summaryrefslogtreecommitdiffstats
path: root/Utilities/cmlibarchive/libarchive/archive_read.c
blob: a65b94dae7447732d5f33bdb2eb5876232d7449e (plain)
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
/*-
 * Copyright (c) 2003-2011 Tim Kientzle
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 *
 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR(S) ``AS IS'' AND ANY EXPRESS OR
 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
 * IN NO EVENT SHALL THE AUTHOR(S) BE LIABLE FOR ANY DIRECT, INDIRECT,
 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */

/*
 * This file contains the "essential" portions of the read API, that
 * is, stuff that will probably always be used by any client that
 * actually needs to read an archive.  Optional pieces have been, as
 * far as possible, separated out into separate files to avoid
 * needlessly bloating statically-linked clients.
 */

#include "archive_platform.h"
__FBSDID("$FreeBSD: head/lib/libarchive/archive_read.c 201157 2009-12-29 05:30:23Z kientzle $");

#ifdef HAVE_ERRNO_H
#include <errno.h>
#endif
#include <stdio.h>
#ifdef HAVE_STDLIB_H
#include <stdlib.h>
#endif
#ifdef HAVE_STRING_H
#include <string.h>
#endif
#ifdef HAVE_UNISTD_H
#include <unistd.h>
#endif

#include "archive.h"
#include "archive_entry.h"
#include "archive_private.h"
#include "archive_read_private.h"

#define minimum(a, b) (a < b ? a : b)

static int	choose_filters(struct archive_read *);
static int	choose_format(struct archive_read *);
static struct archive_vtable *archive_read_vtable(void);
static int64_t	_archive_filter_bytes(struct archive *, int);
static int	_archive_filter_code(struct archive *, int);
static const char *_archive_filter_name(struct archive *, int);
static int  _archive_filter_count(struct archive *);
static int	_archive_read_close(struct archive *);
static int	_archive_read_data_block(struct archive *,
		    const void **, size_t *, int64_t *);
static int	_archive_read_free(struct archive *);
static int	_archive_read_next_header(struct archive *,
		    struct archive_entry **);
static int	_archive_read_next_header2(struct archive *,
		    struct archive_entry *);
static int64_t  advance_file_pointer(struct archive_read_filter *, int64_t);

static struct archive_vtable *
archive_read_vtable(void)
{
	static struct archive_vtable av;
	static int inited = 0;

	if (!inited) {
		av.archive_filter_bytes = _archive_filter_bytes;
		av.archive_filter_code = _archive_filter_code;
		av.archive_filter_name = _archive_filter_name;
		av.archive_filter_count = _archive_filter_count;
		av.archive_read_data_block = _archive_read_data_block;
		av.archive_read_next_header = _archive_read_next_header;
		av.archive_read_next_header2 = _archive_read_next_header2;
		av.archive_free = _archive_read_free;
		av.archive_close = _archive_read_close;
		inited = 1;
	}
	return (&av);
}

/*
 * Allocate, initialize and return a struct archive object.
 */
struct archive *
archive_read_new(void)
{
	struct archive_read *a;

	a = (struct archive_read *)malloc(sizeof(*a));
	if (a == NULL)
		return (NULL);
	memset(a, 0, sizeof(*a));
	a->archive.magic = ARCHIVE_READ_MAGIC;

	a->archive.state = ARCHIVE_STATE_NEW;
	a->entry = archive_entry_new2(&a->archive);
	a->archive.vtable = archive_read_vtable();

	return (&a->archive);
}

/*
 * Record the do-not-extract-to file. This belongs in archive_read_extract.c.
 */
void
archive_read_extract_set_skip_file(struct archive *_a, int64_t d, int64_t i)
{
	struct archive_read *a = (struct archive_read *)_a;

	if (ARCHIVE_OK != __archive_check_magic(_a, ARCHIVE_READ_MAGIC,
		ARCHIVE_STATE_ANY, "archive_read_extract_set_skip_file"))
		return;
	a->skip_file_set = 1;
	a->skip_file_dev = d;
	a->skip_file_ino = i;
}

/*
 * Open the archive
 */
int
archive_read_open(struct archive *a, void *client_data,
    archive_open_callback *client_opener, archive_read_callback *client_reader,
    archive_close_callback *client_closer)
{
	/* Old archive_read_open() is just a thin shell around
	 * archive_read_open1. */
	archive_read_set_open_callback(a, client_opener);
	archive_read_set_read_callback(a, client_reader);
	archive_read_set_close_callback(a, client_closer);
	archive_read_set_callback_data(a, client_data);
	return archive_read_open1(a);
}


int
archive_read_open2(struct archive *a, void *client_data,
    archive_open_callback *client_opener,
    archive_read_callback *client_reader,
    archive_skip_callback *client_skipper,
    archive_close_callback *client_closer)
{
	/* Old archive_read_open2() is just a thin shell around
	 * archive_read_open1. */
	archive_read_set_callback_data(a, client_data);
	archive_read_set_open_callback(a, client_opener);
	archive_read_set_read_callback(a, client_reader);
	archive_read_set_skip_callback(a, client_skipper);
	archive_read_set_close_callback(a, client_closer);
	return archive_read_open1(a);
}

static ssize_t
client_read_proxy(struct archive_read_filter *self, const void **buff)
{
	ssize_t r;
	r = (self->archive->client.reader)(&self->archive->archive,
	    self->data, buff);
	return (r);
}

static int64_t
client_skip_proxy(struct archive_read_filter *self, int64_t request)
{
	if (request < 0)
		__archive_errx(1, "Negative skip requested.");
	if (request == 0)
		return 0;

	if (self->archive->client.skipper != NULL) {
		/* Seek requests over 1GiB are broken down into
		 * multiple seeks.  This avoids overflows when the
		 * requests get passed through 32-bit arguments. */
		int64_t skip_limit = (int64_t)1 << 30;
		int64_t total = 0;
		for (;;) {
			int64_t get, ask = request;
			if (ask > skip_limit)
				ask = skip_limit;
			get = (self->archive->client.skipper)
				(&self->archive->archive, self->data, ask);
			if (get == 0)
				return (total);
			request -= get;
			total += get;
		}
	} else if (self->archive->client.seeker != NULL
		&& request > 64 * 1024) {
		/* If the client provided a seeker but not a skipper,
		 * we can use the seeker to skip forward.
		 *
		 * Note: This isn't always a good idea.  The client
		 * skipper is allowed to skip by less than requested
		 * if it needs to maintain block alignment.  The
		 * seeker is not allowed to play such games, so using
		 * the seeker here may be a performance loss compared
		 * to just reading and discarding.  That's why we
		 * only do this for skips of over 64k.
		 */
		int64_t before = self->position;
		int64_t after = (self->archive->client.seeker)
		    (&self->archive->archive, self->data, request, SEEK_CUR);
		if (after != before + request)
			return ARCHIVE_FATAL;
		return after - before;
	}
	return 0;
}

static int64_t
client_seek_proxy(struct archive_read_filter *self, int64_t offset, int whence)
{
	/* DO NOT use the skipper here!  If we transparently handled
	 * forward seek here by using the skipper, that will break
	 * other libarchive code that assumes a successful forward
	 * seek means it can also seek backwards.
	 */
	if (self->archive->client.seeker == NULL)
		return (ARCHIVE_FAILED);
	return (self->archive->client.seeker)(&self->archive->archive,
	    self->data, offset, whence);
}

static int
client_close_proxy(struct archive_read_filter *self)
{
	int r = ARCHIVE_OK, r2;
	unsigned int i;

	if (self->archive->client.closer == NULL)
		return (r);
	for (i = 0; i < self->archive->client.nodes; i++)
	{
		r2 = (self->archive->client.closer)
			((struct archive *)self->archive,
				self->archive->client.dataset[i].data);
		if (r > r2)
			r = r2;
	}
	return (r);
}

static int
client_open_proxy(struct archive_read_filter *self)
{
  int r = ARCHIVE_OK;
	if (self->archive->client.opener != NULL)
		r = (self->archive->client.opener)(
		    (struct archive *)self->archive, self->data);
	return (r);
}

static int
client_switch_proxy(struct archive_read_filter *self, unsigned int iindex)
{
  int r1 = ARCHIVE_OK, r2 = ARCHIVE_OK;
	void *data2 = NULL;

	/* Don't do anything if already in the specified data node */
	if (self->archive->client.cursor == iindex)
		return (ARCHIVE_OK);

	self->archive->client.cursor = iindex;
	data2 = self->archive->client.dataset[self->archive->client.cursor].data;
	if (self->archive->client.switcher != NULL)
	{
		r1 = r2 = (self->archive->client.switcher)
			((struct archive *)self->archive, self->data, data2);
		self->data = data2;
	}
	else
	{
		/* Attempt to call close and open instead */
		if (self->archive->client.closer != NULL)
			r1 = (self->archive->client.closer)
				((struct archive *)self->archive, self->data);
		self->data = data2;
		if (self->archive->client.opener != NULL)
			r2 = (self->archive->client.opener)
				((struct archive *)self->archive, self->data);
	}
	return (r1 < r2) ? r1 : r2;
}

int
archive_read_set_open_callback(struct archive *_a,
    archive_open_callback *client_opener)
{
	struct archive_read *a = (struct archive_read *)_a;
	archive_check_magic(_a, ARCHIVE_READ_MAGIC, ARCHIVE_STATE_NEW,
	    "archive_read_set_open_callback");
	a->client.opener = client_opener;
	return ARCHIVE_OK;
}

int
archive_read_set_read_callback(struct archive *_a,
    archive_read_callback *client_reader)
{
	struct archive_read *a = (struct archive_read *)_a;
	archive_check_magic(_a, ARCHIVE_READ_MAGIC, ARCHIVE_STATE_NEW,
	    "archive_read_set_read_callback");
	a->client.reader = client_reader;
	return ARCHIVE_OK;
}

int
archive_read_set_skip_callback(struct archive *_a,
    archive_skip_callback *client_skipper)
{
	struct archive_read *a = (struct archive_read *)_a;
	archive_check_magic(_a, ARCHIVE_READ_MAGIC, ARCHIVE_STATE_NEW,
	    "archive_read_set_skip_callback");
	a->client.skipper = client_skipper;
	return ARCHIVE_OK;
}

int
archive_read_set_seek_callback(struct archive *_a,
    archive_seek_callback *client_seeker)
{
	struct archive_read *a = (struct archive_read *)_a;
	archive_check_magic(_a, ARCHIVE_READ_MAGIC, ARCHIVE_STATE_NEW,
	    "archive_read_set_seek_callback");
	a->client.seeker = client_seeker;
	return ARCHIVE_OK;
}

int
archive_read_set_close_callback(struct archive *_a,
    archive_close_callback *client_closer)
{
	struct archive_read *a = (struct archive_read *)_a;
	archive_check_magic(_a, ARCHIVE_READ_MAGIC, ARCHIVE_STATE_NEW,
	    "archive_read_set_close_callback");
	a->client.closer = client_closer;
	return ARCHIVE_OK;
}

int
archive_read_set_switch_callback(struct archive *_a,
    archive_switch_callback *client_switcher)
{
	struct archive_read *a = (struct archive_read *)_a;
	archive_check_magic(_a, ARCHIVE_READ_MAGIC, ARCHIVE_STATE_NEW,
	    "archive_read_set_switch_callback");
	a->client.switcher = client_switcher;
	return ARCHIVE_OK;
}

int
archive_read_set_callback_data(struct archive *_a, void *client_data)
{
	return archive_read_set_callback_data2(_a, client_data, 0);
}

int
archive_read_set_callback_data2(struct archive *_a, void *client_data,
    unsigned int iindex)
{
	struct archive_read *a = (struct archive_read *)_a;
	archive_check_magic(_a, ARCHIVE_READ_MAGIC, ARCHIVE_STATE_NEW,
	    "archive_read_set_callback_data2");

	if (a->client.nodes == 0)
	{
		a->client.dataset = (struct archive_read_data_node *)
		    calloc(1, sizeof(*a->client.dataset));
		if (a->client.dataset == NULL)
		{
			archive_set_error(&a->archive, ENOMEM,
				"No memory.");
			return ARCHIVE_FATAL;
		}
		a->client.nodes = 1;
	}

	if (iindex > a->client.nodes - 1)
	{
		archive_set_error(&a->archive, EINVAL,
			"Invalid index specified.");
		return ARCHIVE_FATAL;
	}
	a->client.dataset[iindex].data = client_data;
	a->client.dataset[iindex].begin_position = -1;
	a->client.dataset[iindex].total_size = -1;
	return ARCHIVE_OK;
}

int
archive_read_add_callback_data(struct archive *_a, void *client_data,
    unsigned int iindex)
{
	struct archive_read *a = (struct archive_read *)_a;
	void *p;
	unsigned int i;

	archive_check_magic(_a, ARCHIVE_READ_MAGIC, ARCHIVE_STATE_NEW,
	    "archive_read_add_callback_data");
	if (iindex > a->client.nodes) {
		archive_set_error(&a->archive, EINVAL,
			"Invalid index specified.");
		return ARCHIVE_FATAL;
	}
	p = realloc(a->client.dataset, sizeof(*a->client.dataset)
		* (++(a->client.nodes)));
	if (p == NULL) {
		archive_set_error(&a->archive, ENOMEM,
			"No memory.");
		return ARCHIVE_FATAL;
	}
	a->client.dataset = (struct archive_read_data_node *)p;
	for (i = a->client.nodes - 1; i > iindex && i > 0; i--) {
		a->client.dataset[i].data = a->client.dataset[i-1].data;
		a->client.dataset[i].begin_position = -1;
		a->client.dataset[i].total_size = -1;
	}
	a->client.dataset[iindex].data = client_data;
	a->client.dataset[iindex].begin_position = -1;
	a->client.dataset[iindex].total_size = -1;
	return ARCHIVE_OK;
}

int
archive_read_append_callback_data(struct archive *_a, void *client_data)
{
	struct archive_read *a = (struct archive_read *)_a;
	return archive_read_add_callback_data(_a, client_data, a->client.nodes);
}

int
archive_read_prepend_callback_data(struct archive *_a, void *client_data)
{
	return archive_read_add_callback_data(_a, client_data, 0);
}

int
archive_read_open1(struct archive *_a)
{
	struct archive_read *a = (struct archive_read *)_a;
	struct archive_read_filter *filter, *tmp;
	int slot, e;
	unsigned int i;

	archive_check_magic(_a, ARCHIVE_READ_MAGIC, ARCHIVE_STATE_NEW,
	    "archive_read_open");
	archive_clear_error(&a->archive);

	if (a->client.reader == NULL) {
		archive_set_error(&a->archive, EINVAL,
		    "No reader function provided to archive_read_open");
		a->archive.state = ARCHIVE_STATE_FATAL;
		return (ARCHIVE_FATAL);
	}

	/* Open data source. */
	if (a->client.opener != NULL) {
		e = (a->client.opener)(&a->archive, a->client.dataset[0].data);
		if (e != 0) {
			/* If the open failed, call the closer to clean up. */
			if (a->client.closer) {
				for (i = 0; i < a->client.nodes; i++)
					(a->client.closer)(&a->archive,
					    a->client.dataset[i].data);
			}
			return (e);
		}
	}

	filter = calloc(1, sizeof(*filter));
	if (filter == NULL)
		return (ARCHIVE_FATAL);
	filter->bidder = NULL;
	filter->upstream = NULL;
	filter->archive = a;
	filter->data = a->client.dataset[0].data;
	filter->open = client_open_proxy;
	filter->read = client_read_proxy;
	filter->skip = client_skip_proxy;
	filter->seek = client_seek_proxy;
	filter->close = client_close_proxy;
	filter->sswitch = client_switch_proxy;
	filter->name = "none";
	filter->code = ARCHIVE_FILTER_NONE;

	a->client.dataset[0].begin_position = 0;
	if (!a->filter || !a->bypass_filter_bidding)
	{
		a->filter = filter;
		/* Build out the input pipeline. */
		e = choose_filters(a);
		if (e < ARCHIVE_WARN) {
			a->archive.state = ARCHIVE_STATE_FATAL;
			return (ARCHIVE_FATAL);
		}
	}
	else
	{
		/* Need to add "NONE" type filter at the end of the filter chain */
		tmp = a->filter;
		while (tmp->upstream)
			tmp = tmp->upstream;
		tmp->upstream = filter;
	}

	if (!a->format)
	{
		slot = choose_format(a);
		if (slot < 0) {
			__archive_read_close_filters(a);
			a->archive.state = ARCHIVE_STATE_FATAL;
			return (ARCHIVE_FATAL);
		}
		a->format = &(a->formats[slot]);
	}

	a->archive.state = ARCHIVE_STATE_HEADER;

	/* Ensure libarchive starts from the first node in a multivolume set */
	client_switch_proxy(a->filter, 0);
	return (e);
}

/*
 * Allow each registered stream transform to bid on whether
 * it wants to handle this stream.  Repeat until we've finished
 * building the pipeline.
 */
static int
choose_filters(struct archive_read *a)
{
	int number_bidders, i, bid, best_bid;
	struct archive_read_filter_bidder *bidder, *best_bidder;
	struct archive_read_filter *filter;
	ssize_t avail;
	int r;

	for (;;) {
		number_bidders = sizeof(a->bidders) / sizeof(a->bidders[0]);

		best_bid = 0;
		best_bidder = NULL;

		bidder = a->bidders;
		for (i = 0; i < number_bidders; i++, bidder++) {
			if (bidder->bid != NULL) {
				bid = (bidder->bid)(bidder, a->filter);
				if (bid > best_bid) {
					best_bid = bid;
					best_bidder = bidder;
				}
			}
		}

		/* If no bidder, we're done. */
		if (best_bidder == NULL) {
			/* Verify the filter by asking it for some data. */
			__archive_read_filter_ahead(a->filter, 1, &avail);
			if (avail < 0) {
				__archive_read_close_filters(a);
				__archive_read_free_filters(a);
				return (ARCHIVE_FATAL);
			}
			a->archive.compression_name = a->filter->name;
			a->archive.compression_code = a->filter->code;
			return (ARCHIVE_OK);
		}

		filter
		    = (struct archive_read_filter *)calloc(1, sizeof(*filter));
		if (filter == NULL)
			return (ARCHIVE_FATAL);
		filter->bidder = best_bidder;
		filter->archive = a;
		filter->upstream = a->filter;
		a->filter = filter;
		r = (best_bidder->init)(a->filter);
		if (r != ARCHIVE_OK) {
			__archive_read_close_filters(a);
			__archive_read_free_filters(a);
			return (ARCHIVE_FATAL);
		}
	}
}

/*
 * Read header of next entry.
 */
static int
_archive_read_next_header2(struct archive *_a, struct archive_entry *entry)
{
	struct archive_read *a = (struct archive_read *)_a;
	int r1 = ARCHIVE_OK, r2;

	archive_check_magic(_a, ARCHIVE_READ_MAGIC,
	    ARCHIVE_STATE_HEADER | ARCHIVE_STATE_DATA,
	    "archive_read_next_header");

	archive_entry_clear(entry);
	archive_clear_error(&a->archive);

	/*
	 * If client didn't consume entire data, skip any remainder
	 * (This is especially important for GNU incremental directories.)
	 */
	if (a->archive.state == ARCHIVE_STATE_DATA) {
		r1 = archive_read_data_skip(&a->archive);
		if (r1 == ARCHIVE_EOF)
			archive_set_error(&a->archive, EIO,
			    "Premature end-of-file.");
		if (r1 == ARCHIVE_EOF || r1 == ARCHIVE_FATAL) {
			a->archive.state = ARCHIVE_STATE_FATAL;
			return (ARCHIVE_FATAL);
		}
	}

	/* Record start-of-header offset in uncompressed stream. */
	a->header_position = a->filter->position;

	++_a->file_count;
	r2 = (a->format->read_header)(a, entry);

	/*
	 * EOF and FATAL are persistent at this layer.  By
	 * modifying the state, we guarantee that future calls to
	 * read a header or read data will fail.
	 */
	switch (r2) {
	case ARCHIVE_EOF:
		a->archive.state = ARCHIVE_STATE_EOF;
		--_a->file_count;/* Revert a file counter. */
		break;
	case ARCHIVE_OK:
		a->archive.state = ARCHIVE_STATE_DATA;
		break;
	case ARCHIVE_WARN:
		a->archive.state = ARCHIVE_STATE_DATA;
		break;
	case ARCHIVE_RETRY:
		break;
	case ARCHIVE_FATAL:
		a->archive.state = ARCHIVE_STATE_FATAL;
		break;
	}

	a->read_data_output_offset = 0;
	a->read_data_remaining = 0;
	a->read_data_is_posix_read = 0;
	a->read_data_requested = 0;
	a->data_start_node = a->client.cursor;
	/* EOF always wins; otherwise return the worst error. */
	return (r2 < r1 || r2 == ARCHIVE_EOF) ? r2 : r1;
}

int
_archive_read_next_header(struct archive *_a, struct archive_entry **entryp)
{
	int ret;
	struct archive_read *a = (struct archive_read *)_a;
	*entryp = NULL;
	ret = _archive_read_next_header2(_a, a->entry);
	*entryp = a->entry;
	return ret;
}

/*
 * Allow each registered format to bid on whether it wants to handle
 * the next entry.  Return index of winning bidder.
 */
static int
choose_format(struct archive_read *a)
{
	int slots;
	int i;
	int bid, best_bid;
	int best_bid_slot;

	slots = sizeof(a->formats) / sizeof(a->formats[0]);
	best_bid = -1;
	best_bid_slot = -1;

	/* Set up a->format for convenience of bidders. */
	a->format = &(a->formats[0]);
	for (i = 0; i < slots; i++, a->format++) {
		if (a->format->bid) {
			bid = (a->format->bid)(a, best_bid);
			if (bid == ARCHIVE_FATAL)
				return (ARCHIVE_FATAL);
			if (a->filter->position != 0)
				__archive_read_seek(a, 0, SEEK_SET);
			if ((bid > best_bid) || (best_bid_slot < 0)) {
				best_bid = bid;
				best_bid_slot = i;
			}
		}
	}

	/*
	 * There were no bidders; this is a serious programmer error
	 * and demands a quick and definitive abort.
	 */
	if (best_bid_slot < 0) {
		archive_set_error(&a->archive, ARCHIVE_ERRNO_FILE_FORMAT,
		    "No formats registered");
		return (ARCHIVE_FATAL);
	}

	/*
	 * There were bidders, but no non-zero bids; this means we
	 * can't support this stream.
	 */
	if (best_bid < 1) {
		archive_set_error(&a->archive, ARCHIVE_ERRNO_FILE_FORMAT,
		    "Unrecognized archive format");
		return (ARCHIVE_FATAL);
	}

	return (best_bid_slot);
}

/*
 * Return the file offset (within the uncompressed data stream) where
 * the last header started.
 */
int64_t
archive_read_header_position(struct archive *_a)
{
	struct archive_read *a = (struct archive_read *)_a;
	archive_check_magic(_a, ARCHIVE_READ_MAGIC,
	    ARCHIVE_STATE_ANY, "archive_read_header_position");
	return (a->header_position);
}

/*
 * Returns 1 if the archive contains at least one encrypted entry.
 * If the archive format not support encryption at all
 * ARCHIVE_READ_FORMAT_ENCRYPTION_UNSUPPORTED is returned.
 * If for any other reason (e.g. not enough data read so far)
 * we cannot say whether there are encrypted entries, then
 * ARCHIVE_READ_FORMAT_ENCRYPTION_DONT_KNOW is returned.
 * In general, this function will return values below zero when the
 * reader is uncertain or totally uncapable of encryption support.
 * When this function returns 0 you can be sure that the reader
 * supports encryption detection but no encrypted entries have
 * been found yet.
 *
 * NOTE: If the metadata/header of an archive is also encrypted, you
 * cannot rely on the number of encrypted entries. That is why this
 * function does not return the number of encrypted entries but#
 * just shows that there are some.
 */
int
archive_read_has_encrypted_entries(struct archive *_a)
{
	struct archive_read *a = (struct archive_read *)_a;
	int format_supports_encryption = archive_read_format_capabilities(_a)
			& (ARCHIVE_READ_FORMAT_CAPS_ENCRYPT_DATA | ARCHIVE_READ_FORMAT_CAPS_ENCRYPT_METADATA);

	if (!_a || !format_supports_encryption) {
		/* Format in general doesn't support encryption */
		return ARCHIVE_READ_FORMAT_ENCRYPTION_UNSUPPORTED;
	}

	/* A reader potentially has read enough data now. */
	if (a->format && a->format->has_encrypted_entries) {
		return (a->format->has_encrypted_entries)(a);
	}

	/* For any other reason we cannot say how many entries are there. */
	return ARCHIVE_READ_FORMAT_ENCRYPTION_DONT_KNOW;
}

/*
 * Returns a bitmask of capabilities that are supported by the archive format reader.
 * If the reader has no special capabilities, ARCHIVE_READ_FORMAT_CAPS_NONE is returned.
 */
int
archive_read_format_capabilities(struct archive *_a)
{
	struct archive_read *a = (struct archive_read *)_a;
	if (a && a->format && a->format->format_capabilties) {
		return (a->format->format_capabilties)(a);
	}
	return ARCHIVE_READ_FORMAT_CAPS_NONE;
}

/*
 * Read data from an archive entry, using a read(2)-style interface.
 * This is a convenience routine that just calls
 * archive_read_data_block and copies the results into the client
 * buffer, filling any gaps with zero bytes.  Clients using this
 * API can be completely ignorant of sparse-file issues; sparse files
 * will simply be padded with nulls.
 *
 * DO NOT intermingle calls to this function and archive_read_data_block
 * to read a single entry body.
 */
ssize_t
archive_read_data(struct archive *_a, void *buff, size_t s)
{
	struct archive_read *a = (struct archive_read *)_a;
	char	*dest;
	const void *read_buf;
	size_t	 bytes_read;
	size_t	 len;
	int	 r;

	bytes_read = 0;
	dest = (char *)buff;

	while (s > 0) {
		if (a->read_data_remaining == 0) {
			read_buf = a->read_data_block;
			a->read_data_is_posix_read = 1;
			a->read_data_requested = s;
			r = _archive_read_data_block(&a->archive, &read_buf,
			    &a->read_data_remaining, &a->read_data_offset);
			a->read_data_block = read_buf;
			if (r == ARCHIVE_EOF)
				return (bytes_read);
			/*
			 * Error codes are all negative, so the status
			 * return here cannot be confused with a valid
			 * byte count.  (ARCHIVE_OK is zero.)
			 */
			if (r < ARCHIVE_OK)
				return (r);
		}

		if (a->read_data_offset < a->read_data_output_offset) {
			archive_set_error(&a->archive, ARCHIVE_ERRNO_FILE_FORMAT,
			    "Encountered out-of-order sparse blocks");
			return (ARCHIVE_RETRY);
		}

		/* Compute the amount of zero padding needed. */
		if (a->read_data_output_offset + (int64_t)s <
		    a->read_data_offset) {
			len = s;
		} else if (a->read_data_output_offset <
		    a->read_data_offset) {
			len = (size_t)(a->read_data_offset -
			    a->read_data_output_offset);
		} else
			len = 0;

		/* Add zeroes. */
		memset(dest, 0, len);
		s -= len;
		a->read_data_output_offset += len;
		dest += len;
		bytes_read += len;

		/* Copy data if there is any space left. */
		if (s > 0) {
			len = a->read_data_remaining;
			if (len > s)
				len = s;
			memcpy(dest, a->read_data_block, len);
			s -= len;
			a->read_data_block += len;
			a->read_data_remaining -= len;
			a->read_data_output_offset += len;
			a->read_data_offset += len;
			dest += len;
			bytes_read += len;
		}
	}
	a->read_data_is_posix_read = 0;
	a->read_data_requested = 0;
	return (bytes_read);
}

/*
 * Skip over all remaining data in this entry.
 */
int
archive_read_data_skip(struct archive *_a)
{
	struct archive_read *a = (struct archive_read *)_a;
	int r;
	const void *buff;
	size_t size;
	int64_t offset;

	archive_check_magic(_a, ARCHIVE_READ_MAGIC, ARCHIVE_STATE_DATA,
	    "archive_read_data_skip");

	if (a->format->read_data_skip != NULL)
		r = (a->format->read_data_skip)(a);
	else {
		while ((r = archive_read_data_block(&a->archive,
			    &buff, &size, &offset))
		    == ARCHIVE_OK)
			;
	}

	if (r == ARCHIVE_EOF)
		r = ARCHIVE_OK;

	a->archive.state = ARCHIVE_STATE_HEADER;
	return (r);
}

int64_t
archive_seek_data(struct archive *_a, int64_t offset, int whence)
{
	struct archive_read *a = (struct archive_read *)_a;
	archive_check_magic(_a, ARCHIVE_READ_MAGIC, ARCHIVE_STATE_DATA,
	    "archive_seek_data_block");

	if (a->format->seek_data == NULL) {
		archive_set_error(&a->archive, ARCHIVE_ERRNO_PROGRAMMER,
		    "Internal error: "
		    "No format_seek_data_block function registered");
		return (ARCHIVE_FATAL);
	}

	return (a->format->seek_data)(a, offset, whence);
}

/*
 * Read the next block of entry data from the archive.
 * This is a zero-copy interface; the client receives a pointer,
 * size, and file offset of the next available block of data.
 *
 * Returns ARCHIVE_OK if the operation is successful, ARCHIVE_EOF if
 * the end of entry is encountered.
 */
static int
_archive_read_data_block(struct archive *_a,
    const void **buff, size_t *size, int64_t *offset)
{
	struct archive_read *a = (struct archive_read *)_a;
	archive_check_magic(_a, ARCHIVE_READ_MAGIC, ARCHIVE_STATE_DATA,
	    "archive_read_data_block");

	if (a->format->read_data == NULL) {
		archive_set_error(&a->archive, ARCHIVE_ERRNO_PROGRAMMER,
		    "Internal error: "
		    "No format_read_data_block function registered");
		return (ARCHIVE_FATAL);
	}

	return (a->format->read_data)(a, buff, size, offset);
}

int
__archive_read_close_filters(struct archive_read *a)
{
	struct archive_read_filter *f = a->filter;
	int r = ARCHIVE_OK;
	/* Close each filter in the pipeline. */
	while (f != NULL) {
		struct archive_read_filter *t = f->upstream;
		if (!f->closed && f->close != NULL) {
			int r1 = (f->close)(f);
			f->closed = 1;
			if (r1 < r)
				r = r1;
		}
		free(f->buffer);
		f->buffer = NULL;
		f = t;
	}
	return r;
}

void
__archive_read_free_filters(struct archive_read *a)
{
	while (a->filter != NULL) {
		struct archive_read_filter *t = a->filter->upstream;
		free(a->filter);
		a->filter = t;
	}
}

/*
 * return the count of # of filters in use
 */
static int
_archive_filter_count(struct archive *_a)
{
	struct archive_read *a = (struct archive_read *)_a;
	struct archive_read_filter *p = a->filter;
	int count = 0;
	while(p) {
		count++;
		p = p->upstream;
	}
	return count;
}

/*
 * Close the file and all I/O.
 */
static int
_archive_read_close(struct archive *_a)
{
	struct archive_read *a = (struct archive_read *)_a;
	int r = ARCHIVE_OK, r1 = ARCHIVE_OK;

	archive_check_magic(&a->archive, ARCHIVE_READ_MAGIC,
	    ARCHIVE_STATE_ANY | ARCHIVE_STATE_FATAL, "archive_read_close");
	if (a->archive.state == ARCHIVE_STATE_CLOSED)
		return (ARCHIVE_OK);
	archive_clear_error(&a->archive);
	a->archive.state = ARCHIVE_STATE_CLOSED;

	/* TODO: Clean up the formatters. */

	/* Release the filter objects. */
	r1 = __archive_read_close_filters(a);
	if (r1 < r)
		r = r1;

	return (r);
}

/*
 * Release memory and other resources.
 */
static int
_archive_read_free(struct archive *_a)
{
	struct archive_read *a = (struct archive_read *)_a;
	int i, n;
	int slots;
	int r = ARCHIVE_OK;

	if (_a == NULL)
		return (ARCHIVE_OK);
	archive_check_magic(_a, ARCHIVE_READ_MAGIC,
	    ARCHIVE_STATE_ANY | ARCHIVE_STATE_FATAL, "archive_read_free");
	if (a->archive.state != ARCHIVE_STATE_CLOSED
	    && a->archive.state != ARCHIVE_STATE_FATAL)
		r = archive_read_close(&a->archive);

	/* Call cleanup functions registered by optional components. */
	if (a->cleanup_archive_extract != NULL)
		r = (a->cleanup_archive_extract)(a);

	/* Cleanup format-specific data. */
	slots = sizeof(a->formats) / sizeof(a->formats[0]);
	for (i = 0; i < slots; i++) {
		a->format = &(a->formats[i]);
		if (a->formats[i].cleanup)
			(a->formats[i].cleanup)(a);
	}

	/* Free the filters */
	__archive_read_free_filters(a);

	/* Release the bidder objects. */
	n = sizeof(a->bidders)/sizeof(a->bidders[0]);
	for (i = 0; i < n; i++) {
		if (a->bidders[i].free != NULL) {
			int r1 = (a->bidders[i].free)(&a->bidders[i]);
			if (r1 < r)
				r = r1;
		}
	}

	archive_string_free(&a->archive.error_string);
	if (a->entry)
		archive_entry_free(a->entry);
	a->archive.magic = 0;
	__archive_clean(&a->archive);
	free(a->client.dataset);
	free(a);
	return (r);
}

static struct archive_read_filter *
get_filter(struct archive *_a, int n)
{
	struct archive_read *a = (struct archive_read *)_a;
	struct archive_read_filter *f = a->filter;
	/* We use n == -1 for 'the last filter', which is always the
	 * client proxy. */
	if (n == -1 && f != NULL) {
		struct archive_read_filter *last = f;
		f = f->upstream;
		while (f != NULL) {
			last = f;
			f = f->upstream;
		}
		return (last);
	}
	if (n < 0)
		return NULL;
	while (n > 0 && f != NULL) {
		f = f->upstream;
		--n;
	}
	return (f);
}

static int
_archive_filter_code(struct archive *_a, int n)
{
	struct archive_read_filter *f = get_filter(_a, n);
	return f == NULL ? -1 : f->code;
}

static const char *
_archive_filter_name(struct archive *_a, int n)
{
	struct archive_read_filter *f = get_filter(_a, n);
	return f != NULL ? f->name : NULL;
}

static int64_t
_archive_filter_bytes(struct archive *_a, int n)
{
	struct archive_read_filter *f = get_filter(_a, n);
	return f == NULL ? -1 : f->position;
}

/*
 * Used internally by read format handlers to register their bid and
 * initialization functions.
 */
int
__archive_read_register_format(struct archive_read *a,
    void *format_data,
    const char *name,
    int (*bid)(struct archive_read *, int),
    int (*options)(struct archive_read *, const char *, const char *),
    int (*read_header)(struct archive_read *, struct archive_entry *),
    int (*read_data)(struct archive_read *, const void **, size_t *, int64_t *),
    int (*read_data_skip)(struct archive_read *),
    int64_t (*seek_data)(struct archive_read *, int64_t, int),
    int (*cleanup)(struct archive_read *),
    int (*format_capabilities)(struct archive_read *),
    int (*has_encrypted_entries)(struct archive_read *))
{
	int i, number_slots;

	archive_check_magic(&a->archive,
	    ARCHIVE_READ_MAGIC, ARCHIVE_STATE_NEW,
	    "__archive_read_register_format");

	number_slots = sizeof(a->formats) / sizeof(a->formats[0]);

	for (i = 0; i < number_slots; i++) {
		if (a->formats[i].bid == bid)
			return (ARCHIVE_WARN); /* We've already installed */
		if (a->formats[i].bid == NULL) {
			a->formats[i].bid = bid;
			a->formats[i].options = options;
			a->formats[i].read_header = read_header;
			a->formats[i].read_data = read_data;
			a->formats[i].read_data_skip = read_data_skip;
			a->formats[i].seek_data = seek_data;
			a->formats[i].cleanup = cleanup;
			a->formats[i].data = format_data;
			a->formats[i].name = name;
			a->formats[i].format_capabilties = format_capabilities;
			a->formats[i].has_encrypted_entries = has_encrypted_entries;
			return (ARCHIVE_OK);
		}
	}

	archive_set_error(&a->archive, ENOMEM,
	    "Not enough slots for format registration");
	return (ARCHIVE_FATAL);
}

/*
 * Used internally by decompression routines to register their bid and
 * initialization functions.
 */
int
__archive_read_get_bidder(struct archive_read *a,
    struct archive_read_filter_bidder **bidder)
{
	int i, number_slots;

	number_slots = sizeof(a->bidders) / sizeof(a->bidders[0]);

	for (i = 0; i < number_slots; i++) {
		if (a->bidders[i].bid == NULL) {
			memset(a->bidders + i, 0, sizeof(a->bidders[0]));
			*bidder = (a->bidders + i);
			return (ARCHIVE_OK);
		}
	}

	archive_set_error(&a->archive, ENOMEM,
	    "Not enough slots for filter registration");
	return (ARCHIVE_FATAL);
}

/*
 * The next section implements the peek/consume internal I/O
 * system used by archive readers.  This system allows simple
 * read-ahead for consumers while preserving zero-copy operation
 * most of the time.
 *
 * The two key operations:
 *  * The read-ahead function returns a pointer to a block of data
 *    that satisfies a minimum request.
 *  * The consume function advances the file pointer.
 *
 * In the ideal case, filters generate blocks of data
 * and __archive_read_ahead() just returns pointers directly into
 * those blocks.  Then __archive_read_consume() just bumps those
 * pointers.  Only if your request would span blocks does the I/O
 * layer use a copy buffer to provide you with a contiguous block of
 * data.
 *
 * A couple of useful idioms:
 *  * "I just want some data."  Ask for 1 byte and pay attention to
 *    the "number of bytes available" from __archive_read_ahead().
 *    Consume whatever you actually use.
 *  * "I want to output a large block of data."  As above, ask for 1 byte,
 *    emit all that's available (up to whatever limit you have), consume
 *    it all, then repeat until you're done.  This effectively means that
 *    you're passing along the blocks that came from your provider.
 *  * "I want to peek ahead by a large amount."  Ask for 4k or so, then
 *    double and repeat until you get an error or have enough.  Note
 *    that the I/O layer will likely end up expanding its copy buffer
 *    to fit your request, so use this technique cautiously.  This
 *    technique is used, for example, by some of the format tasting
 *    code that has uncertain look-ahead needs.
 */

/*
 * Looks ahead in the input stream:
 *  * If 'avail' pointer is provided, that returns number of bytes available
 *    in the current buffer, which may be much larger than requested.
 *  * If end-of-file, *avail gets set to zero.
 *  * If error, *avail gets error code.
 *  * If request can be met, returns pointer to data.
 *  * If minimum request cannot be met, returns NULL.
 *
 * Note: If you just want "some data", ask for 1 byte and pay attention
 * to *avail, which will have the actual amount available.  If you
 * know exactly how many bytes you need, just ask for that and treat
 * a NULL return as an error.
 *
 * Important:  This does NOT move the file pointer.  See
 * __archive_read_consume() below.
 */
const void *
__archive_read_ahead(struct archive_read *a, size_t min, ssize_t *avail)
{
	return (__archive_read_filter_ahead(a->filter, min, avail));
}

const void *
__archive_read_filter_ahead(struct archive_read_filter *filter,
    size_t min, ssize_t *avail)
{
	ssize_t bytes_read;
	size_t tocopy;

	if (filter->fatal) {
		if (avail)
			*avail = ARCHIVE_FATAL;
		return (NULL);
	}

	/*
	 * Keep pulling more data until we can satisfy the request.
	 */
	for (;;) {

		/*
		 * If we can satisfy from the copy buffer (and the
		 * copy buffer isn't empty), we're done.  In particular,
		 * note that min == 0 is a perfectly well-defined
		 * request.
		 */
		if (filter->avail >= min && filter->avail > 0) {
			if (avail != NULL)
				*avail = filter->avail;
			return (filter->next);
		}

		/*
		 * We can satisfy directly from client buffer if everything
		 * currently in the copy buffer is still in the client buffer.
		 */
		if (filter->client_total >= filter->client_avail + filter->avail
		    && filter->client_avail + filter->avail >= min) {
			/* "Roll back" to client buffer. */
			filter->client_avail += filter->avail;
			filter->client_next -= filter->avail;
			/* Copy buffer is now empty. */
			filter->avail = 0;
			filter->next = filter->buffer;
			/* Return data from client buffer. */
			if (avail != NULL)
				*avail = filter->client_avail;
			return (filter->client_next);
		}

		/* Move data forward in copy buffer if necessary. */
		if (filter->next > filter->buffer &&
		    filter->next + min > filter->buffer + filter->buffer_size) {
			if (filter->avail > 0)
				memmove(filter->buffer, filter->next,
				    filter->avail);
			filter->next = filter->buffer;
		}

		/* If we've used up the client data, get more. */
		if (filter->client_avail <= 0) {
			if (filter->end_of_file) {
				if (avail != NULL)
					*avail = 0;
				return (NULL);
			}
			bytes_read = (filter->read)(filter,
			    &filter->client_buff);
			if (bytes_read < 0) {		/* Read error. */
				filter->client_total = filter->client_avail = 0;
				filter->client_next =
				    filter->client_buff = NULL;
				filter->fatal = 1;
				if (avail != NULL)
					*avail = ARCHIVE_FATAL;
				return (NULL);
			}
			if (bytes_read == 0) {
				/* Check for another client object first */
				if (filter->archive->client.cursor !=
				      filter->archive->client.nodes - 1) {
					if (client_switch_proxy(filter,
					    filter->archive->client.cursor + 1)
					    == ARCHIVE_OK)
						continue;
				}
				/* Premature end-of-file. */
				filter->client_total = filter->client_avail = 0;
				filter->client_next =
				    filter->client_buff = NULL;
				filter->end_of_file = 1;
				/* Return whatever we do have. */
				if (avail != NULL)
					*avail = filter->avail;
				return (NULL);
			}
			filter->client_total = bytes_read;
			filter->client_avail = filter->client_total;
			filter->client_next = filter->client_buff;
		} else {
			/*
			 * We can't satisfy the request from the copy
			 * buffer or the existing client data, so we
			 * need to copy more client data over to the
			 * copy buffer.
			 */

			/* Ensure the buffer is big enough. */
			if (min > filter->buffer_size) {
				size_t s, t;
				char *p;

				/* Double the buffer; watch for overflow. */
				s = t = filter->buffer_size;
				if (s == 0)
					s = min;
				while (s < min) {
					t *= 2;
					if (t <= s) { /* Integer overflow! */
						archive_set_error(
						    &filter->archive->archive,
						    ENOMEM,
						    "Unable to allocate copy"
						    " buffer");
						filter->fatal = 1;
						if (avail != NULL)
							*avail = ARCHIVE_FATAL;
						return (NULL);
					}
					s = t;
				}
				/* Now s >= min, so allocate a new buffer. */
				p = (char *)malloc(s);
				if (p == NULL) {
					archive_set_error(
						&filter->archive->archive,
						ENOMEM,
					    "Unable to allocate copy buffer");
					filter->fatal = 1;
					if (avail != NULL)
						*avail = ARCHIVE_FATAL;
					return (NULL);
				}
				/* Move data into newly-enlarged buffer. */
				if (filter->avail > 0)
					memmove(p, filter->next, filter->avail);
				free(filter->buffer);
				filter->next = filter->buffer = p;
				filter->buffer_size = s;
			}

			/* We can add client data to copy buffer. */
			/* First estimate: copy to fill rest of buffer. */
			tocopy = (filter->buffer + filter->buffer_size)
			    - (filter->next + filter->avail);
			/* Don't waste time buffering more than we need to. */
			if (tocopy + filter->avail > min)
				tocopy = min - filter->avail;
			/* Don't copy more than is available. */
			if (tocopy > filter->client_avail)
				tocopy = filter->client_avail;

			memcpy(filter->next + filter->avail,
			    filter->client_next, tocopy);
			/* Remove this data from client buffer. */
			filter->client_next += tocopy;
			filter->client_avail -= tocopy;
			/* add it to copy buffer. */
			filter->avail += tocopy;
		}
	}
}

/*
 * Move the file pointer forward.
 */
int64_t
__archive_read_consume(struct archive_read *a, int64_t request)
{
	return (__archive_read_filter_consume(a->filter, request));
}

int64_t
__archive_read_filter_consume(struct archive_read_filter * filter,
    int64_t request)
{
	int64_t skipped;

	if (request == 0)
		return 0;

	skipped = advance_file_pointer(filter, request);
	if (skipped == request)
		return (skipped);
	/* We hit EOF before we satisfied the skip request. */
	if (skipped < 0)  /* Map error code to 0 for error message below. */
		skipped = 0;
	archive_set_error(&filter->archive->archive,
	    ARCHIVE_ERRNO_MISC,
	    "Truncated input file (needed %jd bytes, only %jd available)",
	    (intmax_t)request, (intmax_t)skipped);
	return (ARCHIVE_FATAL);
}

/*
 * Advance the file pointer by the amount requested.
 * Returns the amount actually advanced, which may be less than the
 * request if EOF is encountered first.
 * Returns a negative value if there's an I/O error.
 */
static int64_t
advance_file_pointer(struct archive_read_filter *filter, int64_t request)
{
	int64_t bytes_skipped, total_bytes_skipped = 0;
	ssize_t bytes_read;
	size_t min;

	if (filter->fatal)
		return (-1);

	/* Use up the copy buffer first. */
	if (filter->avail > 0) {
		min = (size_t)minimum(request, (int64_t)filter->avail);
		filter->next += min;
		filter->avail -= min;
		request -= min;
		filter->position += min;
		total_bytes_skipped += min;
	}

	/* Then use up the client buffer. */
	if (filter->client_avail > 0) {
		min = (size_t)minimum(request, (int64_t)filter->client_avail);
		filter->client_next += min;
		filter->client_avail -= min;
		request -= min;
		filter->position += min;
		total_bytes_skipped += min;
	}
	if (request == 0)
		return (total_bytes_skipped);

	/* If there's an optimized skip function, use it. */
	if (filter->skip != NULL) {
		bytes_skipped = (filter->skip)(filter, request);
		if (bytes_skipped < 0) {	/* error */
			filter->fatal = 1;
			return (bytes_skipped);
		}
		filter->position += bytes_skipped;
		total_bytes_skipped += bytes_skipped;
		request -= bytes_skipped;
		if (request == 0)
			return (total_bytes_skipped);
	}

	/* Use ordinary reads as necessary to complete the request. */
	for (;;) {
		bytes_read = (filter->read)(filter, &filter->client_buff);
		if (bytes_read < 0) {
			filter->client_buff = NULL;
			filter->fatal = 1;
			return (bytes_read);
		}

		if (bytes_read == 0) {
			if (filter->archive->client.cursor !=
			      filter->archive->client.nodes - 1) {
				if (client_switch_proxy(filter,
				    filter->archive->client.cursor + 1)
				    == ARCHIVE_OK)
					continue;
			}
			filter->client_buff = NULL;
			filter->end_of_file = 1;
			return (total_bytes_skipped);
		}

		if (bytes_read >= request) {
			filter->client_next =
			    ((const char *)filter->client_buff) + request;
			filter->client_avail = (size_t)(bytes_read - request);
			filter->client_total = bytes_read;
			total_bytes_skipped += request;
			filter->position += request;
			return (total_bytes_skipped);
		}

		filter->position += bytes_read;
		total_bytes_skipped += bytes_read;
		request -= bytes_read;
	}
}

/**
 * Returns ARCHIVE_FAILED if seeking isn't supported.
 */
int64_t
__archive_read_seek(struct archive_read *a, int64_t offset, int whence)
{
	return __archive_read_filter_seek(a->filter, offset, whence);
}

int64_t
__archive_read_filter_seek(struct archive_read_filter *filter, int64_t offset,
    int whence)
{
	struct archive_read_client *client;
	int64_t r;
	unsigned int cursor;

	if (filter->closed || filter->fatal)
		return (ARCHIVE_FATAL);
	if (filter->seek == NULL)
		return (ARCHIVE_FAILED);

	client = &(filter->archive->client);
	switch (whence) {
	case SEEK_CUR:
		/* Adjust the offset and use SEEK_SET instead */
		offset += filter->position;			
	case SEEK_SET:
		cursor = 0;
		while (1)
		{
			if (client->dataset[cursor].begin_position < 0 ||
			    client->dataset[cursor].total_size < 0 ||
			    client->dataset[cursor].begin_position +
			      client->dataset[cursor].total_size - 1 > offset ||
			    cursor + 1 >= client->nodes)
				break;
			r = client->dataset[cursor].begin_position +
				client->dataset[cursor].total_size;
			client->dataset[++cursor].begin_position = r;
		}
		while (1) {
			r = client_switch_proxy(filter, cursor);
			if (r != ARCHIVE_OK)
				return r;
			if ((r = client_seek_proxy(filter, 0, SEEK_END)) < 0)
				return r;
			client->dataset[cursor].total_size = r;
			if (client->dataset[cursor].begin_position +
			    client->dataset[cursor].total_size - 1 > offset ||
			    cursor + 1 >= client->nodes)
				break;
			r = client->dataset[cursor].begin_position +
				client->dataset[cursor].total_size;
			client->dataset[++cursor].begin_position = r;
		}
		offset -= client->dataset[cursor].begin_position;
		if (offset < 0
		    || offset > client->dataset[cursor].total_size)
			return ARCHIVE_FATAL;
		if ((r = client_seek_proxy(filter, offset, SEEK_SET)) < 0)
			return r;
		break;

	case SEEK_END:
		cursor = 0;
		while (1) {
			if (client->dataset[cursor].begin_position < 0 ||
			    client->dataset[cursor].total_size < 0 ||
			    cursor + 1 >= client->nodes)
				break;
			r = client->dataset[cursor].begin_position +
				client->dataset[cursor].total_size;
			client->dataset[++cursor].begin_position = r;
		}
		while (1) {
			r = client_switch_proxy(filter, cursor);
			if (r != ARCHIVE_OK)
				return r;
			if ((r = client_seek_proxy(filter, 0, SEEK_END)) < 0)
				return r;
			client->dataset[cursor].total_size = r;
			r = client->dataset[cursor].begin_position +
				client->dataset[cursor].total_size;
			if (cursor + 1 >= client->nodes)
				break;
			client->dataset[++cursor].begin_position = r;
		}
		while (1) {
			if (r + offset >=
			    client->dataset[cursor].begin_position)
				break;
			offset += client->dataset[cursor].total_size;
			if (cursor == 0)
				break;
			cursor--;
			r = client->dataset[cursor].begin_position +
				client->dataset[cursor].total_size;
		}
		offset = (r + offset) - client->dataset[cursor].begin_position;
		if ((r = client_switch_proxy(filter, cursor)) != ARCHIVE_OK)
			return r;
		r = client_seek_proxy(filter, offset, SEEK_SET);
		if (r < ARCHIVE_OK)
			return r;
		break;

	default:
		return (ARCHIVE_FATAL);
	}
	r += client->dataset[cursor].begin_position;

	if (r >= 0) {
		/*
		 * Ouch.  Clearing the buffer like this hurts, especially
		 * at bid time.  A lot of our efficiency at bid time comes
		 * from having bidders reuse the data we've already read.
		 *
		 * TODO: If the seek request is in data we already
		 * have, then don't call the seek callback.
		 *
		 * TODO: Zip seeks to end-of-file at bid time.  If
		 * other formats also start doing this, we may need to
		 * find a way for clients to fudge the seek offset to
		 * a block boundary.
		 *
		 * Hmmm... If whence was SEEK_END, we know the file
		 * size is (r - offset).  Can we use that to simplify
		 * the TODO items above?
		 */
		filter->avail = filter->client_avail = 0;
		filter->next = filter->buffer;
		filter->position = r;
		filter->end_of_file = 0;
	}
	return r;
}