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
|
# Copyright (C) 2005, 2006, 2007 Canonical Ltd
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
"""Knit versionedfile implementation.
A knit is a versioned file implementation that supports efficient append only
updates.
Knit file layout:
lifeless: the data file is made up of "delta records". each delta record has a delta header
that contains; (1) a version id, (2) the size of the delta (in lines), and (3) the digest of
the -expanded data- (ie, the delta applied to the parent). the delta also ends with a
end-marker; simply "end VERSION"
delta can be line or full contents.a
... the 8's there are the index number of the annotation.
version robertc@robertcollins.net-20051003014215-ee2990904cc4c7ad 7 c7d23b2a5bd6ca00e8e266cec0ec228158ee9f9e
59,59,3
8
8 if ie.executable:
8 e.set('executable', 'yes')
130,130,2
8 if elt.get('executable') == 'yes':
8 ie.executable = True
end robertc@robertcollins.net-20051003014215-ee2990904cc4c7ad
whats in an index:
09:33 < jrydberg> lifeless: each index is made up of a tuple of; version id, options, position, size, parents
09:33 < jrydberg> lifeless: the parents are currently dictionary compressed
09:33 < jrydberg> lifeless: (meaning it currently does not support ghosts)
09:33 < lifeless> right
09:33 < jrydberg> lifeless: the position and size is the range in the data file
so the index sequence is the dictionary compressed sequence number used
in the deltas to provide line annotation
"""
# TODOS:
# 10:16 < lifeless> make partial index writes safe
# 10:16 < lifeless> implement 'knit.check()' like weave.check()
# 10:17 < lifeless> record known ghosts so we can detect when they are filled in rather than the current 'reweave
# always' approach.
# move sha1 out of the content so that join is faster at verifying parents
# record content length ?
from copy import copy
from cStringIO import StringIO
from itertools import izip, chain
import operator
import os
import sys
import warnings
from zlib import Z_DEFAULT_COMPRESSION
import bzrlib
from bzrlib.lazy_import import lazy_import
lazy_import(globals(), """
from bzrlib import (
annotate,
lru_cache,
pack,
trace,
)
""")
from bzrlib import (
cache_utf8,
debug,
diff,
errors,
osutils,
patiencediff,
progress,
merge,
ui,
)
from bzrlib.errors import (
FileExists,
NoSuchFile,
KnitError,
InvalidRevisionId,
KnitCorrupt,
KnitHeaderError,
RevisionNotPresent,
RevisionAlreadyPresent,
)
from bzrlib.tuned_gzip import GzipFile, bytes_to_gzip
from bzrlib.osutils import (
contains_whitespace,
contains_linebreaks,
sha_string,
sha_strings,
)
from bzrlib.symbol_versioning import DEPRECATED_PARAMETER, deprecated_passed
from bzrlib.tsort import topo_sort
import bzrlib.ui
import bzrlib.weave
from bzrlib.versionedfile import VersionedFile, InterVersionedFile
# TODO: Split out code specific to this format into an associated object.
# TODO: Can we put in some kind of value to check that the index and data
# files belong together?
# TODO: accommodate binaries, perhaps by storing a byte count
# TODO: function to check whole file
# TODO: atomically append data, then measure backwards from the cursor
# position after writing to work out where it was located. we may need to
# bypass python file buffering.
DATA_SUFFIX = '.knit'
INDEX_SUFFIX = '.kndx'
class KnitContent(object):
"""Content of a knit version to which deltas can be applied."""
def annotate(self):
"""Return a list of (origin, text) tuples."""
return list(self.annotate_iter())
def apply_delta(self, delta, new_version_id):
"""Apply delta to this object to become new_version_id."""
raise NotImplementedError(self.apply_delta)
def line_delta_iter(self, new_lines):
"""Generate line-based delta from this content to new_lines."""
new_texts = new_lines.text()
old_texts = self.text()
s = patiencediff.PatienceSequenceMatcher(None, old_texts, new_texts)
for tag, i1, i2, j1, j2 in s.get_opcodes():
if tag == 'equal':
continue
# ofrom, oto, length, data
yield i1, i2, j2 - j1, new_lines._lines[j1:j2]
def line_delta(self, new_lines):
return list(self.line_delta_iter(new_lines))
@staticmethod
def get_line_delta_blocks(knit_delta, source, target):
"""Extract SequenceMatcher.get_matching_blocks() from a knit delta"""
target_len = len(target)
s_pos = 0
t_pos = 0
for s_begin, s_end, t_len, new_text in knit_delta:
true_n = s_begin - s_pos
n = true_n
if n > 0:
# knit deltas do not provide reliable info about whether the
# last line of a file matches, due to eol handling.
if source[s_pos + n -1] != target[t_pos + n -1]:
n-=1
if n > 0:
yield s_pos, t_pos, n
t_pos += t_len + true_n
s_pos = s_end
n = target_len - t_pos
if n > 0:
if source[s_pos + n -1] != target[t_pos + n -1]:
n-=1
if n > 0:
yield s_pos, t_pos, n
yield s_pos + (target_len - t_pos), target_len, 0
class AnnotatedKnitContent(KnitContent):
"""Annotated content."""
def __init__(self, lines):
self._lines = lines
def annotate_iter(self):
"""Yield tuples of (origin, text) for each content line."""
return iter(self._lines)
def apply_delta(self, delta, new_version_id):
"""Apply delta to this object to become new_version_id."""
offset = 0
lines = self._lines
for start, end, count, delta_lines in delta:
lines[offset+start:offset+end] = delta_lines
offset = offset + (start - end) + count
def strip_last_line_newline(self):
line = self._lines[-1][1].rstrip('\n')
self._lines[-1] = (self._lines[-1][0], line)
def text(self):
try:
return [text for origin, text in self._lines]
except ValueError, e:
# most commonly (only?) caused by the internal form of the knit
# missing annotation information because of a bug - see thread
# around 20071015
raise KnitCorrupt(self,
"line in annotated knit missing annotation information: %s"
% (e,))
def copy(self):
return AnnotatedKnitContent(self._lines[:])
class PlainKnitContent(KnitContent):
"""Unannotated content.
When annotate[_iter] is called on this content, the same version is reported
for all lines. Generally, annotate[_iter] is not useful on PlainKnitContent
objects.
"""
def __init__(self, lines, version_id):
self._lines = lines
self._version_id = version_id
def annotate_iter(self):
"""Yield tuples of (origin, text) for each content line."""
for line in self._lines:
yield self._version_id, line
def apply_delta(self, delta, new_version_id):
"""Apply delta to this object to become new_version_id."""
offset = 0
lines = self._lines
for start, end, count, delta_lines in delta:
lines[offset+start:offset+end] = delta_lines
offset = offset + (start - end) + count
self._version_id = new_version_id
def copy(self):
return PlainKnitContent(self._lines[:], self._version_id)
def strip_last_line_newline(self):
self._lines[-1] = self._lines[-1].rstrip('\n')
def text(self):
return self._lines
class KnitAnnotateFactory(object):
"""Factory for creating annotated Content objects."""
annotated = True
def make(self, lines, version_id):
num_lines = len(lines)
return AnnotatedKnitContent(zip([version_id] * num_lines, lines))
def parse_fulltext(self, content, version_id):
"""Convert fulltext to internal representation
fulltext content is of the format
revid(utf8) plaintext\n
internal representation is of the format:
(revid, plaintext)
"""
# TODO: jam 20070209 The tests expect this to be returned as tuples,
# but the code itself doesn't really depend on that.
# Figure out a way to not require the overhead of turning the
# list back into tuples.
lines = [tuple(line.split(' ', 1)) for line in content]
return AnnotatedKnitContent(lines)
def parse_line_delta_iter(self, lines):
return iter(self.parse_line_delta(lines))
def parse_line_delta(self, lines, version_id, plain=False):
"""Convert a line based delta into internal representation.
line delta is in the form of:
intstart intend intcount
1..count lines:
revid(utf8) newline\n
internal representation is
(start, end, count, [1..count tuples (revid, newline)])
:param plain: If True, the lines are returned as a plain
list without annotations, not as a list of (origin, content) tuples, i.e.
(start, end, count, [1..count newline])
"""
result = []
lines = iter(lines)
next = lines.next
cache = {}
def cache_and_return(line):
origin, text = line.split(' ', 1)
return cache.setdefault(origin, origin), text
# walk through the lines parsing.
# Note that the plain test is explicitly pulled out of the
# loop to minimise any performance impact
if plain:
for header in lines:
start, end, count = [int(n) for n in header.split(',')]
contents = [next().split(' ', 1)[1] for i in xrange(count)]
result.append((start, end, count, contents))
else:
for header in lines:
start, end, count = [int(n) for n in header.split(',')]
contents = [tuple(next().split(' ', 1)) for i in xrange(count)]
result.append((start, end, count, contents))
return result
def get_fulltext_content(self, lines):
"""Extract just the content lines from a fulltext."""
return (line.split(' ', 1)[1] for line in lines)
def get_linedelta_content(self, lines):
"""Extract just the content from a line delta.
This doesn't return all of the extra information stored in a delta.
Only the actual content lines.
"""
lines = iter(lines)
next = lines.next
for header in lines:
header = header.split(',')
count = int(header[2])
for i in xrange(count):
origin, text = next().split(' ', 1)
yield text
def lower_fulltext(self, content):
"""convert a fulltext content record into a serializable form.
see parse_fulltext which this inverts.
"""
# TODO: jam 20070209 We only do the caching thing to make sure that
# the origin is a valid utf-8 line, eventually we could remove it
return ['%s %s' % (o, t) for o, t in content._lines]
def lower_line_delta(self, delta):
"""convert a delta into a serializable form.
See parse_line_delta which this inverts.
"""
# TODO: jam 20070209 We only do the caching thing to make sure that
# the origin is a valid utf-8 line, eventually we could remove it
out = []
for start, end, c, lines in delta:
out.append('%d,%d,%d\n' % (start, end, c))
out.extend(origin + ' ' + text
for origin, text in lines)
return out
def annotate_iter(self, knit, version_id):
content = knit._get_content(version_id)
return content.annotate_iter()
class KnitPlainFactory(object):
"""Factory for creating plain Content objects."""
annotated = False
def make(self, lines, version_id):
return PlainKnitContent(lines, version_id)
def parse_fulltext(self, content, version_id):
"""This parses an unannotated fulltext.
Note that this is not a noop - the internal representation
has (versionid, line) - its just a constant versionid.
"""
return self.make(content, version_id)
def parse_line_delta_iter(self, lines, version_id):
cur = 0
num_lines = len(lines)
while cur < num_lines:
header = lines[cur]
cur += 1
start, end, c = [int(n) for n in header.split(',')]
yield start, end, c, lines[cur:cur+c]
cur += c
def parse_line_delta(self, lines, version_id):
return list(self.parse_line_delta_iter(lines, version_id))
def get_fulltext_content(self, lines):
"""Extract just the content lines from a fulltext."""
return iter(lines)
def get_linedelta_content(self, lines):
"""Extract just the content from a line delta.
This doesn't return all of the extra information stored in a delta.
Only the actual content lines.
"""
lines = iter(lines)
next = lines.next
for header in lines:
header = header.split(',')
count = int(header[2])
for i in xrange(count):
yield next()
def lower_fulltext(self, content):
return content.text()
def lower_line_delta(self, delta):
out = []
for start, end, c, lines in delta:
out.append('%d,%d,%d\n' % (start, end, c))
out.extend(lines)
return out
def annotate_iter(self, knit, version_id):
return annotate_knit(knit, version_id)
def make_empty_knit(transport, relpath):
"""Construct a empty knit at the specified location."""
k = KnitVersionedFile(transport, relpath, 'w', KnitPlainFactory)
class KnitVersionedFile(VersionedFile):
"""Weave-like structure with faster random access.
A knit stores a number of texts and a summary of the relationships
between them. Texts are identified by a string version-id. Texts
are normally stored and retrieved as a series of lines, but can
also be passed as single strings.
Lines are stored with the trailing newline (if any) included, to
avoid special cases for files with no final newline. Lines are
composed of 8-bit characters, not unicode. The combination of
these approaches should mean any 'binary' file can be safely
stored and retrieved.
"""
def __init__(self, relpath, transport, file_mode=None, access_mode=None,
factory=None, delta=True, create=False, create_parent_dir=False,
delay_create=False, dir_mode=None, index=None, access_method=None):
"""Construct a knit at location specified by relpath.
:param create: If not True, only open an existing knit.
:param create_parent_dir: If True, create the parent directory if
creating the file fails. (This is used for stores with
hash-prefixes that may not exist yet)
:param delay_create: The calling code is aware that the knit won't
actually be created until the first data is stored.
:param index: An index to use for the knit.
"""
if access_mode is None:
access_mode = 'w'
super(KnitVersionedFile, self).__init__(access_mode)
assert access_mode in ('r', 'w'), "invalid mode specified %r" % access_mode
self.transport = transport
self.filename = relpath
self.factory = factory or KnitAnnotateFactory()
self.writable = (access_mode == 'w')
self.delta = delta
self._max_delta_chain = 200
if index is None:
self._index = _KnitIndex(transport, relpath + INDEX_SUFFIX,
access_mode, create=create, file_mode=file_mode,
create_parent_dir=create_parent_dir, delay_create=delay_create,
dir_mode=dir_mode)
else:
self._index = index
if access_method is None:
_access = _KnitAccess(transport, relpath + DATA_SUFFIX, file_mode, dir_mode,
((create and not len(self)) and delay_create), create_parent_dir)
else:
_access = access_method
if create and not len(self) and not delay_create:
_access.create()
self._data = _KnitData(_access)
def __repr__(self):
return '%s(%s)' % (self.__class__.__name__,
self.transport.abspath(self.filename))
def _check_should_delta(self, first_parents):
"""Iterate back through the parent listing, looking for a fulltext.
This is used when we want to decide whether to add a delta or a new
fulltext. It searches for _max_delta_chain parents. When it finds a
fulltext parent, it sees if the total size of the deltas leading up to
it is large enough to indicate that we want a new full text anyway.
Return True if we should create a new delta, False if we should use a
full text.
"""
delta_size = 0
fulltext_size = None
delta_parents = first_parents
for count in xrange(self._max_delta_chain):
parent = delta_parents[0]
method = self._index.get_method(parent)
index, pos, size = self._index.get_position(parent)
if method == 'fulltext':
fulltext_size = size
break
delta_size += size
delta_parents = self._index.get_parents(parent)
else:
# We couldn't find a fulltext, so we must create a new one
return False
return fulltext_size > delta_size
def _add_raw_records(self, records, data):
"""Add all the records 'records' with data pre-joined in 'data'.
:param records: A list of tuples(version_id, options, parents, size).
:param data: The data for the records. When it is written, the records
are adjusted to have pos pointing into data by the sum of
the preceding records sizes.
"""
# write all the data
raw_record_sizes = [record[3] for record in records]
positions = self._data.add_raw_records(raw_record_sizes, data)
offset = 0
index_entries = []
for (version_id, options, parents, size), access_memo in zip(
records, positions):
index_entries.append((version_id, options, access_memo, parents))
if self._data._do_cache:
self._data._cache[version_id] = data[offset:offset+size]
offset += size
self._index.add_versions(index_entries)
def enable_cache(self):
"""Start caching data for this knit"""
self._data.enable_cache()
def clear_cache(self):
"""Clear the data cache only."""
self._data.clear_cache()
def copy_to(self, name, transport):
"""See VersionedFile.copy_to()."""
# copy the current index to a temp index to avoid racing with local
# writes
transport.put_file_non_atomic(name + INDEX_SUFFIX + '.tmp',
self.transport.get(self._index._filename))
# copy the data file
f = self._data._open_file()
try:
transport.put_file(name + DATA_SUFFIX, f)
finally:
f.close()
# move the copied index into place
transport.move(name + INDEX_SUFFIX + '.tmp', name + INDEX_SUFFIX)
def create_empty(self, name, transport, mode=None):
return KnitVersionedFile(name, transport, factory=self.factory,
delta=self.delta, create=True)
def get_data_stream(self, required_versions):
"""Get a data stream for the specified versions.
Versions may be returned in any order, not necessarily the order
specified. They are returned in a partial order by compression
parent, so that the deltas can be applied as the data stream is
inserted; however note that compression parents will not be sent
unless they were specifically requested, as the client may already
have them.
:param required_versions: The exact set of versions to be extracted.
Unlike some other knit methods, this is not used to generate a
transitive closure, rather it is used precisely as given.
:returns: format_signature, list of (version, options, length, parents),
reader_callable.
"""
required_version_set = frozenset(required_versions)
version_index = {}
# list of revisions that can just be sent without waiting for their
# compression parent
ready_to_send = []
# map from revision to the children based on it
deferred = {}
# first, read all relevant index data, enough to sort into the right
# order to return
for version_id in required_versions:
options = self._index.get_options(version_id)
parents = self._index.get_parents_with_ghosts(version_id)
index_memo = self._index.get_position(version_id)
version_index[version_id] = (index_memo, options, parents)
if ('line-delta' in options
and parents[0] in required_version_set):
# must wait until the parent has been sent
deferred.setdefault(parents[0], []). \
append(version_id)
else:
# either a fulltext, or a delta whose parent the client did
# not ask for and presumably already has
ready_to_send.append(version_id)
# build a list of results to return, plus instructions for data to
# read from the file
copy_queue_records = []
temp_version_list = []
while ready_to_send:
# XXX: pushing and popping lists may be a bit inefficient
version_id = ready_to_send.pop(0)
(index_memo, options, parents) = version_index[version_id]
copy_queue_records.append((version_id, index_memo))
none, data_pos, data_size = index_memo
temp_version_list.append((version_id, options, data_size,
parents))
if version_id in deferred:
# now we can send all the children of this revision - we could
# put them in anywhere, but we hope that sending them soon
# after the fulltext will give good locality in the receiver
ready_to_send[:0] = deferred.pop(version_id)
assert len(deferred) == 0, \
"Still have compressed child versions waiting to be sent"
# XXX: The stream format is such that we cannot stream it - we have to
# know the length of all the data a-priori.
raw_datum = []
result_version_list = []
for (version_id, raw_data), \
(version_id2, options, _, parents) in \
izip(self._data.read_records_iter_raw(copy_queue_records),
temp_version_list):
assert version_id == version_id2, \
'logic error, inconsistent results'
raw_datum.append(raw_data)
result_version_list.append(
(version_id, options, len(raw_data), parents))
# provide a callback to get data incrementally.
pseudo_file = StringIO(''.join(raw_datum))
def read(length):
if length is None:
return pseudo_file.read()
else:
return pseudo_file.read(length)
return (self.get_format_signature(), result_version_list, read)
def _extract_blocks(self, version_id, source, target):
if self._index.get_method(version_id) != 'line-delta':
return None
parent, sha1, noeol, delta = self.get_delta(version_id)
return KnitContent.get_line_delta_blocks(delta, source, target)
def get_delta(self, version_id):
"""Get a delta for constructing version from some other version."""
self.check_not_reserved_id(version_id)
parents = self.get_parents(version_id)
if len(parents):
parent = parents[0]
else:
parent = None
index_memo = self._index.get_position(version_id)
data, sha1 = self._data.read_records(((version_id, index_memo),))[version_id]
noeol = 'no-eol' in self._index.get_options(version_id)
if 'fulltext' == self._index.get_method(version_id):
new_content = self.factory.parse_fulltext(data, version_id)
if parent is not None:
reference_content = self._get_content(parent)
old_texts = reference_content.text()
else:
old_texts = []
new_texts = new_content.text()
delta_seq = patiencediff.PatienceSequenceMatcher(None, old_texts,
new_texts)
return parent, sha1, noeol, self._make_line_delta(delta_seq, new_content)
else:
delta = self.factory.parse_line_delta(data, version_id)
return parent, sha1, noeol, delta
def get_format_signature(self):
"""See VersionedFile.get_format_signature()."""
if self.factory.annotated:
annotated_part = "annotated"
else:
annotated_part = "plain"
return "knit-%s" % (annotated_part,)
def get_graph_with_ghosts(self):
"""See VersionedFile.get_graph_with_ghosts()."""
graph_items = self._index.get_graph()
return dict(graph_items)
def get_sha1(self, version_id):
return self.get_sha1s([version_id])[0]
def get_sha1s(self, version_ids):
"""See VersionedFile.get_sha1()."""
record_map = self._get_record_map(version_ids)
# record entry 2 is the 'digest'.
return [record_map[v][2] for v in version_ids]
@staticmethod
def get_suffixes():
"""See VersionedFile.get_suffixes()."""
return [DATA_SUFFIX, INDEX_SUFFIX]
def has_ghost(self, version_id):
"""True if there is a ghost reference in the file to version_id."""
# maybe we have it
if self.has_version(version_id):
return False
# optimisable if needed by memoising the _ghosts set.
items = self._index.get_graph()
for node, parents in items:
for parent in parents:
if parent not in self._index._cache:
if parent == version_id:
return True
return False
def insert_data_stream(self, (format, data_list, reader_callable)):
"""Insert knit records from a data stream into this knit.
If a version in the stream is already present in this knit, it will not
be inserted a second time. It will be checked for consistency with the
stored version however, and may cause a KnitCorrupt error to be raised
if the data in the stream disagrees with the already stored data.
:seealso: get_data_stream
"""
if format != self.get_format_signature():
if 'knit' in debug.debug_flags:
trace.mutter(
'incompatible format signature inserting to %r', self)
source = self._knit_from_datastream(
(format, data_list, reader_callable))
self.join(source)
return
for version_id, options, length, parents in data_list:
if self.has_version(version_id):
# First check: the list of parents.
my_parents = self.get_parents_with_ghosts(version_id)
if my_parents != parents:
# XXX: KnitCorrupt is not quite the right exception here.
raise KnitCorrupt(
self.filename,
'parents list %r from data stream does not match '
'already recorded parents %r for %s'
% (parents, my_parents, version_id))
# Also check the SHA-1 of the fulltext this content will
# produce.
raw_data = reader_callable(length)
my_fulltext_sha1 = self.get_sha1(version_id)
df, rec = self._data._parse_record_header(version_id, raw_data)
stream_fulltext_sha1 = rec[3]
if my_fulltext_sha1 != stream_fulltext_sha1:
# Actually, we don't know if it's this knit that's corrupt,
# or the data stream we're trying to insert.
raise KnitCorrupt(
self.filename, 'sha-1 does not match %s' % version_id)
else:
if 'line-delta' in options:
# Make sure that this knit record is actually useful: a
# line-delta is no use unless we have its parent.
# Fetching from a broken repository with this problem
# shouldn't break the target repository.
#
# See https://bugs.launchpad.net/bzr/+bug/164443
if not self._index.has_version(parents[0]):
raise KnitCorrupt(
self.filename,
'line-delta from stream '
'for version %s '
'references '
'missing parent %s\n'
'Try running "bzr check" '
'on the source repository, and "bzr reconcile" '
'if necessary.' %
(version_id, parents[0]))
self._add_raw_records(
[(version_id, options, parents, length)],
reader_callable(length))
def _knit_from_datastream(self, (format, data_list, reader_callable)):
"""Create a knit object from a data stream.
This method exists to allow conversion of data streams that do not
match the signature of this knit. Generally it will be slower and use
more memory to use this method to insert data, but it will work.
:seealso: get_data_stream for details on datastreams.
:return: A knit versioned file which can be used to join the datastream
into self.
"""
if format == "knit-plain":
factory = KnitPlainFactory()
elif format == "knit-annotated":
factory = KnitAnnotateFactory()
else:
raise errors.KnitDataStreamUnknown(format)
index = _StreamIndex(data_list)
access = _StreamAccess(reader_callable, index, self, factory)
return KnitVersionedFile(self.filename, self.transport,
factory=factory, index=index, access_method=access)
def versions(self):
"""See VersionedFile.versions."""
if 'evil' in debug.debug_flags:
trace.mutter_callsite(2, "versions scales with size of history")
return self._index.get_versions()
def has_version(self, version_id):
"""See VersionedFile.has_version."""
if 'evil' in debug.debug_flags:
trace.mutter_callsite(2, "has_version is a LBYL scenario")
return self._index.has_version(version_id)
__contains__ = has_version
def _merge_annotations(self, content, parents, parent_texts={},
delta=None, annotated=None,
left_matching_blocks=None):
"""Merge annotations for content. This is done by comparing
the annotations based on changed to the text.
"""
if left_matching_blocks is not None:
delta_seq = diff._PrematchedMatcher(left_matching_blocks)
else:
delta_seq = None
if annotated:
for parent_id in parents:
merge_content = self._get_content(parent_id, parent_texts)
if (parent_id == parents[0] and delta_seq is not None):
seq = delta_seq
else:
seq = patiencediff.PatienceSequenceMatcher(
None, merge_content.text(), content.text())
for i, j, n in seq.get_matching_blocks():
if n == 0:
continue
# this appears to copy (origin, text) pairs across to the
# new content for any line that matches the last-checked
# parent.
content._lines[j:j+n] = merge_content._lines[i:i+n]
if delta:
if delta_seq is None:
reference_content = self._get_content(parents[0], parent_texts)
new_texts = content.text()
old_texts = reference_content.text()
delta_seq = patiencediff.PatienceSequenceMatcher(
None, old_texts, new_texts)
return self._make_line_delta(delta_seq, content)
def _make_line_delta(self, delta_seq, new_content):
"""Generate a line delta from delta_seq and new_content."""
diff_hunks = []
for op in delta_seq.get_opcodes():
if op[0] == 'equal':
continue
diff_hunks.append((op[1], op[2], op[4]-op[3], new_content._lines[op[3]:op[4]]))
return diff_hunks
def _get_components_positions(self, version_ids):
"""Produce a map of position data for the components of versions.
This data is intended to be used for retrieving the knit records.
A dict of version_id to (method, data_pos, data_size, next) is
returned.
method is the way referenced data should be applied.
data_pos is the position of the data in the knit.
data_size is the size of the data in the knit.
next is the build-parent of the version, or None for fulltexts.
"""
component_data = {}
for version_id in version_ids:
cursor = version_id
while cursor is not None and cursor not in component_data:
method = self._index.get_method(cursor)
if method == 'fulltext':
next = None
else:
next = self.get_parents_with_ghosts(cursor)[0]
index_memo = self._index.get_position(cursor)
component_data[cursor] = (method, index_memo, next)
cursor = next
return component_data
def _get_content(self, version_id, parent_texts={}):
"""Returns a content object that makes up the specified
version."""
cached_version = parent_texts.get(version_id, None)
if cached_version is not None:
if not self.has_version(version_id):
raise RevisionNotPresent(version_id, self.filename)
return cached_version
text_map, contents_map = self._get_content_maps([version_id])
return contents_map[version_id]
def _check_versions_present(self, version_ids):
"""Check that all specified versions are present."""
self._index.check_versions_present(version_ids)
def _add_lines_with_ghosts(self, version_id, parents, lines, parent_texts,
nostore_sha, random_id, check_content):
"""See VersionedFile.add_lines_with_ghosts()."""
self._check_add(version_id, lines, random_id, check_content)
return self._add(version_id, lines, parents, self.delta,
parent_texts, None, nostore_sha, random_id)
def _add_lines(self, version_id, parents, lines, parent_texts,
left_matching_blocks, nostore_sha, random_id, check_content):
"""See VersionedFile.add_lines."""
self._check_add(version_id, lines, random_id, check_content)
self._check_versions_present(parents)
return self._add(version_id, lines[:], parents, self.delta,
parent_texts, left_matching_blocks, nostore_sha, random_id)
def _check_add(self, version_id, lines, random_id, check_content):
"""check that version_id and lines are safe to add."""
if contains_whitespace(version_id):
raise InvalidRevisionId(version_id, self.filename)
self.check_not_reserved_id(version_id)
# Technically this could be avoided if we are happy to allow duplicate
# id insertion when other things than bzr core insert texts, but it
# seems useful for folk using the knit api directly to have some safety
# blanket that we can disable.
if not random_id and self.has_version(version_id):
raise RevisionAlreadyPresent(version_id, self.filename)
if check_content:
self._check_lines_not_unicode(lines)
self._check_lines_are_lines(lines)
def _add(self, version_id, lines, parents, delta, parent_texts,
left_matching_blocks, nostore_sha, random_id):
"""Add a set of lines on top of version specified by parents.
If delta is true, compress the text as a line-delta against
the first parent.
Any versions not present will be converted into ghosts.
"""
# first thing, if the content is something we don't need to store, find
# that out.
line_bytes = ''.join(lines)
digest = sha_string(line_bytes)
if nostore_sha == digest:
raise errors.ExistingContent
present_parents = []
if parent_texts is None:
parent_texts = {}
for parent in parents:
if self.has_version(parent):
present_parents.append(parent)
# can only compress against the left most present parent.
if (delta and
(len(present_parents) == 0 or
present_parents[0] != parents[0])):
delta = False
text_length = len(line_bytes)
options = []
if lines:
if lines[-1][-1] != '\n':
# copy the contents of lines.
lines = lines[:]
options.append('no-eol')
lines[-1] = lines[-1] + '\n'
line_bytes += '\n'
if delta:
# To speed the extract of texts the delta chain is limited
# to a fixed number of deltas. This should minimize both
# I/O and the time spend applying deltas.
delta = self._check_should_delta(present_parents)
assert isinstance(version_id, str)
content = self.factory.make(lines, version_id)
if delta or (self.factory.annotated and len(present_parents) > 0):
# Merge annotations from parent texts if needed.
delta_hunks = self._merge_annotations(content, present_parents,
parent_texts, delta, self.factory.annotated,
left_matching_blocks)
if delta:
options.append('line-delta')
store_lines = self.factory.lower_line_delta(delta_hunks)
size, bytes = self._data._record_to_data(version_id, digest,
store_lines)
else:
options.append('fulltext')
# isinstance is slower and we have no hierarchy.
if self.factory.__class__ == KnitPlainFactory:
# Use the already joined bytes saving iteration time in
# _record_to_data.
size, bytes = self._data._record_to_data(version_id, digest,
lines, [line_bytes])
else:
# get mixed annotation + content and feed it into the
# serialiser.
store_lines = self.factory.lower_fulltext(content)
size, bytes = self._data._record_to_data(version_id, digest,
store_lines)
access_memo = self._data.add_raw_records([size], bytes)[0]
self._index.add_versions(
((version_id, options, access_memo, parents),),
random_id=random_id)
return digest, text_length, content
def check(self, progress_bar=None):
"""See VersionedFile.check()."""
def _clone_text(self, new_version_id, old_version_id, parents):
"""See VersionedFile.clone_text()."""
# FIXME RBC 20060228 make fast by only inserting an index with null
# delta.
self.add_lines(new_version_id, parents, self.get_lines(old_version_id))
def get_lines(self, version_id):
"""See VersionedFile.get_lines()."""
return self.get_line_list([version_id])[0]
def _get_record_map(self, version_ids):
"""Produce a dictionary of knit records.
The keys are version_ids, the values are tuples of (method, content,
digest, next).
method is the way the content should be applied.
content is a KnitContent object.
digest is the SHA1 digest of this version id after all steps are done
next is the build-parent of the version, i.e. the leftmost ancestor.
If the method is fulltext, next will be None.
"""
position_map = self._get_components_positions(version_ids)
# c = component_id, m = method, i_m = index_memo, n = next
records = [(c, i_m) for c, (m, i_m, n) in position_map.iteritems()]
record_map = {}
for component_id, content, digest in \
self._data.read_records_iter(records):
method, index_memo, next = position_map[component_id]
record_map[component_id] = method, content, digest, next
return record_map
def get_text(self, version_id):
"""See VersionedFile.get_text"""
return self.get_texts([version_id])[0]
def get_texts(self, version_ids):
return [''.join(l) for l in self.get_line_list(version_ids)]
def get_line_list(self, version_ids):
"""Return the texts of listed versions as a list of strings."""
for version_id in version_ids:
self.check_not_reserved_id(version_id)
text_map, content_map = self._get_content_maps(version_ids)
return [text_map[v] for v in version_ids]
_get_lf_split_line_list = get_line_list
def _get_content_maps(self, version_ids):
"""Produce maps of text and KnitContents
:return: (text_map, content_map) where text_map contains the texts for
the requested versions and content_map contains the KnitContents.
Both dicts take version_ids as their keys.
"""
# FUTURE: This function could be improved for the 'extract many' case
# by tracking each component and only doing the copy when the number of
# children than need to apply delta's to it is > 1 or it is part of the
# final output.
version_ids = list(version_ids)
multiple_versions = len(version_ids) != 1
record_map = self._get_record_map(version_ids)
text_map = {}
content_map = {}
final_content = {}
for version_id in version_ids:
components = []
cursor = version_id
while cursor is not None:
method, data, digest, next = record_map[cursor]
components.append((cursor, method, data, digest))
if cursor in content_map:
break
cursor = next
content = None
for component_id, method, data, digest in reversed(components):
if component_id in content_map:
content = content_map[component_id]
else:
if method == 'fulltext':
assert content is None
content = self.factory.parse_fulltext(data, version_id)
elif method == 'line-delta':
delta = self.factory.parse_line_delta(data, version_id)
if multiple_versions:
# only doing this when we want multiple versions
# output avoids list copies - which reference and
# dereference many strings.
content = content.copy()
content.apply_delta(delta, version_id)
if multiple_versions:
content_map[component_id] = content
if 'no-eol' in self._index.get_options(version_id):
if multiple_versions:
content = content.copy()
content.strip_last_line_newline()
final_content[version_id] = content
# digest here is the digest from the last applied component.
text = content.text()
actual_sha = sha_strings(text)
if actual_sha != digest:
raise KnitCorrupt(self.filename,
'\n sha-1 %s'
'\n of reconstructed text does not match'
'\n expected %s'
'\n for version %s' %
(actual_sha, digest, version_id))
text_map[version_id] = text
return text_map, final_content
def iter_lines_added_or_present_in_versions(self, version_ids=None,
pb=None):
"""See VersionedFile.iter_lines_added_or_present_in_versions()."""
if version_ids is None:
version_ids = self.versions()
if pb is None:
pb = progress.DummyProgress()
# we don't care about inclusions, the caller cares.
# but we need to setup a list of records to visit.
# we need version_id, position, length
version_id_records = []
requested_versions = set(version_ids)
# filter for available versions
for version_id in requested_versions:
if not self.has_version(version_id):
raise RevisionNotPresent(version_id, self.filename)
# get a in-component-order queue:
for version_id in self.versions():
if version_id in requested_versions:
index_memo = self._index.get_position(version_id)
version_id_records.append((version_id, index_memo))
total = len(version_id_records)
for version_idx, (version_id, data, sha_value) in \
enumerate(self._data.read_records_iter(version_id_records)):
pb.update('Walking content.', version_idx, total)
method = self._index.get_method(version_id)
assert method in ('fulltext', 'line-delta')
if method == 'fulltext':
line_iterator = self.factory.get_fulltext_content(data)
else:
line_iterator = self.factory.get_linedelta_content(data)
# XXX: It might be more efficient to yield (version_id,
# line_iterator) in the future. However for now, this is a simpler
# change to integrate into the rest of the codebase. RBC 20071110
for line in line_iterator:
yield line, version_id
pb.update('Walking content.', total, total)
def iter_parents(self, version_ids):
"""Iterate through the parents for many version ids.
:param version_ids: An iterable yielding version_ids.
:return: An iterator that yields (version_id, parents). Requested
version_ids not present in the versioned file are simply skipped.
The order is undefined, allowing for different optimisations in
the underlying implementation.
"""
return self._index.iter_parents(version_ids)
def num_versions(self):
"""See VersionedFile.num_versions()."""
return self._index.num_versions()
__len__ = num_versions
def annotate_iter(self, version_id):
"""See VersionedFile.annotate_iter."""
return self.factory.annotate_iter(self, version_id)
def get_parents(self, version_id):
"""See VersionedFile.get_parents."""
# perf notes:
# optimism counts!
# 52554 calls in 1264 872 internal down from 3674
try:
return self._index.get_parents(version_id)
except KeyError:
raise RevisionNotPresent(version_id, self.filename)
def get_parents_with_ghosts(self, version_id):
"""See VersionedFile.get_parents."""
try:
return self._index.get_parents_with_ghosts(version_id)
except KeyError:
raise RevisionNotPresent(version_id, self.filename)
def get_ancestry(self, versions, topo_sorted=True):
"""See VersionedFile.get_ancestry."""
if isinstance(versions, basestring):
versions = [versions]
if not versions:
return []
return self._index.get_ancestry(versions, topo_sorted)
def get_ancestry_with_ghosts(self, versions):
"""See VersionedFile.get_ancestry_with_ghosts."""
if isinstance(versions, basestring):
versions = [versions]
if not versions:
return []
return self._index.get_ancestry_with_ghosts(versions)
def plan_merge(self, ver_a, ver_b):
"""See VersionedFile.plan_merge."""
ancestors_b = set(self.get_ancestry(ver_b, topo_sorted=False))
ancestors_a = set(self.get_ancestry(ver_a, topo_sorted=False))
annotated_a = self.annotate(ver_a)
annotated_b = self.annotate(ver_b)
return merge._plan_annotate_merge(annotated_a, annotated_b,
ancestors_a, ancestors_b)
class _KnitComponentFile(object):
"""One of the files used to implement a knit database"""
def __init__(self, transport, filename, mode, file_mode=None,
create_parent_dir=False, dir_mode=None):
self._transport = transport
self._filename = filename
self._mode = mode
self._file_mode = file_mode
self._dir_mode = dir_mode
self._create_parent_dir = create_parent_dir
self._need_to_create = False
def _full_path(self):
"""Return the full path to this file."""
return self._transport.base + self._filename
def check_header(self, fp):
line = fp.readline()
if line == '':
# An empty file can actually be treated as though the file doesn't
# exist yet.
raise errors.NoSuchFile(self._full_path())
if line != self.HEADER:
raise KnitHeaderError(badline=line,
filename=self._transport.abspath(self._filename))
def __repr__(self):
return '%s(%s)' % (self.__class__.__name__, self._filename)
class _KnitIndex(_KnitComponentFile):
"""Manages knit index file.
The index is already kept in memory and read on startup, to enable
fast lookups of revision information. The cursor of the index
file is always pointing to the end, making it easy to append
entries.
_cache is a cache for fast mapping from version id to a Index
object.
_history is a cache for fast mapping from indexes to version ids.
The index data format is dictionary compressed when it comes to
parent references; a index entry may only have parents that with a
lover index number. As a result, the index is topological sorted.
Duplicate entries may be written to the index for a single version id
if this is done then the latter one completely replaces the former:
this allows updates to correct version and parent information.
Note that the two entries may share the delta, and that successive
annotations and references MUST point to the first entry.
The index file on disc contains a header, followed by one line per knit
record. The same revision can be present in an index file more than once.
The first occurrence gets assigned a sequence number starting from 0.
The format of a single line is
REVISION_ID FLAGS BYTE_OFFSET LENGTH( PARENT_ID|PARENT_SEQUENCE_ID)* :\n
REVISION_ID is a utf8-encoded revision id
FLAGS is a comma separated list of flags about the record. Values include
no-eol, line-delta, fulltext.
BYTE_OFFSET is the ascii representation of the byte offset in the data file
that the the compressed data starts at.
LENGTH is the ascii representation of the length of the data file.
PARENT_ID a utf-8 revision id prefixed by a '.' that is a parent of
REVISION_ID.
PARENT_SEQUENCE_ID the ascii representation of the sequence number of a
revision id already in the knit that is a parent of REVISION_ID.
The ' :' marker is the end of record marker.
partial writes:
when a write is interrupted to the index file, it will result in a line
that does not end in ' :'. If the ' :' is not present at the end of a line,
or at the end of the file, then the record that is missing it will be
ignored by the parser.
When writing new records to the index file, the data is preceded by '\n'
to ensure that records always start on new lines even if the last write was
interrupted. As a result its normal for the last line in the index to be
missing a trailing newline. One can be added with no harmful effects.
"""
HEADER = "# bzr knit index 8\n"
# speed of knit parsing went from 280 ms to 280 ms with slots addition.
# __slots__ = ['_cache', '_history', '_transport', '_filename']
def _cache_version(self, version_id, options, pos, size, parents):
"""Cache a version record in the history array and index cache.
This is inlined into _load_data for performance. KEEP IN SYNC.
(It saves 60ms, 25% of the __init__ overhead on local 4000 record
indexes).
"""
# only want the _history index to reference the 1st index entry
# for version_id
if version_id not in self._cache:
index = len(self._history)
self._history.append(version_id)
else:
index = self._cache[version_id][5]
self._cache[version_id] = (version_id,
options,
pos,
size,
parents,
index)
def __init__(self, transport, filename, mode, create=False, file_mode=None,
create_parent_dir=False, delay_create=False, dir_mode=None):
_KnitComponentFile.__init__(self, transport, filename, mode,
file_mode=file_mode,
create_parent_dir=create_parent_dir,
dir_mode=dir_mode)
self._cache = {}
# position in _history is the 'official' index for a revision
# but the values may have come from a newer entry.
# so - wc -l of a knit index is != the number of unique names
# in the knit.
self._history = []
try:
fp = self._transport.get(self._filename)
try:
# _load_data may raise NoSuchFile if the target knit is
# completely empty.
_load_data(self, fp)
finally:
fp.close()
except NoSuchFile:
if mode != 'w' or not create:
raise
elif delay_create:
self._need_to_create = True
else:
self._transport.put_bytes_non_atomic(
self._filename, self.HEADER, mode=self._file_mode)
def get_graph(self):
"""Return a list of the node:parents lists from this knit index."""
return [(vid, idx[4]) for vid, idx in self._cache.iteritems()]
def get_ancestry(self, versions, topo_sorted=True):
"""See VersionedFile.get_ancestry."""
# get a graph of all the mentioned versions:
graph = {}
pending = set(versions)
cache = self._cache
while pending:
version = pending.pop()
# trim ghosts
try:
parents = [p for p in cache[version][4] if p in cache]
except KeyError:
raise RevisionNotPresent(version, self._filename)
# if not completed and not a ghost
pending.update([p for p in parents if p not in graph])
graph[version] = parents
if not topo_sorted:
return graph.keys()
return topo_sort(graph.items())
def get_ancestry_with_ghosts(self, versions):
"""See VersionedFile.get_ancestry_with_ghosts."""
# get a graph of all the mentioned versions:
self.check_versions_present(versions)
cache = self._cache
graph = {}
pending = set(versions)
while pending:
version = pending.pop()
try:
parents = cache[version][4]
except KeyError:
# ghost, fake it
graph[version] = []
else:
# if not completed
pending.update([p for p in parents if p not in graph])
graph[version] = parents
return topo_sort(graph.items())
def iter_parents(self, version_ids):
"""Iterate through the parents for many version ids.
:param version_ids: An iterable yielding version_ids.
:return: An iterator that yields (version_id, parents). Requested
version_ids not present in the versioned file are simply skipped.
The order is undefined, allowing for different optimisations in
the underlying implementation.
"""
for version_id in version_ids:
try:
yield version_id, tuple(self.get_parents(version_id))
except KeyError:
pass
def num_versions(self):
return len(self._history)
__len__ = num_versions
def get_versions(self):
"""Get all the versions in the file. not topologically sorted."""
return self._history
def _version_list_to_index(self, versions):
result_list = []
cache = self._cache
for version in versions:
if version in cache:
# -- inlined lookup() --
result_list.append(str(cache[version][5]))
# -- end lookup () --
else:
result_list.append('.' + version)
return ' '.join(result_list)
def add_version(self, version_id, options, index_memo, parents):
"""Add a version record to the index."""
self.add_versions(((version_id, options, index_memo, parents),))
def add_versions(self, versions, random_id=False):
"""Add multiple versions to the index.
:param versions: a list of tuples:
(version_id, options, pos, size, parents).
:param random_id: If True the ids being added were randomly generated
and no check for existence will be performed.
"""
lines = []
orig_history = self._history[:]
orig_cache = self._cache.copy()
try:
for version_id, options, (index, pos, size), parents in versions:
line = "\n%s %s %s %s %s :" % (version_id,
','.join(options),
pos,
size,
self._version_list_to_index(parents))
assert isinstance(line, str), \
'content must be utf-8 encoded: %r' % (line,)
lines.append(line)
self._cache_version(version_id, options, pos, size, parents)
if not self._need_to_create:
self._transport.append_bytes(self._filename, ''.join(lines))
else:
sio = StringIO()
sio.write(self.HEADER)
sio.writelines(lines)
sio.seek(0)
self._transport.put_file_non_atomic(self._filename, sio,
create_parent_dir=self._create_parent_dir,
mode=self._file_mode,
dir_mode=self._dir_mode)
self._need_to_create = False
except:
# If any problems happen, restore the original values and re-raise
self._history = orig_history
self._cache = orig_cache
raise
def has_version(self, version_id):
"""True if the version is in the index."""
return version_id in self._cache
def get_position(self, version_id):
"""Return details needed to access the version.
.kndx indices do not support split-out data, so return None for the
index field.
:return: a tuple (None, data position, size) to hand to the access
logic to get the record.
"""
entry = self._cache[version_id]
return None, entry[2], entry[3]
def get_method(self, version_id):
"""Return compression method of specified version."""
try:
options = self._cache[version_id][1]
except KeyError:
raise RevisionNotPresent(version_id, self._filename)
if 'fulltext' in options:
return 'fulltext'
else:
if 'line-delta' not in options:
raise errors.KnitIndexUnknownMethod(self._full_path(), options)
return 'line-delta'
def get_options(self, version_id):
"""Return a list representing options.
e.g. ['foo', 'bar']
"""
return self._cache[version_id][1]
def get_parents(self, version_id):
"""Return parents of specified version ignoring ghosts."""
return [parent for parent in self._cache[version_id][4]
if parent in self._cache]
def get_parents_with_ghosts(self, version_id):
"""Return parents of specified version with ghosts."""
return self._cache[version_id][4]
def check_versions_present(self, version_ids):
"""Check that all specified versions are present."""
cache = self._cache
for version_id in version_ids:
if version_id not in cache:
raise RevisionNotPresent(version_id, self._filename)
class KnitGraphIndex(object):
"""A knit index that builds on GraphIndex."""
def __init__(self, graph_index, deltas=False, parents=True, add_callback=None):
"""Construct a KnitGraphIndex on a graph_index.
:param graph_index: An implementation of bzrlib.index.GraphIndex.
:param deltas: Allow delta-compressed records.
:param add_callback: If not None, allow additions to the index and call
this callback with a list of added GraphIndex nodes:
[(node, value, node_refs), ...]
:param parents: If True, record knits parents, if not do not record
parents.
"""
self._graph_index = graph_index
self._deltas = deltas
self._add_callback = add_callback
self._parents = parents
if deltas and not parents:
raise KnitCorrupt(self, "Cannot do delta compression without "
"parent tracking.")
def _get_entries(self, keys, check_present=False):
"""Get the entries for keys.
:param keys: An iterable of index keys, - 1-tuples.
"""
keys = set(keys)
found_keys = set()
if self._parents:
for node in self._graph_index.iter_entries(keys):
yield node
found_keys.add(node[1])
else:
# adapt parentless index to the rest of the code.
for node in self._graph_index.iter_entries(keys):
yield node[0], node[1], node[2], ()
found_keys.add(node[1])
if check_present:
missing_keys = keys.difference(found_keys)
if missing_keys:
raise RevisionNotPresent(missing_keys.pop(), self)
def _present_keys(self, version_ids):
return set([
node[1] for node in self._get_entries(version_ids)])
def _parentless_ancestry(self, versions):
"""Honour the get_ancestry API for parentless knit indices."""
wanted_keys = self._version_ids_to_keys(versions)
present_keys = self._present_keys(wanted_keys)
missing = set(wanted_keys).difference(present_keys)
if missing:
raise RevisionNotPresent(missing.pop(), self)
return list(self._keys_to_version_ids(present_keys))
def get_ancestry(self, versions, topo_sorted=True):
"""See VersionedFile.get_ancestry."""
if not self._parents:
return self._parentless_ancestry(versions)
# XXX: This will do len(history) index calls - perhaps
# it should be altered to be a index core feature?
# get a graph of all the mentioned versions:
graph = {}
ghosts = set()
versions = self._version_ids_to_keys(versions)
pending = set(versions)
while pending:
# get all pending nodes
this_iteration = pending
new_nodes = self._get_entries(this_iteration)
found = set()
pending = set()
for (index, key, value, node_refs) in new_nodes:
# dont ask for ghosties - otherwise
# we we can end up looping with pending
# being entirely ghosted.
graph[key] = [parent for parent in node_refs[0]
if parent not in ghosts]
# queue parents
for parent in graph[key]:
# dont examine known nodes again
if parent in graph:
continue
pending.add(parent)
found.add(key)
ghosts.update(this_iteration.difference(found))
if versions.difference(graph):
raise RevisionNotPresent(versions.difference(graph).pop(), self)
if topo_sorted:
result_keys = topo_sort(graph.items())
else:
result_keys = graph.iterkeys()
return [key[0] for key in result_keys]
def get_ancestry_with_ghosts(self, versions):
"""See VersionedFile.get_ancestry."""
if not self._parents:
return self._parentless_ancestry(versions)
# XXX: This will do len(history) index calls - perhaps
# it should be altered to be a index core feature?
# get a graph of all the mentioned versions:
graph = {}
versions = self._version_ids_to_keys(versions)
pending = set(versions)
while pending:
# get all pending nodes
this_iteration = pending
new_nodes = self._get_entries(this_iteration)
pending = set()
for (index, key, value, node_refs) in new_nodes:
graph[key] = node_refs[0]
# queue parents
for parent in graph[key]:
# dont examine known nodes again
if parent in graph:
continue
pending.add(parent)
missing_versions = this_iteration.difference(graph)
missing_needed = versions.intersection(missing_versions)
if missing_needed:
raise RevisionNotPresent(missing_needed.pop(), self)
for missing_version in missing_versions:
# add a key, no parents
graph[missing_version] = []
pending.discard(missing_version) # don't look for it
result_keys = topo_sort(graph.items())
return [key[0] for key in result_keys]
def get_graph(self):
"""Return a list of the node:parents lists from this knit index."""
if not self._parents:
return [(key, ()) for key in self.get_versions()]
result = []
for index, key, value, refs in self._graph_index.iter_all_entries():
result.append((key[0], tuple([ref[0] for ref in refs[0]])))
return result
def iter_parents(self, version_ids):
"""Iterate through the parents for many version ids.
:param version_ids: An iterable yielding version_ids.
:return: An iterator that yields (version_id, parents). Requested
version_ids not present in the versioned file are simply skipped.
The order is undefined, allowing for different optimisations in
the underlying implementation.
"""
if self._parents:
all_nodes = set(self._get_entries(self._version_ids_to_keys(version_ids)))
all_parents = set()
present_parents = set()
for node in all_nodes:
all_parents.update(node[3][0])
# any node we are querying must be present
present_parents.add(node[1])
unknown_parents = all_parents.difference(present_parents)
present_parents.update(self._present_keys(unknown_parents))
for node in all_nodes:
parents = []
for parent in node[3][0]:
if parent in present_parents:
parents.append(parent[0])
yield node[1][0], tuple(parents)
else:
for node in self._get_entries(self._version_ids_to_keys(version_ids)):
yield node[1][0], ()
def num_versions(self):
return len(list(self._graph_index.iter_all_entries()))
__len__ = num_versions
def get_versions(self):
"""Get all the versions in the file. not topologically sorted."""
return [node[1][0] for node in self._graph_index.iter_all_entries()]
def has_version(self, version_id):
"""True if the version is in the index."""
return len(self._present_keys(self._version_ids_to_keys([version_id]))) == 1
def _keys_to_version_ids(self, keys):
return tuple(key[0] for key in keys)
def get_position(self, version_id):
"""Return details needed to access the version.
:return: a tuple (index, data position, size) to hand to the access
logic to get the record.
"""
node = self._get_node(version_id)
bits = node[2][1:].split(' ')
return node[0], int(bits[0]), int(bits[1])
def get_method(self, version_id):
"""Return compression method of specified version."""
if not self._deltas:
return 'fulltext'
return self._parent_compression(self._get_node(version_id)[3][1])
def _parent_compression(self, reference_list):
# use the second reference list to decide if this is delta'd or not.
if len(reference_list):
return 'line-delta'
else:
return 'fulltext'
def _get_node(self, version_id):
try:
return list(self._get_entries(self._version_ids_to_keys([version_id])))[0]
except IndexError:
raise RevisionNotPresent(version_id, self)
def get_options(self, version_id):
"""Return a list representing options.
e.g. ['foo', 'bar']
"""
node = self._get_node(version_id)
if not self._deltas:
options = ['fulltext']
else:
options = [self._parent_compression(node[3][1])]
if node[2][0] == 'N':
options.append('no-eol')
return options
def get_parents(self, version_id):
"""Return parents of specified version ignoring ghosts."""
parents = list(self.iter_parents([version_id]))
if not parents:
# missing key
raise errors.RevisionNotPresent(version_id, self)
return parents[0][1]
def get_parents_with_ghosts(self, version_id):
"""Return parents of specified version with ghosts."""
nodes = list(self._get_entries(self._version_ids_to_keys([version_id]),
check_present=True))
if not self._parents:
return ()
return self._keys_to_version_ids(nodes[0][3][0])
def check_versions_present(self, version_ids):
"""Check that all specified versions are present."""
keys = self._version_ids_to_keys(version_ids)
present = self._present_keys(keys)
missing = keys.difference(present)
if missing:
raise RevisionNotPresent(missing.pop(), self)
def add_version(self, version_id, options, access_memo, parents):
"""Add a version record to the index."""
return self.add_versions(((version_id, options, access_memo, parents),))
def add_versions(self, versions, random_id=False):
"""Add multiple versions to the index.
This function does not insert data into the Immutable GraphIndex
backing the KnitGraphIndex, instead it prepares data for insertion by
the caller and checks that it is safe to insert then calls
self._add_callback with the prepared GraphIndex nodes.
:param versions: a list of tuples:
(version_id, options, pos, size, parents).
:param random_id: If True the ids being added were randomly generated
and no check for existence will be performed.
"""
if not self._add_callback:
raise errors.ReadOnlyError(self)
# we hope there are no repositories with inconsistent parentage
# anymore.
# check for dups
keys = {}
for (version_id, options, access_memo, parents) in versions:
index, pos, size = access_memo
key = (version_id, )
parents = tuple((parent, ) for parent in parents)
if 'no-eol' in options:
value = 'N'
else:
value = ' '
value += "%d %d" % (pos, size)
if not self._deltas:
if 'line-delta' in options:
raise KnitCorrupt(self, "attempt to add line-delta in non-delta knit")
if self._parents:
if self._deltas:
if 'line-delta' in options:
node_refs = (parents, (parents[0],))
else:
node_refs = (parents, ())
else:
node_refs = (parents, )
else:
if parents:
raise KnitCorrupt(self, "attempt to add node with parents "
"in parentless index.")
node_refs = ()
keys[key] = (value, node_refs)
if not random_id:
present_nodes = self._get_entries(keys)
for (index, key, value, node_refs) in present_nodes:
if (value, node_refs) != keys[key]:
raise KnitCorrupt(self, "inconsistent details in add_versions"
": %s %s" % ((value, node_refs), keys[key]))
del keys[key]
result = []
if self._parents:
for key, (value, node_refs) in keys.iteritems():
result.append((key, value, node_refs))
else:
for key, (value, node_refs) in keys.iteritems():
result.append((key, value))
self._add_callback(result)
def _version_ids_to_keys(self, version_ids):
return set((version_id, ) for version_id in version_ids)
class _KnitAccess(object):
"""Access to knit records in a .knit file."""
def __init__(self, transport, filename, _file_mode, _dir_mode,
_need_to_create, _create_parent_dir):
"""Create a _KnitAccess for accessing and inserting data.
:param transport: The transport the .knit is located on.
:param filename: The filename of the .knit.
"""
self._transport = transport
self._filename = filename
self._file_mode = _file_mode
self._dir_mode = _dir_mode
self._need_to_create = _need_to_create
self._create_parent_dir = _create_parent_dir
def add_raw_records(self, sizes, raw_data):
"""Add raw knit bytes to a storage area.
The data is spooled to whereever the access method is storing data.
:param sizes: An iterable containing the size of each raw data segment.
:param raw_data: A bytestring containing the data.
:return: A list of memos to retrieve the record later. Each memo is a
tuple - (index, pos, length), where the index field is always None
for the .knit access method.
"""
assert type(raw_data) == str, \
'data must be plain bytes was %s' % type(raw_data)
if not self._need_to_create:
base = self._transport.append_bytes(self._filename, raw_data)
else:
self._transport.put_bytes_non_atomic(self._filename, raw_data,
create_parent_dir=self._create_parent_dir,
mode=self._file_mode,
dir_mode=self._dir_mode)
self._need_to_create = False
base = 0
result = []
for size in sizes:
result.append((None, base, size))
base += size
return result
def create(self):
"""IFF this data access has its own storage area, initialise it.
:return: None.
"""
self._transport.put_bytes_non_atomic(self._filename, '',
mode=self._file_mode)
def open_file(self):
"""IFF this data access can be represented as a single file, open it.
For knits that are not mapped to a single file on disk this will
always return None.
:return: None or a file handle.
"""
try:
return self._transport.get(self._filename)
except NoSuchFile:
pass
return None
def get_raw_records(self, memos_for_retrieval):
"""Get the raw bytes for a records.
:param memos_for_retrieval: An iterable containing the (index, pos,
length) memo for retrieving the bytes. The .knit method ignores
the index as there is always only a single file.
:return: An iterator over the bytes of the records.
"""
read_vector = [(pos, size) for (index, pos, size) in memos_for_retrieval]
for pos, data in self._transport.readv(self._filename, read_vector):
yield data
class _PackAccess(object):
"""Access to knit records via a collection of packs."""
def __init__(self, index_to_packs, writer=None):
"""Create a _PackAccess object.
:param index_to_packs: A dict mapping index objects to the transport
and file names for obtaining data.
:param writer: A tuple (pack.ContainerWriter, write_index) which
contains the pack to write, and the index that reads from it will
be associated with.
"""
if writer:
self.container_writer = writer[0]
self.write_index = writer[1]
else:
self.container_writer = None
self.write_index = None
self.indices = index_to_packs
def add_raw_records(self, sizes, raw_data):
"""Add raw knit bytes to a storage area.
The data is spooled to the container writer in one bytes-record per
raw data item.
:param sizes: An iterable containing the size of each raw data segment.
:param raw_data: A bytestring containing the data.
:return: A list of memos to retrieve the record later. Each memo is a
tuple - (index, pos, length), where the index field is the
write_index object supplied to the PackAccess object.
"""
assert type(raw_data) == str, \
'data must be plain bytes was %s' % type(raw_data)
result = []
offset = 0
for size in sizes:
p_offset, p_length = self.container_writer.add_bytes_record(
raw_data[offset:offset+size], [])
offset += size
result.append((self.write_index, p_offset, p_length))
return result
def create(self):
"""Pack based knits do not get individually created."""
def get_raw_records(self, memos_for_retrieval):
"""Get the raw bytes for a records.
:param memos_for_retrieval: An iterable containing the (index, pos,
length) memo for retrieving the bytes. The Pack access method
looks up the pack to use for a given record in its index_to_pack
map.
:return: An iterator over the bytes of the records.
"""
# first pass, group into same-index requests
request_lists = []
current_index = None
for (index, offset, length) in memos_for_retrieval:
if current_index == index:
current_list.append((offset, length))
else:
if current_index is not None:
request_lists.append((current_index, current_list))
current_index = index
current_list = [(offset, length)]
# handle the last entry
if current_index is not None:
request_lists.append((current_index, current_list))
for index, offsets in request_lists:
transport, path = self.indices[index]
reader = pack.make_readv_reader(transport, path, offsets)
for names, read_func in reader.iter_records():
yield read_func(None)
def open_file(self):
"""Pack based knits have no single file."""
return None
def set_writer(self, writer, index, (transport, packname)):
"""Set a writer to use for adding data."""
if index is not None:
self.indices[index] = (transport, packname)
self.container_writer = writer
self.write_index = index
class _StreamAccess(object):
"""A Knit Access object that provides data from a datastream.
It also provides a fallback to present as unannotated data, annotated data
from a *backing* access object.
This is triggered by a index_memo which is pointing to a different index
than this was constructed with, and is used to allow extracting full
unannotated texts for insertion into annotated knits.
"""
def __init__(self, reader_callable, stream_index, backing_knit,
orig_factory):
"""Create a _StreamAccess object.
:param reader_callable: The reader_callable from the datastream.
This is called to buffer all the data immediately, for
random access.
:param stream_index: The index the data stream this provides access to
which will be present in native index_memo's.
:param backing_knit: The knit object that will provide access to
annotated texts which are not available in the stream, so as to
create unannotated texts.
:param orig_factory: The original content factory used to generate the
stream. This is used for checking whether the thunk code for
supporting _copy_texts will generate the correct form of data.
"""
self.data = reader_callable(None)
self.stream_index = stream_index
self.backing_knit = backing_knit
self.orig_factory = orig_factory
def get_raw_records(self, memos_for_retrieval):
"""Get the raw bytes for a records.
:param memos_for_retrieval: An iterable containing the (thunk_flag,
index, start, end) memo for retrieving the bytes.
:return: An iterator over the bytes of the records.
"""
# use a generator for memory friendliness
for thunk_flag, version_id, start, end in memos_for_retrieval:
if version_id is self.stream_index:
yield self.data[start:end]
continue
# we have been asked to thunk. This thunking only occurs when
# we are obtaining plain texts from an annotated backing knit
# so that _copy_texts will work.
# We could improve performance here by scanning for where we need
# to do this and using get_line_list, then interleaving the output
# as desired. However, for now, this is sufficient.
if self.orig_factory.__class__ != KnitPlainFactory:
raise errors.KnitCorrupt(
self, 'Bad thunk request %r' % version_id)
lines = self.backing_knit.get_lines(version_id)
line_bytes = ''.join(lines)
digest = sha_string(line_bytes)
if lines:
if lines[-1][-1] != '\n':
lines[-1] = lines[-1] + '\n'
line_bytes += '\n'
orig_options = list(self.backing_knit._index.get_options(version_id))
if 'fulltext' not in orig_options:
if 'line-delta' not in orig_options:
raise errors.KnitCorrupt(self,
'Unknown compression method %r' % orig_options)
orig_options.remove('line-delta')
orig_options.append('fulltext')
# We want plain data, because we expect to thunk only to allow text
# extraction.
size, bytes = self.backing_knit._data._record_to_data(version_id,
digest, lines, line_bytes)
yield bytes
class _StreamIndex(object):
"""A Knit Index object that uses the data map from a datastream."""
def __init__(self, data_list):
"""Create a _StreamIndex object.
:param data_list: The data_list from the datastream.
"""
self.data_list = data_list
self._by_version = {}
pos = 0
for key, options, length, parents in data_list:
self._by_version[key] = options, (pos, pos + length), parents
pos += length
def get_ancestry(self, versions, topo_sorted):
"""Get an ancestry list for versions."""
if topo_sorted:
# Not needed for basic joins
raise NotImplementedError(self.get_ancestry)
# get a graph of all the mentioned versions:
# Little ugly - basically copied from KnitIndex, but don't want to
# accidentally incorporate too much of that index's code.
ancestry = set()
pending = set(versions)
cache = self._by_version
while pending:
version = pending.pop()
# trim ghosts
try:
parents = [p for p in cache[version][2] if p in cache]
except KeyError:
raise RevisionNotPresent(version, self)
# if not completed and not a ghost
pending.update([p for p in parents if p not in ancestry])
ancestry.add(version)
return list(ancestry)
def get_method(self, version_id):
"""Return compression method of specified version."""
try:
options = self._by_version[version_id][0]
except KeyError:
# Strictly speaking this should check in the backing knit, but
# until we have a test to discriminate, this will do.
return 'fulltext'
if 'fulltext' in options:
return 'fulltext'
elif 'line-delta' in options:
return 'line-delta'
else:
raise errors.KnitIndexUnknownMethod(self, options)
def get_options(self, version_id):
"""Return a list representing options.
e.g. ['foo', 'bar']
"""
return self._by_version[version_id][0]
def get_parents_with_ghosts(self, version_id):
"""Return parents of specified version with ghosts."""
return self._by_version[version_id][2]
def get_position(self, version_id):
"""Return details needed to access the version.
_StreamAccess has the data as a big array, so we return slice
coordinates into that (as index_memo's are opaque outside the
index and matching access class).
:return: a tuple (thunk_flag, index, start, end). If thunk_flag is
False, index will be self, otherwise it will be a version id.
"""
try:
start, end = self._by_version[version_id][1]
return False, self, start, end
except KeyError:
# Signal to the access object to handle this from the backing knit.
return (True, version_id, None, None)
def get_versions(self):
"""Get all the versions in the stream."""
return self._by_version.keys()
def iter_parents(self, version_ids):
"""Iterate through the parents for many version ids.
:param version_ids: An iterable yielding version_ids.
:return: An iterator that yields (version_id, parents). Requested
version_ids not present in the versioned file are simply skipped.
The order is undefined, allowing for different optimisations in
the underlying implementation.
"""
result = []
for version in version_ids:
try:
result.append((version, self._by_version[version][2]))
except KeyError:
pass
return result
class _KnitData(object):
"""Manage extraction of data from a KnitAccess, caching and decompressing.
The KnitData class provides the logic for parsing and using knit records,
making use of an access method for the low level read and write operations.
"""
def __init__(self, access):
"""Create a KnitData object.
:param access: The access method to use. Access methods such as
_KnitAccess manage the insertion of raw records and the subsequent
retrieval of the same.
"""
self._access = access
self._checked = False
# TODO: jam 20060713 conceptually, this could spill to disk
# if the cached size gets larger than a certain amount
# but it complicates the model a bit, so for now just use
# a simple dictionary
self._cache = {}
self._do_cache = False
def enable_cache(self):
"""Enable caching of reads."""
self._do_cache = True
def clear_cache(self):
"""Clear the record cache."""
self._do_cache = False
self._cache = {}
def _open_file(self):
return self._access.open_file()
def _record_to_data(self, version_id, digest, lines, dense_lines=None):
"""Convert version_id, digest, lines into a raw data block.
:param dense_lines: The bytes of lines but in a denser form. For
instance, if lines is a list of 1000 bytestrings each ending in \n,
dense_lines may be a list with one line in it, containing all the
1000's lines and their \n's. Using dense_lines if it is already
known is a win because the string join to create bytes in this
function spends less time resizing the final string.
:return: (len, a StringIO instance with the raw data ready to read.)
"""
# Note: using a string copy here increases memory pressure with e.g.
# ISO's, but it is about 3 seconds faster on a 1.2Ghz intel machine
# when doing the initial commit of a mozilla tree. RBC 20070921
bytes = ''.join(chain(
["version %s %d %s\n" % (version_id,
len(lines),
digest)],
dense_lines or lines,
["end %s\n" % version_id]))
assert bytes.__class__ == str
compressed_bytes = bytes_to_gzip(bytes)
return len(compressed_bytes), compressed_bytes
def add_raw_records(self, sizes, raw_data):
"""Append a prepared record to the data file.
:param sizes: An iterable containing the size of each raw data segment.
:param raw_data: A bytestring containing the data.
:return: a list of index data for the way the data was stored.
See the access method add_raw_records documentation for more
details.
"""
return self._access.add_raw_records(sizes, raw_data)
def _parse_record_header(self, version_id, raw_data):
"""Parse a record header for consistency.
:return: the header and the decompressor stream.
as (stream, header_record)
"""
df = GzipFile(mode='rb', fileobj=StringIO(raw_data))
try:
rec = self._check_header(version_id, df.readline())
except Exception, e:
raise KnitCorrupt(self._access,
"While reading {%s} got %s(%s)"
% (version_id, e.__class__.__name__, str(e)))
return df, rec
def _check_header(self, version_id, line):
rec = line.split()
if len(rec) != 4:
raise KnitCorrupt(self._access,
'unexpected number of elements in record header')
if rec[1] != version_id:
raise KnitCorrupt(self._access,
'unexpected version, wanted %r, got %r'
% (version_id, rec[1]))
return rec
def _parse_record(self, version_id, data):
# profiling notes:
# 4168 calls in 2880 217 internal
# 4168 calls to _parse_record_header in 2121
# 4168 calls to readlines in 330
df = GzipFile(mode='rb', fileobj=StringIO(data))
try:
record_contents = df.readlines()
except Exception, e:
raise KnitCorrupt(self._access,
"While reading {%s} got %s(%s)"
% (version_id, e.__class__.__name__, str(e)))
header = record_contents.pop(0)
rec = self._check_header(version_id, header)
last_line = record_contents.pop()
if len(record_contents) != int(rec[2]):
raise KnitCorrupt(self._access,
'incorrect number of lines %s != %s'
' for version {%s}'
% (len(record_contents), int(rec[2]),
version_id))
if last_line != 'end %s\n' % rec[1]:
raise KnitCorrupt(self._access,
'unexpected version end line %r, wanted %r'
% (last_line, version_id))
df.close()
return record_contents, rec[3]
def read_records_iter_raw(self, records):
"""Read text records from data file and yield raw data.
This unpacks enough of the text record to validate the id is
as expected but thats all.
"""
# setup an iterator of the external records:
# uses readv so nice and fast we hope.
if len(records):
# grab the disk data needed.
if self._cache:
# Don't check _cache if it is empty
needed_offsets = [index_memo for version_id, index_memo
in records
if version_id not in self._cache]
else:
needed_offsets = [index_memo for version_id, index_memo
in records]
raw_records = self._access.get_raw_records(needed_offsets)
for version_id, index_memo in records:
if version_id in self._cache:
# This data has already been validated
data = self._cache[version_id]
else:
data = raw_records.next()
if self._do_cache:
self._cache[version_id] = data
# validate the header
df, rec = self._parse_record_header(version_id, data)
df.close()
yield version_id, data
def read_records_iter(self, records):
"""Read text records from data file and yield result.
The result will be returned in whatever is the fastest to read.
Not by the order requested. Also, multiple requests for the same
record will only yield 1 response.
:param records: A list of (version_id, pos, len) entries
:return: Yields (version_id, contents, digest) in the order
read, not the order requested
"""
if not records:
return
if self._cache:
# Skip records we have alread seen
yielded_records = set()
needed_records = set()
for record in records:
if record[0] in self._cache:
if record[0] in yielded_records:
continue
yielded_records.add(record[0])
data = self._cache[record[0]]
content, digest = self._parse_record(record[0], data)
yield (record[0], content, digest)
else:
needed_records.add(record)
needed_records = sorted(needed_records, key=operator.itemgetter(1))
else:
needed_records = sorted(set(records), key=operator.itemgetter(1))
if not needed_records:
return
# The transport optimizes the fetching as well
# (ie, reads continuous ranges.)
raw_data = self._access.get_raw_records(
[index_memo for version_id, index_memo in needed_records])
for (version_id, index_memo), data in \
izip(iter(needed_records), raw_data):
content, digest = self._parse_record(version_id, data)
if self._do_cache:
self._cache[version_id] = data
yield version_id, content, digest
def read_records(self, records):
"""Read records into a dictionary."""
components = {}
for record_id, content, digest in \
self.read_records_iter(records):
components[record_id] = (content, digest)
return components
class InterKnit(InterVersionedFile):
"""Optimised code paths for knit to knit operations."""
_matching_file_from_factory = KnitVersionedFile
_matching_file_to_factory = KnitVersionedFile
@staticmethod
def is_compatible(source, target):
"""Be compatible with knits. """
try:
return (isinstance(source, KnitVersionedFile) and
isinstance(target, KnitVersionedFile))
except AttributeError:
return False
def _copy_texts(self, pb, msg, version_ids, ignore_missing=False):
"""Copy texts to the target by extracting and adding them one by one.
see join() for the parameter definitions.
"""
version_ids = self._get_source_version_ids(version_ids, ignore_missing)
graph = self.source.get_graph(version_ids)
order = topo_sort(graph.items())
def size_of_content(content):
return sum(len(line) for line in content.text())
# Cache at most 10MB of parent texts
parent_cache = lru_cache.LRUSizeCache(max_size=10*1024*1024,
compute_size=size_of_content)
# TODO: jam 20071116 It would be nice to have a streaming interface to
# get multiple texts from a source. The source could be smarter
# about how it handled intermediate stages.
# get_line_list() or make_mpdiffs() seem like a possibility, but
# at the moment they extract all full texts into memory, which
# causes us to store more than our 3x fulltext goal.
# Repository.iter_files_bytes() may be another possibility
to_process = [version for version in order
if version not in self.target]
total = len(to_process)
pb = ui.ui_factory.nested_progress_bar()
try:
for index, version in enumerate(to_process):
pb.update('Converting versioned data', index, total)
sha1, num_bytes, parent_text = self.target.add_lines(version,
self.source.get_parents_with_ghosts(version),
self.source.get_lines(version),
parent_texts=parent_cache)
parent_cache[version] = parent_text
finally:
pb.finished()
return total
def join(self, pb=None, msg=None, version_ids=None, ignore_missing=False):
"""See InterVersionedFile.join."""
assert isinstance(self.source, KnitVersionedFile)
assert isinstance(self.target, KnitVersionedFile)
# If the source and target are mismatched w.r.t. annotations vs
# plain, the data needs to be converted accordingly
if self.source.factory.annotated == self.target.factory.annotated:
converter = None
elif self.source.factory.annotated:
converter = self._anno_to_plain_converter
else:
# We're converting from a plain to an annotated knit. Copy them
# across by full texts.
return self._copy_texts(pb, msg, version_ids, ignore_missing)
version_ids = self._get_source_version_ids(version_ids, ignore_missing)
if not version_ids:
return 0
pb = ui.ui_factory.nested_progress_bar()
try:
version_ids = list(version_ids)
if None in version_ids:
version_ids.remove(None)
self.source_ancestry = set(self.source.get_ancestry(version_ids,
topo_sorted=False))
this_versions = set(self.target._index.get_versions())
# XXX: For efficiency we should not look at the whole index,
# we only need to consider the referenced revisions - they
# must all be present, or the method must be full-text.
# TODO, RBC 20070919
needed_versions = self.source_ancestry - this_versions
if not needed_versions:
return 0
full_list = topo_sort(self.source.get_graph())
version_list = [i for i in full_list if (not self.target.has_version(i)
and i in needed_versions)]
# plan the join:
copy_queue = []
copy_queue_records = []
copy_set = set()
for version_id in version_list:
options = self.source._index.get_options(version_id)
parents = self.source._index.get_parents_with_ghosts(version_id)
# check that its will be a consistent copy:
for parent in parents:
# if source has the parent, we must :
# * already have it or
# * have it scheduled already
# otherwise we don't care
assert (self.target.has_version(parent) or
parent in copy_set or
not self.source.has_version(parent))
index_memo = self.source._index.get_position(version_id)
copy_queue_records.append((version_id, index_memo))
copy_queue.append((version_id, options, parents))
copy_set.add(version_id)
# data suck the join:
count = 0
total = len(version_list)
raw_datum = []
raw_records = []
for (version_id, raw_data), \
(version_id2, options, parents) in \
izip(self.source._data.read_records_iter_raw(copy_queue_records),
copy_queue):
assert version_id == version_id2, 'logic error, inconsistent results'
count = count + 1
pb.update("Joining knit", count, total)
if converter:
size, raw_data = converter(raw_data, version_id, options,
parents)
else:
size = len(raw_data)
raw_records.append((version_id, options, parents, size))
raw_datum.append(raw_data)
self.target._add_raw_records(raw_records, ''.join(raw_datum))
return count
finally:
pb.finished()
def _anno_to_plain_converter(self, raw_data, version_id, options,
parents):
"""Convert annotated content to plain content."""
data, digest = self.source._data._parse_record(version_id, raw_data)
if 'fulltext' in options:
content = self.source.factory.parse_fulltext(data, version_id)
lines = self.target.factory.lower_fulltext(content)
else:
delta = self.source.factory.parse_line_delta(data, version_id,
plain=True)
lines = self.target.factory.lower_line_delta(delta)
return self.target._data._record_to_data(version_id, digest, lines)
InterVersionedFile.register_optimiser(InterKnit)
class WeaveToKnit(InterVersionedFile):
"""Optimised code paths for weave to knit operations."""
_matching_file_from_factory = bzrlib.weave.WeaveFile
_matching_file_to_factory = KnitVersionedFile
@staticmethod
def is_compatible(source, target):
"""Be compatible with weaves to knits."""
try:
return (isinstance(source, bzrlib.weave.Weave) and
isinstance(target, KnitVersionedFile))
except AttributeError:
return False
def join(self, pb=None, msg=None, version_ids=None, ignore_missing=False):
"""See InterVersionedFile.join."""
assert isinstance(self.source, bzrlib.weave.Weave)
assert isinstance(self.target, KnitVersionedFile)
version_ids = self._get_source_version_ids(version_ids, ignore_missing)
if not version_ids:
return 0
pb = ui.ui_factory.nested_progress_bar()
try:
version_ids = list(version_ids)
self.source_ancestry = set(self.source.get_ancestry(version_ids))
this_versions = set(self.target._index.get_versions())
needed_versions = self.source_ancestry - this_versions
if not needed_versions:
return 0
full_list = topo_sort(self.source.get_graph())
version_list = [i for i in full_list if (not self.target.has_version(i)
and i in needed_versions)]
# do the join:
count = 0
total = len(version_list)
for version_id in version_list:
pb.update("Converting to knit", count, total)
parents = self.source.get_parents(version_id)
# check that its will be a consistent copy:
for parent in parents:
# if source has the parent, we must already have it
assert (self.target.has_version(parent))
self.target.add_lines(
version_id, parents, self.source.get_lines(version_id))
count = count + 1
return count
finally:
pb.finished()
InterVersionedFile.register_optimiser(WeaveToKnit)
# Deprecated, use PatienceSequenceMatcher instead
KnitSequenceMatcher = patiencediff.PatienceSequenceMatcher
def annotate_knit(knit, revision_id):
"""Annotate a knit with no cached annotations.
This implementation is for knits with no cached annotations.
It will work for knits with cached annotations, but this is not
recommended.
"""
ancestry = knit.get_ancestry(revision_id)
fulltext = dict(zip(ancestry, knit.get_line_list(ancestry)))
annotations = {}
for candidate in ancestry:
if candidate in annotations:
continue
parents = knit.get_parents(candidate)
if len(parents) == 0:
blocks = None
elif knit._index.get_method(candidate) != 'line-delta':
blocks = None
else:
parent, sha1, noeol, delta = knit.get_delta(candidate)
blocks = KnitContent.get_line_delta_blocks(delta,
fulltext[parents[0]], fulltext[candidate])
annotations[candidate] = list(annotate.reannotate([annotations[p]
for p in parents], fulltext[candidate], candidate, blocks))
return iter(annotations[revision_id])
try:
from bzrlib._knit_load_data_c import _load_data_c as _load_data
except ImportError:
from bzrlib._knit_load_data_py import _load_data_py as _load_data
|