'directory', None))])
        builder.build_snapshot('B', ['A'], [])
        builder.build_snapshot('C', ['B'], [])
        builder.build_snapshot('D', ['C'], [])
        builder.finish_series()
        b = builder.get_branch()
        b.lock_read()
        self.addCleanup(b.unlock)
        vf = b.repository.inventories
        return versionedfile.OrderingVersionedFilesDecorator(vf, key_priority)

    def test_get_empty(self):
        vf = self.get_ordering_vf({})
        self.assertEqual([], vf.calls)

    def test_get_record_stream_topological(self):
        vf = self.get_ordering_vf({('A',): 3, ('B',): 2, ('C',): 4, ('D',): 1})
        request_keys = [('B',), ('C',), ('D',), ('A',)]
        keys = [r.key for r in vf.get_record_stream(request_keys,
                                    'topological', False)]
        # We should have gotten the keys in topological order
        self.assertEqual([('A',), ('B',), ('C',), ('D',)], keys)
        # And recorded that the request was made
        self.assertEqual([('get_record_stream', request_keys, 'topological',
                           False)], vf.calls)

    def test_get_record_stream_ordered(self):
        vf = self.get_ordering_vf({('A',): 3, ('B',): 2, ('C',): 4, ('D',): 1})
        request_keys = [('B',), ('C',), ('D',), ('A',)]
        keys = [r.key for r in vf.get_record_stream(request_keys,
                                   'unordered', False)]
        # They should be returned based on their priority
        self.assertEqual([('D',), ('B',), ('A',), ('C',)], keys)
        # And the request recorded
        self.assertEqual([('get_record_stream', request_keys, 'unordered',
                           False)], vf.calls)

    def test_get_record_stream_implicit_order(self):
        vf = self.get_ordering_vf({('B',): 2, ('D',): 1})
        request_keys = [('B',), ('C',), ('D',), ('A',)]
        keys = [r.key for r in vf.get_record_stream(request_keys,
                                   'unordered', False)]
        # A and C are not in the map, so they get sorted to the front. A comes
        # before C alphabetically, so it comes back first
        self.assertEqual([('A',), ('C',), ('D',), ('B',)], keys)
        # And the request recorded
        self.assertEqual([('get_record_stream', request_keys, 'unordered',
                           False)], vf.calls)