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
|
/***************************************************************************
* _ _ ____ _
* Project ___| | | | _ \| |
* / __| | | | |_) | |
* | (__| |_| | _ <| |___
* \___|\___/|_| \_\_____|
*
* Copyright (C) 1998 - 2007, Daniel Stenberg, <daniel@haxx.se>, et al.
*
* This software is licensed as described in the file COPYING, which
* you should have received as part of this distribution. The terms
* are also available at http://curl.haxx.se/docs/copyright.html.
*
* You may opt to use, copy, modify, merge, publish, distribute and/or sell
* copies of the Software, and permit persons to whom the Software is
* furnished to do so, under the terms of the COPYING file.
*
* This software is distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY
* KIND, either express or implied.
*
* $Id$
***************************************************************************/
#include "setup.h"
/* -- WIN32 approved -- */
#include <stdio.h>
#include <string.h>
#include <stdarg.h>
#include <stdlib.h>
#include <ctype.h>
#include <errno.h>
#include "strtoofft.h"
#include "strequal.h"
#ifdef WIN32
#include <time.h>
#include <io.h>
#else
#ifdef HAVE_SYS_SOCKET_H
#include <sys/socket.h>
#endif
#ifdef HAVE_NETINET_IN_H
#include <netinet/in.h>
#endif
#ifdef HAVE_SYS_TIME_H
#include <sys/time.h>
#endif
#ifdef HAVE_UNISTD_H
#include <unistd.h>
#endif
#include <netdb.h>
#ifdef HAVE_ARPA_INET_H
#include <arpa/inet.h>
#endif
#ifdef HAVE_NET_IF_H
#include <net/if.h>
#endif
#ifdef HAVE_SYS_IOCTL_H
#include <sys/ioctl.h>
#endif
#include <signal.h>
#ifdef HAVE_SYS_PARAM_H
#include <sys/param.h>
#endif
#ifdef HAVE_SYS_SELECT_H
#include <sys/select.h>
#endif
#ifndef HAVE_SOCKET
#error "We can't compile without socket() support!"
#endif
#endif /* WIN32 */
#include "urldata.h"
#include <curl/curl.h>
#include "netrc.h"
#include "content_encoding.h"
#include "hostip.h"
#include "transfer.h"
#include "sendf.h"
#include "speedcheck.h"
#include "progress.h"
#include "http.h"
#include "url.h"
#include "getinfo.h"
#include "sslgen.h"
#include "http_digest.h"
#include "http_ntlm.h"
#include "http_negotiate.h"
#include "share.h"
#include "memory.h"
#include "select.h"
#include "multiif.h"
#include "easyif.h" /* for Curl_convert_to_network prototype */
#define _MPRINTF_REPLACE /* use our functions only */
#include <curl/mprintf.h>
/* The last #include file should be: */
#include "memdebug.h"
#define CURL_TIMEOUT_EXPECT_100 1000 /* counting ms here */
/*
* This function will call the read callback to fill our buffer with data
* to upload.
*/
CURLcode Curl_fillreadbuffer(struct connectdata *conn, int bytes, int *nreadp)
{
struct SessionHandle *data = conn->data;
size_t buffersize = (size_t)bytes;
int nread;
if(conn->bits.upload_chunky) {
/* if chunked Transfer-Encoding */
buffersize -= (8 + 2 + 2); /* 32bit hex + CRLF + CRLF */
data->reqdata.upload_fromhere += 10; /* 32bit hex + CRLF */
}
/* this function returns a size_t, so we typecast to int to prevent warnings
with picky compilers */
nread = (int)conn->fread(data->reqdata.upload_fromhere, 1,
buffersize, conn->fread_in);
if(nread == CURL_READFUNC_ABORT) {
failf(data, "operation aborted by callback\n");
return CURLE_ABORTED_BY_CALLBACK;
}
if(!conn->bits.forbidchunk && conn->bits.upload_chunky) {
/* if chunked Transfer-Encoding */
char hexbuffer[11];
int hexlen = snprintf(hexbuffer, sizeof(hexbuffer),
"%x\r\n", nread);
/* move buffer pointer */
data->reqdata.upload_fromhere -= hexlen;
nread += hexlen;
/* copy the prefix to the buffer */
memcpy(data->reqdata.upload_fromhere, hexbuffer, hexlen);
/* always append CRLF to the data */
memcpy(data->reqdata.upload_fromhere + nread, "\r\n", 2);
if((nread - hexlen) == 0) {
/* mark this as done once this chunk is transfered */
data->reqdata.keep.upload_done = TRUE;
}
nread+=2; /* for the added CRLF */
}
*nreadp = nread;
#ifdef CURL_DOES_CONVERSIONS
if(data->set.prefer_ascii) {
CURLcode res;
res = Curl_convert_to_network(data, data->reqdata.upload_fromhere, nread);
/* Curl_convert_to_network calls failf if unsuccessful */
if(res != CURLE_OK) {
return(res);
}
}
#endif /* CURL_DOES_CONVERSIONS */
return CURLE_OK;
}
/*
* checkhttpprefix()
*
* Returns TRUE if member of the list matches prefix of string
*/
static bool
checkhttpprefix(struct SessionHandle *data,
const char *s)
{
struct curl_slist *head = data->set.http200aliases;
bool rc = FALSE;
#ifdef CURL_DOES_CONVERSIONS
/* convert from the network encoding using a scratch area */
char *scratch = calloc(1, strlen(s)+1);
if (NULL == scratch) {
failf (data, "Failed to calloc memory for conversion!");
return FALSE; /* can't return CURLE_OUT_OF_MEMORY so return FALSE */
}
strcpy(scratch, s);
if (CURLE_OK != Curl_convert_from_network(data, scratch, strlen(s)+1)) {
/* Curl_convert_from_network calls failf if unsuccessful */
free(scratch);
return FALSE; /* can't return CURLE_foobar so return FALSE */
}
s = scratch;
#endif /* CURL_DOES_CONVERSIONS */
while (head) {
if (checkprefix(head->data, s)) {
rc = TRUE;
break;
}
head = head->next;
}
if ((rc != TRUE) && (checkprefix("HTTP/", s))) {
rc = TRUE;
}
#ifdef CURL_DOES_CONVERSIONS
free(scratch);
#endif /* CURL_DOES_CONVERSIONS */
return rc;
}
/*
* Curl_readrewind() rewinds the read stream. This typically (so far) only
* used for HTTP POST/PUT with multi-pass authentication when a sending was
* denied and a resend is necessary.
*/
CURLcode Curl_readrewind(struct connectdata *conn)
{
struct SessionHandle *data = conn->data;
conn->bits.rewindaftersend = FALSE; /* we rewind now */
/* We have sent away data. If not using CURLOPT_POSTFIELDS or
CURLOPT_HTTPPOST, call app to rewind
*/
if(data->set.postfields ||
(data->set.httpreq == HTTPREQ_POST_FORM))
; /* do nothing */
else {
if(data->set.ioctl) {
curlioerr err;
err = (data->set.ioctl) (data, CURLIOCMD_RESTARTREAD,
data->set.ioctl_client);
infof(data, "the ioctl callback returned %d\n", (int)err);
if(err) {
/* FIXME: convert to a human readable error message */
failf(data, "ioctl callback returned error %d\n", (int)err);
return CURLE_SEND_FAIL_REWIND;
}
}
else {
/* If no CURLOPT_READFUNCTION is used, we know that we operate on a
given FILE * stream and we can actually attempt to rewind that
ourself with fseek() */
if(data->set.fread == (curl_read_callback)fread) {
if(-1 != fseek(data->set.in, 0, SEEK_SET))
/* successful rewind */
return CURLE_OK;
}
/* no callback set or failure aboe, makes us fail at once */
failf(data, "necessary data rewind wasn't possible\n");
return CURLE_SEND_FAIL_REWIND;
}
}
return CURLE_OK;
}
static int data_pending(struct connectdata *conn)
{
return Curl_ssl_data_pending(conn, FIRSTSOCKET);
}
#ifndef MIN
#define MIN(a,b) (a < b ? a : b)
#endif
static void read_rewind(struct connectdata *conn,
size_t thismuch)
{
conn->read_pos -= thismuch;
conn->bits.stream_was_rewound = TRUE;
#ifdef CURLDEBUG
{
char buf[512 + 1];
size_t show;
show = MIN(conn->buf_len - conn->read_pos, sizeof(buf)-1);
memcpy(buf, conn->master_buffer + conn->read_pos, show);
buf[show] = '\0';
DEBUGF(infof(conn->data,
"Buffer after stream rewind (read_pos = %d): [%s]",
conn->read_pos, buf));
}
#endif
}
/*
* Curl_readwrite() is the low-level function to be called when data is to
* be read and written to/from the connection.
*/
CURLcode Curl_readwrite(struct connectdata *conn,
bool *done)
{
struct SessionHandle *data = conn->data;
struct Curl_transfer_keeper *k = &data->reqdata.keep;
CURLcode result;
ssize_t nread; /* number of bytes read */
int didwhat=0;
curl_socket_t fd_read;
curl_socket_t fd_write;
int select_res;
curl_off_t contentlength;
/* only use the proper socket if the *_HOLD bit is not set simultaneously as
then we are in rate limiting state in that transfer direction */
if((k->keepon & (KEEP_READ|KEEP_READ_HOLD)) == KEEP_READ)
fd_read = conn->sockfd;
else
fd_read = CURL_SOCKET_BAD;
if((k->keepon & (KEEP_WRITE|KEEP_WRITE_HOLD)) == KEEP_WRITE)
fd_write = conn->writesockfd;
else
fd_write = CURL_SOCKET_BAD;
select_res = Curl_socket_ready(fd_read, fd_write, 0);
if(select_res == CSELECT_ERR) {
failf(data, "select/poll returned error");
return CURLE_SEND_ERROR;
}
do {
/* We go ahead and do a read if we have a readable socket or if
the stream was rewound (in which case we have data in a
buffer) */
if((k->keepon & KEEP_READ) &&
((select_res & CSELECT_IN) || conn->bits.stream_was_rewound)) {
/* read */
bool is_empty_data = FALSE;
/* This is where we loop until we have read everything there is to
read or we get a EWOULDBLOCK */
do {
size_t buffersize = data->set.buffer_size?
data->set.buffer_size : BUFSIZE;
size_t bytestoread = buffersize;
int readrc;
if (k->size != -1 && !k->header) {
/* make sure we don't read "too much" if we can help it since we
might be pipelining and then someone else might want to read what
follows! */
curl_off_t totalleft = k->size - k->bytecount;
if(totalleft < (curl_off_t)bytestoread)
bytestoread = (size_t)totalleft;
}
/* receive data from the network! */
readrc = Curl_read(conn, conn->sockfd, k->buf, bytestoread, &nread);
/* subzero, this would've blocked */
if(0 > readrc)
break; /* get out of loop */
/* get the CURLcode from the int */
result = (CURLcode)readrc;
if(result>0)
return result;
if ((k->bytecount == 0) && (k->writebytecount == 0)) {
Curl_pgrsTime(data, TIMER_STARTTRANSFER);
if(k->wait100_after_headers)
/* set time stamp to compare with when waiting for the 100 */
k->start100 = Curl_tvnow();
}
didwhat |= KEEP_READ;
/* indicates data of zero size, i.e. empty file */
is_empty_data = (bool)((nread == 0) && (k->bodywrites == 0));
/* NULL terminate, allowing string ops to be used */
if (0 < nread || is_empty_data) {
k->buf[nread] = 0;
}
else if (0 >= nread) {
/* if we receive 0 or less here, the server closed the connection
and we bail out from this! */
k->keepon &= ~KEEP_READ;
break;
}
/* Default buffer to use when we write the buffer, it may be changed
in the flow below before the actual storing is done. */
k->str = k->buf;
/* Since this is a two-state thing, we check if we are parsing
headers at the moment or not. */
if (k->header) {
/* we are in parse-the-header-mode */
bool stop_reading = FALSE;
/* header line within buffer loop */
do {
size_t hbufp_index;
size_t rest_length;
size_t full_length;
int writetype;
/* str_start is start of line within buf */
k->str_start = k->str;
/* data is in network encoding so use 0x0a instead of '\n' */
k->end_ptr = memchr(k->str_start, 0x0a, nread);
if (!k->end_ptr) {
/* Not a complete header line within buffer, append the data to
the end of the headerbuff. */
if (k->hbuflen + nread >= data->state.headersize) {
/* We enlarge the header buffer as it is too small */
char *newbuff;
size_t newsize=CURLMAX((k->hbuflen+nread)*3/2,
data->state.headersize*2);
hbufp_index = k->hbufp - data->state.headerbuff;
newbuff = (char *)realloc(data->state.headerbuff, newsize);
if(!newbuff) {
failf (data, "Failed to alloc memory for big header!");
return CURLE_OUT_OF_MEMORY;
}
data->state.headersize=newsize;
data->state.headerbuff = newbuff;
k->hbufp = data->state.headerbuff + hbufp_index;
}
memcpy(k->hbufp, k->str, nread);
k->hbufp += nread;
k->hbuflen += nread;
if (!k->headerline && (k->hbuflen>5)) {
/* make a first check that this looks like a HTTP header */
if(!checkhttpprefix(data, data->state.headerbuff)) {
/* this is not the beginning of a HTTP first header line */
k->header = FALSE;
k->badheader = HEADER_ALLBAD;
break;
}
}
break; /* read more and try again */
}
/* decrease the size of the remaining (supposed) header line */
rest_length = (k->end_ptr - k->str)+1;
nread -= (ssize_t)rest_length;
k->str = k->end_ptr + 1; /* move past new line */
full_length = k->str - k->str_start;
/*
* We're about to copy a chunk of data to the end of the
* already received header. We make sure that the full string
* fit in the allocated header buffer, or else we enlarge
* it.
*/
if (k->hbuflen + full_length >=
data->state.headersize) {
char *newbuff;
size_t newsize=CURLMAX((k->hbuflen+full_length)*3/2,
data->state.headersize*2);
hbufp_index = k->hbufp - data->state.headerbuff;
newbuff = (char *)realloc(data->state.headerbuff, newsize);
if(!newbuff) {
failf (data, "Failed to alloc memory for big header!");
return CURLE_OUT_OF_MEMORY;
}
data->state.headersize= newsize;
data->state.headerbuff = newbuff;
k->hbufp = data->state.headerbuff + hbufp_index;
}
/* copy to end of line */
memcpy(k->hbufp, k->str_start, full_length);
k->hbufp += full_length;
k->hbuflen += full_length;
*k->hbufp = 0;
k->end_ptr = k->hbufp;
k->p = data->state.headerbuff;
/****
* We now have a FULL header line that p points to
*****/
if(!k->headerline) {
/* the first read header */
if((k->hbuflen>5) &&
!checkhttpprefix(data, data->state.headerbuff)) {
/* this is not the beginning of a HTTP first header line */
k->header = FALSE;
if(nread)
/* since there's more, this is a partial bad header */
k->badheader = HEADER_PARTHEADER;
else {
/* this was all we read so its all a bad header */
k->badheader = HEADER_ALLBAD;
nread = (ssize_t)rest_length;
}
break;
}
}
/* headers are in network encoding so
use 0x0a and 0x0d instead of '\n' and '\r' */
if ((0x0a == *k->p) || (0x0d == *k->p)) {
size_t headerlen;
/* Zero-length header line means end of headers! */
#ifdef CURL_DOES_CONVERSIONS
if (0x0d == *k->p) {
*k->p = '\r'; /* replace with CR in host encoding */
k->p++; /* pass the CR byte */
}
if (0x0a == *k->p) {
*k->p = '\n'; /* replace with LF in host encoding */
k->p++; /* pass the LF byte */
}
#else
if ('\r' == *k->p)
k->p++; /* pass the \r byte */
if ('\n' == *k->p)
k->p++; /* pass the \n byte */
#endif /* CURL_DOES_CONVERSIONS */
if(100 == k->httpcode) {
/*
* We have made a HTTP PUT or POST and this is 1.1-lingo
* that tells us that the server is OK with this and ready
* to receive the data.
* However, we'll get more headers now so we must get
* back into the header-parsing state!
*/
k->header = TRUE;
k->headerline = 0; /* restart the header line counter */
/* if we did wait for this do enable write now! */
if (k->write_after_100_header) {
k->write_after_100_header = FALSE;
k->keepon |= KEEP_WRITE;
}
}
else {
k->header = FALSE; /* no more header to parse! */
if((k->size == -1) && !conn->bits.chunk && !conn->bits.close)
/* When connection is not to get closed, but no
Content-Length nor Content-Encoding chunked have been
received, there is no body in this response. We don't set
stop_reading TRUE since that would also prevent necessary
authentication actions to take place. */
conn->bits.no_body = TRUE;
}
if (417 == k->httpcode) {
/*
* we got: "417 Expectation Failed" this means:
* we have made a HTTP call and our Expect Header
* seems to cause a problem => abort the write operations
* (or prevent them from starting).
*/
k->write_after_100_header = FALSE;
k->keepon &= ~KEEP_WRITE;
}
#ifndef CURL_DISABLE_HTTP
/*
* When all the headers have been parsed, see if we should give
* up and return an error.
*/
if (Curl_http_should_fail(conn)) {
failf (data, "The requested URL returned error: %d",
k->httpcode);
return CURLE_HTTP_RETURNED_ERROR;
}
#endif /* CURL_DISABLE_HTTP */
/* now, only output this if the header AND body are requested:
*/
writetype = CLIENTWRITE_HEADER;
if (data->set.include_header)
writetype |= CLIENTWRITE_BODY;
headerlen = k->p - data->state.headerbuff;
result = Curl_client_write(conn, writetype,
data->state.headerbuff,
headerlen);
if(result)
return result;
data->info.header_size += (long)headerlen;
conn->headerbytecount += (long)headerlen;
conn->deductheadercount =
(100 == k->httpcode)?conn->headerbytecount:0;
if (data->reqdata.resume_from &&
(data->set.httpreq==HTTPREQ_GET) &&
(k->httpcode == 416)) {
/* "Requested Range Not Satisfiable" */
stop_reading = TRUE;
}
#ifndef CURL_DISABLE_HTTP
if(!stop_reading) {
/* Curl_http_auth_act() checks what authentication methods
* that are available and decides which one (if any) to
* use. It will set 'newurl' if an auth metod was picked. */
result = Curl_http_auth_act(conn);
if(result)
return result;
if(conn->bits.rewindaftersend) {
/* We rewind after a complete send, so thus we continue
sending now */
infof(data, "Keep sending data to get tossed away!\n");
k->keepon |= KEEP_WRITE;
}
}
#endif /* CURL_DISABLE_HTTP */
if(!k->header) {
/*
* really end-of-headers.
*
* If we requested a "no body", this is a good time to get
* out and return home.
*/
if(conn->bits.no_body)
stop_reading = TRUE;
else {
/* If we know the expected size of this document, we set the
maximum download size to the size of the expected
document or else, we won't know when to stop reading!
Note that we set the download maximum even if we read a
"Connection: close" header, to make sure that
"Content-Length: 0" still prevents us from attempting to
read the (missing) response-body.
*/
/* According to RFC2616 section 4.4, we MUST ignore
Content-Length: headers if we are now receiving data
using chunked Transfer-Encoding.
*/
if(conn->bits.chunk)
k->size=-1;
}
if(-1 != k->size) {
/* We do this operation even if no_body is true, since this
data might be retrieved later with curl_easy_getinfo()
and its CURLINFO_CONTENT_LENGTH_DOWNLOAD option. */
Curl_pgrsSetDownloadSize(data, k->size);
k->maxdownload = k->size;
}
/* If max download size is *zero* (nothing) we already
have nothing and can safely return ok now! */
if(0 == k->maxdownload)
stop_reading = TRUE;
if(stop_reading) {
/* we make sure that this socket isn't read more now */
k->keepon &= ~KEEP_READ;
}
if(data->set.verbose)
Curl_debug(data, CURLINFO_HEADER_IN,
k->str_start, headerlen, conn);
break; /* exit header line loop */
}
/* We continue reading headers, so reset the line-based
header parsing variables hbufp && hbuflen */
k->hbufp = data->state.headerbuff;
k->hbuflen = 0;
continue;
}
/*
* Checks for special headers coming up.
*/
if (!k->headerline++) {
/* This is the first header, it MUST be the error code line
or else we consider this to be the body right away! */
int httpversion_major;
int nc;
#ifdef CURL_DOES_CONVERSIONS
#define HEADER1 scratch
#define SCRATCHSIZE 21
CURLcode res;
char scratch[SCRATCHSIZE+1]; /* "HTTP/major.minor 123" */
/* We can't really convert this yet because we
don't know if it's the 1st header line or the body.
So we do a partial conversion into a scratch area,
leaving the data at k->p as-is.
*/
strncpy(&scratch[0], k->p, SCRATCHSIZE);
scratch[SCRATCHSIZE] = 0; /* null terminate */
res = Curl_convert_from_network(data,
&scratch[0],
SCRATCHSIZE);
if (CURLE_OK != res) {
/* Curl_convert_from_network calls failf if unsuccessful */
return res;
}
#else
#define HEADER1 k->p /* no conversion needed, just use k->p */
#endif /* CURL_DOES_CONVERSIONS */
nc = sscanf(HEADER1,
" HTTP/%d.%d %3d",
&httpversion_major,
&k->httpversion,
&k->httpcode);
if (nc==3) {
k->httpversion += 10 * httpversion_major;
}
else {
/* this is the real world, not a Nirvana
NCSA 1.5.x returns this crap when asked for HTTP/1.1
*/
nc=sscanf(HEADER1, " HTTP %3d", &k->httpcode);
k->httpversion = 10;
/* If user has set option HTTP200ALIASES,
compare header line against list of aliases
*/
if (!nc) {
if (checkhttpprefix(data, k->p)) {
nc = 1;
k->httpcode = 200;
k->httpversion =
(data->set.httpversion==CURL_HTTP_VERSION_1_0)? 10 : 11;
}
}
}
if (nc) {
data->info.httpcode = k->httpcode;
data->info.httpversion = k->httpversion;
/*
* This code executes as part of processing the header. As a
* result, it's not totally clear how to interpret the
* response code yet as that depends on what other headers may
* be present. 401 and 407 may be errors, but may be OK
* depending on how authentication is working. Other codes
* are definitely errors, so give up here.
*/
if (data->set.http_fail_on_error && (k->httpcode >= 400) &&
((k->httpcode != 401) || !data->set.userpwd) &&
((k->httpcode != 407) || !data->set.proxyuserpwd) ) {
if (data->reqdata.resume_from &&
(data->set.httpreq==HTTPREQ_GET) &&
(k->httpcode == 416)) {
/* "Requested Range Not Satisfiable", just proceed and
pretend this is no error */
}
else {
/* serious error, go home! */
failf (data, "The requested URL returned error: %d",
k->httpcode);
return CURLE_HTTP_RETURNED_ERROR;
}
}
if(k->httpversion == 10)
/* Default action for HTTP/1.0 must be to close, unless
we get one of those fancy headers that tell us the
server keeps it open for us! */
conn->bits.close = TRUE;
switch(k->httpcode) {
case 204:
/* (quote from RFC2616, section 10.2.5): The server has
* fulfilled the request but does not need to return an
* entity-body ... The 204 response MUST NOT include a
* message-body, and thus is always terminated by the first
* empty line after the header fields. */
/* FALLTHROUGH */
case 416: /* Requested Range Not Satisfiable, it has the
Content-Length: set as the "real" document but no
actual response is sent. */
case 304:
/* (quote from RFC2616, section 10.3.5): The 304 response
* MUST NOT contain a message-body, and thus is always
* terminated by the first empty line after the header
* fields. */
k->size=0;
k->maxdownload=0;
k->ignorecl = TRUE; /* ignore Content-Length headers */
break;
default:
/* nothing */
break;
}
}
else {
k->header = FALSE; /* this is not a header line */
break;
}
}
#ifdef CURL_DOES_CONVERSIONS
/* convert from the network encoding */
result = Curl_convert_from_network(data, k->p, strlen(k->p));
if (CURLE_OK != result) {
return(result);
}
/* Curl_convert_from_network calls failf if unsuccessful */
#endif /* CURL_DOES_CONVERSIONS */
/* Check for Content-Length: header lines to get size. Ignore
the header completely if we get a 416 response as then we're
resuming a document that we don't get, and this header contains
info about the true size of the document we didn't get now. */
if (!k->ignorecl && !data->set.ignorecl &&
checkprefix("Content-Length:", k->p)) {
contentlength = curlx_strtoofft(k->p+15, NULL, 10);
if (data->set.max_filesize &&
contentlength > data->set.max_filesize) {
failf(data, "Maximum file size exceeded");
return CURLE_FILESIZE_EXCEEDED;
}
if(contentlength >= 0) {
k->size = contentlength;
k->maxdownload = k->size;
}
else {
/* Negative Content-Length is really odd, and we know it
happens for example when older Apache servers send large
files */
conn->bits.close = TRUE;
infof(data, "Negative content-length: %" FORMAT_OFF_T
", closing after transfer\n", contentlength);
}
}
/* check for Content-Type: header lines to get the mime-type */
else if (checkprefix("Content-Type:", k->p)) {
char *start;
char *end;
size_t len;
/* Find the first non-space letter */
for(start=k->p+13;
*start && ISSPACE(*start);
start++)
; /* empty loop */
/* data is now in the host encoding so
use '\r' and '\n' instead of 0x0d and 0x0a */
end = strchr(start, '\r');
if(!end)
end = strchr(start, '\n');
if(end) {
/* skip all trailing space letters */
for(; ISSPACE(*end) && (end > start); end--)
; /* empty loop */
/* get length of the type */
len = end-start+1;
/* allocate memory of a cloned copy */
Curl_safefree(data->info.contenttype);
data->info.contenttype = malloc(len + 1);
if (NULL == data->info.contenttype)
return CURLE_OUT_OF_MEMORY;
/* copy the content-type string */
memcpy(data->info.contenttype, start, len);
data->info.contenttype[len] = 0; /* zero terminate */
}
}
#ifndef CURL_DISABLE_HTTP
else if((k->httpversion == 10) &&
conn->bits.httpproxy &&
Curl_compareheader(k->p,
"Proxy-Connection:", "keep-alive")) {
/*
* When a HTTP/1.0 reply comes when using a proxy, the
* 'Proxy-Connection: keep-alive' line tells us the
* connection will be kept alive for our pleasure.
* Default action for 1.0 is to close.
*/
conn->bits.close = FALSE; /* don't close when done */
infof(data, "HTTP/1.0 proxy connection set to keep alive!\n");
}
else if((k->httpversion == 11) &&
conn->bits.httpproxy &&
Curl_compareheader(k->p,
"Proxy-Connection:", "close")) {
/*
* We get a HTTP/1.1 response from a proxy and it says it'll
* close down after this transfer.
*/
conn->bits.close = TRUE; /* close when done */
infof(data, "HTTP/1.1 proxy connection set close!\n");
}
else if((k->httpversion == 10) &&
Curl_compareheader(k->p, "Connection:", "keep-alive")) {
/*
* A HTTP/1.0 reply with the 'Connection: keep-alive' line
* tells us the connection will be kept alive for our
* pleasure. Default action for 1.0 is to close.
*
* [RFC2068, section 19.7.1] */
conn->bits.close = FALSE; /* don't close when done */
infof(data, "HTTP/1.0 connection set to keep alive!\n");
}
else if (Curl_compareheader(k->p, "Connection:", "close")) {
/*
* [RFC 2616, section 8.1.2.1]
* "Connection: close" is HTTP/1.1 language and means that
* the connection will close when this request has been
* served.
*/
conn->bits.close = TRUE; /* close when done */
}
else if (Curl_compareheader(k->p,
"Transfer-Encoding:", "chunked")) {
/*
* [RFC 2616, section 3.6.1] A 'chunked' transfer encoding
* means that the server will send a series of "chunks". Each
* chunk starts with line with info (including size of the
* coming block) (terminated with CRLF), then a block of data
* with the previously mentioned size. There can be any amount
* of chunks, and a chunk-data set to zero signals the
* end-of-chunks. */
conn->bits.chunk = TRUE; /* chunks coming our way */
/* init our chunky engine */
Curl_httpchunk_init(conn);
}
else if (checkprefix("Trailer:", k->p) ||
checkprefix("Trailers:", k->p)) {
/*
* This test helps Curl_httpchunk_read() to determine to look
* for well formed trailers after the zero chunksize record. In
* this case a CRLF is required after the zero chunksize record
* when no trailers are sent, or after the last trailer record.
*
* It seems both Trailer: and Trailers: occur in the wild.
*/
conn->bits.trailerHdrPresent = TRUE;
}
else if (checkprefix("Content-Encoding:", k->p) &&
data->set.encoding) {
/*
* Process Content-Encoding. Look for the values: identity,
* gzip, deflate, compress, x-gzip and x-compress. x-gzip and
* x-compress are the same as gzip and compress. (Sec 3.5 RFC
* 2616). zlib cannot handle compress. However, errors are
* handled further down when the response body is processed
*/
char *start;
/* Find the first non-space letter */
for(start=k->p+17;
*start && ISSPACE(*start);
start++)
; /* empty loop */
/* Record the content-encoding for later use */
if (checkprefix("identity", start))
k->content_encoding = IDENTITY;
else if (checkprefix("deflate", start))
k->content_encoding = DEFLATE;
else if (checkprefix("gzip", start)
|| checkprefix("x-gzip", start))
k->content_encoding = GZIP;
else if (checkprefix("compress", start)
|| checkprefix("x-compress", start))
k->content_encoding = COMPRESS;
}
else if (checkprefix("Content-Range:", k->p)) {
/* Content-Range: bytes [num]-
Content-Range: bytes: [num]-
Content-Range: [num]-
The second format was added since Sun's webserver
JavaWebServer/1.1.1 obviously sends the header this way!
The third added since some servers use that!
*/
char *ptr = k->p + 14;
/* Move forward until first digit */
while(*ptr && !ISDIGIT(*ptr))
ptr++;
k->offset = curlx_strtoofft(ptr, NULL, 10);
if (data->reqdata.resume_from == k->offset)
/* we asked for a resume and we got it */
k->content_range = TRUE;
}
#if !defined(CURL_DISABLE_COOKIES)
else if(data->cookies &&
checkprefix("Set-Cookie:", k->p)) {
Curl_share_lock(data, CURL_LOCK_DATA_COOKIE,
CURL_LOCK_ACCESS_SINGLE);
Curl_cookie_add(data,
data->cookies, TRUE, k->p+11,
/* If there is a custom-set Host: name, use it
here, or else use real peer host name. */
conn->allocptr.cookiehost?
conn->allocptr.cookiehost:conn->host.name,
data->reqdata.path);
Curl_share_unlock(data, CURL_LOCK_DATA_COOKIE);
}
#endif
else if(checkprefix("Last-Modified:", k->p) &&
(data->set.timecondition || data->set.get_filetime) ) {
time_t secs=time(NULL);
k->timeofdoc = curl_getdate(k->p+strlen("Last-Modified:"),
&secs);
if(data->set.get_filetime)
data->info.filetime = (long)k->timeofdoc;
}
else if((checkprefix("WWW-Authenticate:", k->p) &&
(401 == k->httpcode)) ||
(checkprefix("Proxy-authenticate:", k->p) &&
(407 == k->httpcode))) {
result = Curl_http_input_auth(conn, k->httpcode, k->p);
if(result)
return result;
}
else if ((k->httpcode >= 300 && k->httpcode < 400) &&
checkprefix("Location:", k->p)) {
if(data->set.http_follow_location) {
/* this is the URL that the server advices us to get instead */
char *ptr;
char *start=k->p;
char backup;
start += 9; /* pass "Location:" */
/* Skip spaces and tabs. We do this to support multiple
white spaces after the "Location:" keyword. */
while(*start && ISSPACE(*start ))
start++;
/* Scan through the string from the end to find the last
non-space. k->end_ptr points to the actual terminating zero
letter, move pointer one letter back and start from
there. This logic strips off trailing whitespace, but keeps
any embedded whitespace. */
ptr = k->end_ptr-1;
while((ptr>=start) && ISSPACE(*ptr))
ptr--;
ptr++;
backup = *ptr; /* store the ending letter */
if(ptr != start) {
*ptr = '\0'; /* zero terminate */
data->reqdata.newurl = strdup(start); /* clone string */
*ptr = backup; /* restore ending letter */
if(!data->reqdata.newurl)
return CURLE_OUT_OF_MEMORY;
}
}
}
#endif /* CURL_DISABLE_HTTP */
/*
* End of header-checks. Write them to the client.
*/
writetype = CLIENTWRITE_HEADER;
if (data->set.include_header)
writetype |= CLIENTWRITE_BODY;
if(data->set.verbose)
Curl_debug(data, CURLINFO_HEADER_IN,
k->p, (size_t)k->hbuflen, conn);
result = Curl_client_write(conn, writetype, k->p, k->hbuflen);
if(result)
return result;
data->info.header_size += (long)k->hbuflen;
conn->headerbytecount += (long)k->hbuflen;
/* reset hbufp pointer && hbuflen */
k->hbufp = data->state.headerbuff;
k->hbuflen = 0;
}
while (!stop_reading && *k->str); /* header line within buffer */
if(stop_reading)
/* We've stopped dealing with input, get out of the do-while loop */
break;
/* We might have reached the end of the header part here, but
there might be a non-header part left in the end of the read
buffer. */
} /* end if header mode */
/* This is not an 'else if' since it may be a rest from the header
parsing, where the beginning of the buffer is headers and the end
is non-headers. */
if (k->str && !k->header && (nread > 0 || is_empty_data)) {
if(0 == k->bodywrites && !is_empty_data) {
/* These checks are only made the first time we are about to
write a piece of the body */
if(conn->protocol&PROT_HTTP) {
/* HTTP-only checks */
if (data->reqdata.newurl) {
if(conn->bits.close) {
/* Abort after the headers if "follow Location" is set
and we're set to close anyway. */
k->keepon &= ~KEEP_READ;
*done = TRUE;
return CURLE_OK;
}
/* We have a new url to load, but since we want to be able
to re-use this connection properly, we read the full
response in "ignore more" */
k->ignorebody = TRUE;
infof(data, "Ignoring the response-body\n");
}
if (data->reqdata.resume_from && !k->content_range &&
(data->set.httpreq==HTTPREQ_GET) &&
!k->ignorebody) {
/* we wanted to resume a download, although the server doesn't
* seem to support this and we did this with a GET (if it
* wasn't a GET we did a POST or PUT resume) */
failf(data, "HTTP server doesn't seem to support "
"byte ranges. Cannot resume.");
return CURLE_HTTP_RANGE_ERROR;
}
if(data->set.timecondition && !data->reqdata.range) {
/* A time condition has been set AND no ranges have been
requested. This seems to be what chapter 13.3.4 of
RFC 2616 defines to be the correct action for a
HTTP/1.1 client */
if((k->timeofdoc > 0) && (data->set.timevalue > 0)) {
switch(data->set.timecondition) {
case CURL_TIMECOND_IFMODSINCE:
default:
if(k->timeofdoc < data->set.timevalue) {
infof(data,
"The requested document is not new enough\n");
*done = TRUE;
return CURLE_OK;
}
break;
case CURL_TIMECOND_IFUNMODSINCE:
if(k->timeofdoc > data->set.timevalue) {
infof(data,
"The requested document is not old enough\n");
*done = TRUE;
return CURLE_OK;
}
break;
} /* switch */
} /* two valid time strings */
} /* we have a time condition */
} /* this is HTTP */
} /* this is the first time we write a body part */
k->bodywrites++;
/* pass data to the debug function before it gets "dechunked" */
if(data->set.verbose) {
if(k->badheader) {
Curl_debug(data, CURLINFO_DATA_IN, data->state.headerbuff,
(size_t)k->hbuflen, conn);
if(k->badheader == HEADER_PARTHEADER)
Curl_debug(data, CURLINFO_DATA_IN,
k->str, (size_t)nread, conn);
}
else
Curl_debug(data, CURLINFO_DATA_IN,
k->str, (size_t)nread, conn);
}
#ifndef CURL_DISABLE_HTTP
if(conn->bits.chunk) {
/*
* Here comes a chunked transfer flying and we need to decode this
* properly. While the name says read, this function both reads
* and writes away the data. The returned 'nread' holds the number
* of actual data it wrote to the client.
*/
CHUNKcode res =
Curl_httpchunk_read(conn, k->str, nread, &nread);
if(CHUNKE_OK < res) {
if(CHUNKE_WRITE_ERROR == res) {
failf(data, "Failed writing data");
return CURLE_WRITE_ERROR;
}
failf(data, "Received problem %d in the chunky parser", res);
return CURLE_RECV_ERROR;
}
else if(CHUNKE_STOP == res) {
size_t dataleft;
/* we're done reading chunks! */
k->keepon &= ~KEEP_READ; /* read no more */
/* There are now possibly N number of bytes at the end of the
str buffer that weren't written to the client.
We DO care about this data if we are pipelining.
Push it back to be read on the next pass. */
dataleft = data->reqdata.proto.http->chunk.dataleft;
if (dataleft != 0) {
infof(conn->data, "Leftovers after chunking. "
" Rewinding %d bytes\n",dataleft);
read_rewind(conn, dataleft);
}
}
/* If it returned OK, we just keep going */
}
#endif /* CURL_DISABLE_HTTP */
if((-1 != k->maxdownload) &&
(k->bytecount + nread >= k->maxdownload)) {
/* The 'excess' amount below can't be more than BUFSIZE which
always will fit in a size_t */
size_t excess = k->bytecount + nread - k->maxdownload;
if (excess > 0 && !k->ignorebody) {
infof(data,
"Rewinding stream by : %d"
" bytes on url %s (size = %" FORMAT_OFF_T
", maxdownload = %" FORMAT_OFF_T
", bytecount = %" FORMAT_OFF_T ", nread = %d)\n",
excess, conn->data->reqdata.path,
k->size, k->maxdownload, k->bytecount, nread);
read_rewind(conn, excess);
}
nread = (ssize_t) (k->maxdownload - k->bytecount);
if(nread < 0 ) /* this should be unusual */
nread = 0;
k->keepon &= ~KEEP_READ; /* we're done reading */
}
k->bytecount += nread;
Curl_pgrsSetDownloadCounter(data, k->bytecount);
if(!conn->bits.chunk && (nread || k->badheader || is_empty_data)) {
/* If this is chunky transfer, it was already written */
if(k->badheader && !k->ignorebody) {
/* we parsed a piece of data wrongly assuming it was a header
and now we output it as body instead */
result = Curl_client_write(conn, CLIENTWRITE_BODY,
data->state.headerbuff,
k->hbuflen);
if(result)
return result;
}
if(k->badheader < HEADER_ALLBAD) {
/* This switch handles various content encodings. If there's an
error here, be sure to check over the almost identical code
in http_chunks.c.
Make sure that ALL_CONTENT_ENCODINGS contains all the
encodings handled here. */
#ifdef HAVE_LIBZ
switch (conn->data->set.http_ce_skip ?
IDENTITY : k->content_encoding) {
case IDENTITY:
#endif
/* This is the default when the server sends no
Content-Encoding header. See Curl_readwrite_init; the
memset() call initializes k->content_encoding to zero. */
if(!k->ignorebody)
result = Curl_client_write(conn, CLIENTWRITE_BODY, k->str,
nread);
#ifdef HAVE_LIBZ
break;
case DEFLATE:
/* Assume CLIENTWRITE_BODY; headers are not encoded. */
if(!k->ignorebody)
result = Curl_unencode_deflate_write(conn, k, nread);
break;
case GZIP:
/* Assume CLIENTWRITE_BODY; headers are not encoded. */
if(!k->ignorebody)
result = Curl_unencode_gzip_write(conn, k, nread);
break;
case COMPRESS:
default:
failf (data, "Unrecognized content encoding type. "
"libcurl understands `identity', `deflate' and `gzip' "
"content encodings.");
result = CURLE_BAD_CONTENT_ENCODING;
break;
}
#endif
}
k->badheader = HEADER_NORMAL; /* taken care of now */
if(result)
return result;
}
} /* if (! header and data to read ) */
if (is_empty_data) {
/* if we received nothing, the server closed the connection and we
are done */
k->keepon &= ~KEEP_READ;
}
} while(data_pending(conn));
} /* if( read from socket ) */
/* If we still have writing to do, we check if we have a writable
socket. */
if((k->keepon & KEEP_WRITE) && (select_res & CSELECT_OUT)) {
/* write */
int i, si;
ssize_t bytes_written;
bool writedone=TRUE;
if ((k->bytecount == 0) && (k->writebytecount == 0))
Curl_pgrsTime(data, TIMER_STARTTRANSFER);
didwhat |= KEEP_WRITE;
/*
* We loop here to do the READ and SEND loop until we run out of
* data to send or until we get EWOULDBLOCK back
*/
do {
/* only read more data if there's no upload data already
present in the upload buffer */
if(0 == data->reqdata.upload_present) {
/* init the "upload from here" pointer */
data->reqdata.upload_fromhere = k->uploadbuf;
if(!k->upload_done) {
/* HTTP pollution, this should be written nicer to become more
protocol agnostic. */
int fillcount;
if(k->wait100_after_headers &&
(data->reqdata.proto.http->sending == HTTPSEND_BODY)) {
/* If this call is to send body data, we must take some action:
We have sent off the full HTTP 1.1 request, and we shall now
go into the Expect: 100 state and await such a header */
k->wait100_after_headers = FALSE; /* headers sent */
k->write_after_100_header = TRUE; /* wait for the header */
k->keepon &= ~KEEP_WRITE; /* disable writing */
k->start100 = Curl_tvnow(); /* timeout count starts now */
didwhat &= ~KEEP_WRITE; /* we didn't write anything actually */
break;
}
result = Curl_fillreadbuffer(conn, BUFSIZE, &fillcount);
if(result)
return result;
nread = (ssize_t)fillcount;
}
else
nread = 0; /* we're done uploading/reading */
/* the signed int typecase of nread of for systems that has
unsigned size_t */
if (nread<=0) {
/* done */
k->keepon &= ~KEEP_WRITE; /* we're done writing */
writedone = TRUE;
if(conn->bits.rewindaftersend) {
result = Curl_readrewind(conn);
if(result)
return result;
}
break;
}
/* store number of bytes available for upload */
data->reqdata.upload_present = nread;
/* convert LF to CRLF if so asked */
#ifdef CURL_DO_LINEEND_CONV
/* always convert if we're FTPing in ASCII mode */
if ((data->set.crlf) || (data->set.prefer_ascii)) {
#else
if (data->set.crlf) {
#endif /* CURL_DO_LINEEND_CONV */
if(data->state.scratch == NULL)
data->state.scratch = malloc(2*BUFSIZE);
if(data->state.scratch == NULL) {
failf (data, "Failed to alloc scratch buffer!");
return CURLE_OUT_OF_MEMORY;
}
/*
* ASCII/EBCDIC Note: This is presumably a text (not binary)
* transfer so the data should already be in ASCII.
* That means the hex values for ASCII CR (0x0d) & LF (0x0a)
* must be used instead of the escape sequences \r & \n.
*/
for(i = 0, si = 0; i < nread; i++, si++) {
if (data->reqdata.upload_fromhere[i] == 0x0a) {
data->state.scratch[si++] = 0x0d;
data->state.scratch[si] = 0x0a;
if (!data->set.crlf) {
/* we're here only because FTP is in ASCII mode...
bump infilesize for the LF we just added */
data->set.infilesize++;
}
}
else
data->state.scratch[si] = data->reqdata.upload_fromhere[i];
}
if(si != nread) {
/* only perform the special operation if we really did replace
anything */
nread = si;
/* upload from the new (replaced) buffer instead */
data->reqdata.upload_fromhere = data->state.scratch;
/* set the new amount too */
data->reqdata.upload_present = nread;
}
}
}
else {
/* We have a partial buffer left from a previous "round". Use
that instead of reading more data */
}
/* write to socket (send away data) */
result = Curl_write(conn,
conn->writesockfd, /* socket to send to */
data->reqdata.upload_fromhere, /* buffer pointer */
data->reqdata.upload_present, /* buffer size */
&bytes_written); /* actually send away */
if(result)
return result;
if(data->set.verbose)
/* show the data before we change the pointer upload_fromhere */
Curl_debug(data, CURLINFO_DATA_OUT, data->reqdata.upload_fromhere,
(size_t)bytes_written, conn);
if(data->reqdata.upload_present != bytes_written) {
/* we only wrote a part of the buffer (if anything), deal with it! */
/* store the amount of bytes left in the buffer to write */
data->reqdata.upload_present -= bytes_written;
/* advance the pointer where to find the buffer when the next send
is to happen */
data->reqdata.upload_fromhere += bytes_written;
writedone = TRUE; /* we are done, stop the loop */
}
else {
/* we've uploaded that buffer now */
data->reqdata.upload_fromhere = k->uploadbuf;
data->reqdata.upload_present = 0; /* no more bytes left */
if(k->upload_done) {
/* switch off writing, we're done! */
k->keepon &= ~KEEP_WRITE; /* we're done writing */
writedone = TRUE;
}
}
k->writebytecount += bytes_written;
Curl_pgrsSetUploadCounter(data, k->writebytecount);
} while(!writedone); /* loop until we're done writing! */
}
} while(0); /* just to break out from! */
k->now = Curl_tvnow();
if(didwhat) {
/* Update read/write counters */
if(k->bytecountp)
*k->bytecountp = k->bytecount; /* read count */
if(k->writebytecountp)
*k->writebytecountp = k->writebytecount; /* write count */
}
else {
/* no read no write, this is a timeout? */
if (k->write_after_100_header) {
/* This should allow some time for the header to arrive, but only a
very short time as otherwise it'll be too much wasted times too
often. */
/* Quoting RFC2616, section "8.2.3 Use of the 100 (Continue) Status":
Therefore, when a client sends this header field to an origin server
(possibly via a proxy) from which it has never seen a 100 (Continue)
status, the client SHOULD NOT wait for an indefinite period before
sending the request body.
*/
long ms = Curl_tvdiff(k->now, k->start100);
if(ms > CURL_TIMEOUT_EXPECT_100) {
/* we've waited long enough, continue anyway */
k->write_after_100_header = FALSE;
k->keepon |= KEEP_WRITE;
}
}
}
if(Curl_pgrsUpdate(conn))
result = CURLE_ABORTED_BY_CALLBACK;
else
result = Curl_speedcheck(data, k->now);
if (result)
return result;
if (data->set.timeout &&
(Curl_tvdiff(k->now, k->start) >= data->set.timeout)) {
if (k->size != -1) {
failf(data, "Operation timed out after %ld milliseconds with %"
FORMAT_OFF_T " out of %" FORMAT_OFF_T " bytes received",
data->set.timeout, k->bytecount, k->size);
} else {
failf(data, "Operation timed out after %ld milliseconds with %"
FORMAT_OFF_T " bytes received",
data->set.timeout, k->bytecount);
}
return CURLE_OPERATION_TIMEOUTED;
}
if(!k->keepon) {
/*
* The transfer has been performed. Just make some general checks before
* returning.
*/
if(!(conn->bits.no_body) && (k->size != -1) &&
(k->bytecount != k->size) &&
#ifdef CURL_DO_LINEEND_CONV
/* Most FTP servers don't adjust their file SIZE response for CRLFs,
so we'll check to see if the discrepancy can be explained
by the number of CRLFs we've changed to LFs.
*/
(k->bytecount != (k->size + data->state.crlf_conversions)) &&
#endif /* CURL_DO_LINEEND_CONV */
!data->reqdata.newurl) {
failf(data, "transfer closed with %" FORMAT_OFF_T
" bytes remaining to read",
k->size - k->bytecount);
return CURLE_PARTIAL_FILE;
}
else if(!(conn->bits.no_body) &&
conn->bits.chunk &&
(data->reqdata.proto.http->chunk.state != CHUNK_STOP)) {
/*
* In chunked mode, return an error if the connection is closed prior to
* the empty (terminiating) chunk is read.
*
* The condition above used to check for
* conn->proto.http->chunk.datasize != 0 which is true after reading
* *any* chunk, not just the empty chunk.
*
*/
failf(data, "transfer closed with outstanding read data remaining");
return CURLE_PARTIAL_FILE;
}
if(Curl_pgrsUpdate(conn))
return CURLE_ABORTED_BY_CALLBACK;
}
/* Now update the "done" boolean we return */
*done = (bool)(0 == (k->keepon&(KEEP_READ|KEEP_WRITE)));
return CURLE_OK;
}
/*
* Curl_readwrite_init() inits the readwrite session. This is inited each time for a
* transfer, sometimes multiple times on the same SessionHandle
*/
CURLcode Curl_readwrite_init(struct connectdata *conn)
{
struct SessionHandle *data = conn->data;
struct Curl_transfer_keeper *k = &data->reqdata.keep;
/* NB: the content encoding software depends on this initialization of
Curl_transfer_keeper.*/
memset(k, 0, sizeof(struct Curl_transfer_keeper));
k->start = Curl_tvnow(); /* start time */
k->now = k->start; /* current time is now */
k->header = TRUE; /* assume header */
k->httpversion = -1; /* unknown at this point */
k->size = data->reqdata.size;
k->maxdownload = data->reqdata.maxdownload;
k->bytecountp = data->reqdata.bytecountp;
k->writebytecountp = data->reqdata.writebytecountp;
k->bytecount = 0;
k->buf = data->state.buffer;
k->uploadbuf = data->state.uploadbuffer;
k->maxfd = (conn->sockfd>conn->writesockfd?
conn->sockfd:conn->writesockfd)+1;
k->hbufp = data->state.headerbuff;
k->ignorebody=FALSE;
Curl_pgrsTime(data, TIMER_PRETRANSFER);
Curl_speedinit(data);
Curl_pgrsSetUploadCounter(data, 0);
Curl_pgrsSetDownloadCounter(data, 0);
if (!conn->bits.getheader) {
k->header = FALSE;
if(k->size > 0)
Curl_pgrsSetDownloadSize(data, k->size);
}
/* we want header and/or body, if neither then don't do this! */
if(conn->bits.getheader || !conn->bits.no_body) {
if(conn->sockfd != CURL_SOCKET_BAD) {
k->keepon |= KEEP_READ;
}
if(conn->writesockfd != CURL_SOCKET_BAD) {
/* HTTP 1.1 magic:
Even if we require a 100-return code before uploading data, we might
need to write data before that since the REQUEST may not have been
finished sent off just yet.
Thus, we must check if the request has been sent before we set the
state info where we wait for the 100-return code
*/
if (data->state.expect100header &&
(data->reqdata.proto.http->sending == HTTPSEND_BODY)) {
/* wait with write until we either got 100-continue or a timeout */
k->write_after_100_header = TRUE;
k->start100 = k->start;
}
else {
if(data->state.expect100header)
/* when we've sent off the rest of the headers, we must await a
100-continue */
k->wait100_after_headers = TRUE;
k->keepon |= KEEP_WRITE;
}
}
}
return CURLE_OK;
}
/*
* Curl_readwrite may get called multiple times. This function is called
* immediately before the first Curl_readwrite. Note that this can't be moved
* to Curl_readwrite_init since that function can get called while another
* pipeline request is in the middle of receiving data.
*
* We init chunking and trailer bits to their default values here immediately
* before receiving any header data for the current request in the pipeline.
*/
void Curl_pre_readwrite(struct connectdata *conn)
{
DEBUGF(infof(conn->data, "Pre readwrite setting chunky header "
"values to default\n"));
conn->bits.chunk=FALSE;
conn->bits.trailerHdrPresent=FALSE;
}
/*
* Curl_single_getsock() gets called by the multi interface code when the app
* has requested to get the sockets for the current connection. This function
* will then be called once for every connection that the multi interface
* keeps track of. This function will only be called for connections that are
* in the proper state to have this information available.
*/
int Curl_single_getsock(struct connectdata *conn,
curl_socket_t *sock, /* points to numsocks number
of sockets */
int numsocks)
{
struct SessionHandle *data = conn->data;
int bitmap = GETSOCK_BLANK;
int index = 0;
if(numsocks < 2)
/* simple check but we might need two slots */
return GETSOCK_BLANK;
if(data->reqdata.keep.keepon & KEEP_READ) {
bitmap |= GETSOCK_READSOCK(index);
sock[index] = conn->sockfd;
}
if(data->reqdata.keep.keepon & KEEP_WRITE) {
if((conn->sockfd != conn->writesockfd) ||
!(data->reqdata.keep.keepon & KEEP_READ)) {
/* only if they are not the same socket or we didn't have a readable
one, we increase index */
if(data->reqdata.keep.keepon & KEEP_READ)
index++; /* increase index if we need two entries */
sock[index] = conn->writesockfd;
}
bitmap |= GETSOCK_WRITESOCK(index);
}
return bitmap;
}
/*
* Transfer()
*
* This function is what performs the actual transfer. It is capable of
* doing both ways simultaneously.
* The transfer must already have been setup by a call to Curl_setup_transfer().
*
* Note that headers are created in a preallocated buffer of a default size.
* That buffer can be enlarged on demand, but it is never shrunken again.
*
* Parts of this function was once written by the friendly Mark Butler
* <butlerm@xmission.com>.
*/
static CURLcode
Transfer(struct connectdata *conn)
{
CURLcode result;
struct SessionHandle *data = conn->data;
struct Curl_transfer_keeper *k = &data->reqdata.keep;
bool done=FALSE;
if(!(conn->protocol & PROT_FILE)) {
/* Only do this if we are not transferring FILE:, since the file: treatment
is different*/
Curl_readwrite_init(conn);
Curl_pre_readwrite(conn);
}
if((conn->sockfd == CURL_SOCKET_BAD) &&
(conn->writesockfd == CURL_SOCKET_BAD))
/* nothing to read, nothing to write, we're already OK! */
return CURLE_OK;
/* we want header and/or body, if neither then don't do this! */
if(!conn->bits.getheader && conn->bits.no_body)
return CURLE_OK;
while (!done) {
curl_socket_t fd_read;
curl_socket_t fd_write;
/* limit-rate logic: if speed exceeds threshold, then do not include fd in
select set. The current speed is recalculated in each Curl_readwrite()
call */
if ((k->keepon & KEEP_WRITE) &&
(!data->set.max_send_speed ||
(data->progress.ulspeed < data->set.max_send_speed) )) {
fd_write = conn->writesockfd;
k->keepon &= ~KEEP_WRITE_HOLD;
}
else {
fd_write = CURL_SOCKET_BAD;
if(k->keepon & KEEP_WRITE)
k->keepon |= KEEP_WRITE_HOLD; /* hold it */
}
if ((k->keepon & KEEP_READ) &&
(!data->set.max_recv_speed ||
(data->progress.dlspeed < data->set.max_recv_speed)) ) {
fd_read = conn->sockfd;
k->keepon &= ~KEEP_READ_HOLD;
}
else {
fd_read = CURL_SOCKET_BAD;
if(k->keepon & KEEP_READ)
k->keepon |= KEEP_READ_HOLD; /* hold it */
}
/* The *_HOLD logic is necessary since even though there might be no
traffic during the select interval, we still call Curl_readwrite() for
the timeout case and if we limit transfer speed we must make sure that
this function doesn't transfer anything while in HOLD status. */
switch (Curl_socket_ready(fd_read, fd_write, 1000)) {
case -1: /* select() error, stop reading */
#ifdef EINTR
/* The EINTR is not serious, and it seems you might get this more
ofen when using the lib in a multi-threaded environment! */
if(SOCKERRNO == EINTR)
;
else
#endif
done = TRUE; /* no more read or write */
continue;
case 0: /* timeout */
default: /* readable descriptors */
result = Curl_readwrite(conn, &done);
break;
}
if(result)
return result;
/* "done" signals to us if the transfer(s) are ready */
}
return CURLE_OK;
}
/*
* Curl_pretransfer() is called immediately before a transfer starts.
*/
CURLcode Curl_pretransfer(struct SessionHandle *data)
{
CURLcode res;
if(!data->change.url) {
/* we can't do anything wihout URL */
failf(data, "No URL set!\n");
return CURLE_URL_MALFORMAT;
}
/* Init the SSL session ID cache here. We do it here since we want to do it
after the *_setopt() calls (that could change the size of the cache) but
before any transfer takes place. */
res = Curl_ssl_initsessions(data, data->set.ssl.numsessions);
if(res)
return res;
data->set.followlocation=0; /* reset the location-follow counter */
data->state.this_is_a_follow = FALSE; /* reset this */
data->state.errorbuf = FALSE; /* no error has occurred */
data->state.authproblem = FALSE;
data->state.authhost.want = data->set.httpauth;
data->state.authproxy.want = data->set.proxyauth;
/* If there is a list of cookie files to read, do it now! */
if(data->change.cookielist) {
Curl_cookie_loadfiles(data);
}
/* Allow data->set.use_port to set which port to use. This needs to be
* disabled for example when we follow Location: headers to URLs using
* different ports! */
data->state.allow_port = TRUE;
#if defined(HAVE_SIGNAL) && defined(SIGPIPE) && !defined(HAVE_MSG_NOSIGNAL)
/*************************************************************
* Tell signal handler to ignore SIGPIPE
*************************************************************/
if(!data->set.no_signal)
data->state.prev_signal = signal(SIGPIPE, SIG_IGN);
#endif
Curl_initinfo(data); /* reset session-specific information "variables" */
Curl_pgrsStartNow(data);
return CURLE_OK;
}
/*
* Curl_posttransfer() is called immediately after a transfer ends
*/
CURLcode Curl_posttransfer(struct SessionHandle *data)
{
#if defined(HAVE_SIGNAL) && defined(SIGPIPE) && !defined(HAVE_MSG_NOSIGNAL)
/* restore the signal handler for SIGPIPE before we get back */
if(!data->set.no_signal)
signal(SIGPIPE, data->state.prev_signal);
#else
(void)data; /* unused parameter */
#endif
if(!(data->progress.flags & PGRS_HIDE) &&
!data->progress.callback)
/* only output if we don't use a progress callback and we're not hidden */
fprintf(data->set.err, "\n");
return CURLE_OK;
}
/*
* strlen_url() returns the length of the given URL if the spaces within the
* URL were properly URL encoded.
*/
static int strlen_url(char *url)
{
char *ptr;
int newlen=0;
bool left=TRUE; /* left side of the ? */
for(ptr=url; *ptr; ptr++) {
switch(*ptr) {
case '?':
left=FALSE;
default:
newlen++;
break;
case ' ':
if(left)
newlen+=3;
else
newlen++;
break;
}
}
return newlen;
}
/* strcpy_url() copies a url to a output buffer and URL-encodes the spaces in
* the source URL accordingly.
*/
static void strcpy_url(char *output, char *url)
{
/* we must add this with whitespace-replacing */
bool left=TRUE;
char *iptr;
char *optr = output;
for(iptr = url; /* read from here */
*iptr; /* until zero byte */
iptr++) {
switch(*iptr) {
case '?':
left=FALSE;
default:
*optr++=*iptr;
break;
case ' ':
if(left) {
*optr++='%'; /* add a '%' */
*optr++='2'; /* add a '2' */
*optr++='0'; /* add a '0' */
}
else
*optr++='+'; /* add a '+' here */
break;
}
}
*optr=0; /* zero terminate output buffer */
}
/*
* Curl_follow() handles the URL redirect magic. Pass in the 'newurl' string
* as given by the remote server and set up the new URL to request.
*/
CURLcode Curl_follow(struct SessionHandle *data,
char *newurl, /* this 'newurl' is the Location: string,
and it must be malloc()ed before passed
here */
bool retry) /* set TRUE if this is a request retry as
opposed to a real redirect following */
{
/* Location: redirect */
char prot[16]; /* URL protocol string storage */
char letter; /* used for a silly sscanf */
size_t newlen;
char *newest;
if(!retry) {
if ((data->set.maxredirs != -1) &&
(data->set.followlocation >= data->set.maxredirs)) {
failf(data,"Maximum (%d) redirects followed", data->set.maxredirs);
return CURLE_TOO_MANY_REDIRECTS;
}
/* mark the next request as a followed location: */
data->state.this_is_a_follow = TRUE;
data->set.followlocation++; /* count location-followers */
}
if(data->set.http_auto_referer) {
/* We are asked to automatically set the previous URL as the
referer when we get the next URL. We pick the ->url field,
which may or may not be 100% correct */
if(data->change.referer_alloc)
/* If we already have an allocated referer, free this first */
free(data->change.referer);
data->change.referer = strdup(data->change.url);
data->change.referer_alloc = TRUE; /* yes, free this later */
}
if(2 != sscanf(newurl, "%15[^?&/:]://%c", prot, &letter)) {
/***
*DANG* this is an RFC 2068 violation. The URL is supposed
to be absolute and this doesn't seem to be that!
***
Instead, we have to TRY to append this new path to the old URL
to the right of the host part. Oh crap, this is doomed to cause
problems in the future...
*/
char *protsep;
char *pathsep;
char *useurl = newurl;
size_t urllen;
/* we must make our own copy of the URL to play with, as it may
point to read-only data */
char *url_clone=strdup(data->change.url);
if(!url_clone)
return CURLE_OUT_OF_MEMORY; /* skip out of this NOW */
/* protsep points to the start of the host name */
protsep=strstr(url_clone, "//");
if(!protsep)
protsep=url_clone;
else
protsep+=2; /* pass the slashes */
if('/' != newurl[0]) {
int level=0;
/* First we need to find out if there's a ?-letter in the URL,
and cut it and the right-side of that off */
pathsep = strchr(protsep, '?');
if(pathsep)
*pathsep=0;
/* we have a relative path to append to the last slash if
there's one available */
pathsep = strrchr(protsep, '/');
if(pathsep)
*pathsep=0;
/* Check if there's any slash after the host name, and if so,
remember that position instead */
pathsep = strchr(protsep, '/');
if(pathsep)
protsep = pathsep+1;
else
protsep = NULL;
/* now deal with one "./" or any amount of "../" in the newurl
and act accordingly */
if((useurl[0] == '.') && (useurl[1] == '/'))
useurl+=2; /* just skip the "./" */
while((useurl[0] == '.') &&
(useurl[1] == '.') &&
(useurl[2] == '/')) {
level++;
useurl+=3; /* pass the "../" */
}
if(protsep) {
while(level--) {
/* cut off one more level from the right of the original URL */
pathsep = strrchr(protsep, '/');
if(pathsep)
*pathsep=0;
else {
*protsep=0;
break;
}
}
}
}
else {
/* We got a new absolute path for this server, cut off from the
first slash */
pathsep = strchr(protsep, '/');
if(pathsep) {
/* When people use badly formatted URLs, such as
"http://www.url.com?dir=/home/daniel" we must not use the first
slash, if there's a ?-letter before it! */
char *sep = strchr(protsep, '?');
if(sep && (sep < pathsep))
pathsep = sep;
*pathsep=0;
}
else {
/* There was no slash. Now, since we might be operating on a badly
formatted URL, such as "http://www.url.com?id=2380" which doesn't
use a slash separator as it is supposed to, we need to check for a
?-letter as well! */
pathsep = strchr(protsep, '?');
if(pathsep)
*pathsep=0;
}
}
/* If the new part contains a space, this is a mighty stupid redirect
but we still make an effort to do "right". To the left of a '?'
letter we replace each space with %20 while it is replaced with '+'
on the right side of the '?' letter.
*/
newlen = strlen_url(useurl);
urllen = strlen(url_clone);
newest=(char *)malloc( urllen + 1 + /* possible slash */
newlen + 1 /* zero byte */);
if(!newest) {
free(url_clone); /* don't leak this */
return CURLE_OUT_OF_MEMORY; /* go out from this */
}
/* copy over the root url part */
memcpy(newest, url_clone, urllen);
/* check if we need to append a slash */
if(('/' == useurl[0]) || (protsep && !*protsep))
;
else
newest[urllen++]='/';
/* then append the new piece on the right side */
strcpy_url(&newest[urllen], useurl);
free(newurl); /* newurl is the allocated pointer */
free(url_clone);
newurl = newest;
}
else {
/* This is an absolute URL, don't allow the custom port number */
data->state.allow_port = FALSE;
if(strchr(newurl, ' ')) {
/* This new URL contains at least one space, this is a mighty stupid
redirect but we still make an effort to do "right". */
newlen = strlen_url(newurl);
newest = malloc(newlen+1); /* get memory for this */
if(newest) {
strcpy_url(newest, newurl); /* create a space-free URL */
free(newurl); /* that was no good */
newurl = newest; /* use this instead now */
}
}
}
if(data->change.url_alloc)
free(data->change.url);
else
data->change.url_alloc = TRUE; /* the URL is allocated */
data->change.url = newurl;
newurl = NULL; /* don't free! */
infof(data, "Issue another request to this URL: '%s'\n", data->change.url);
/*
* We get here when the HTTP code is 300-399 (and 401). We need to perform
* differently based on exactly what return code there was.
*
* News from 7.10.6: we can also get here on a 401 or 407, in case we act on
* a HTTP (proxy-) authentication scheme other than Basic.
*/
switch(data->info.httpcode) {
/* 401 - Act on a www-authentication, we keep on moving and do the
Authorization: XXXX header in the HTTP request code snippet */
/* 407 - Act on a proxy-authentication, we keep on moving and do the
Proxy-Authorization: XXXX header in the HTTP request code snippet */
/* 300 - Multiple Choices */
/* 306 - Not used */
/* 307 - Temporary Redirect */
default: /* for all above (and the unknown ones) */
/* Some codes are explicitly mentioned since I've checked RFC2616 and they
* seem to be OK to POST to.
*/
break;
case 301: /* Moved Permanently */
/* (quote from RFC2616, section 10.3.2):
*
* Note: When automatically redirecting a POST request after receiving a
* 301 status code, some existing HTTP/1.0 user agents will erroneously
* change it into a GET request.
*
* ----
*
* Warning: Because most of importants user agents do this obvious RFC2616
* violation, many webservers expect this misbehavior. So these servers
* often answers to a POST request with an error page. To be sure that
* libcurl gets the page that most user agents would get, libcurl has to
* force GET:
*/
if( data->set.httpreq == HTTPREQ_POST
|| data->set.httpreq == HTTPREQ_POST_FORM) {
infof(data,
"Violate RFC 2616/10.3.2 and switch from POST to GET\n");
data->set.httpreq = HTTPREQ_GET;
}
break;
case 302: /* Found */
/* (From 10.3.3)
Note: RFC 1945 and RFC 2068 specify that the client is not allowed
to change the method on the redirected request. However, most
existing user agent implementations treat 302 as if it were a 303
response, performing a GET on the Location field-value regardless
of the original request method. The status codes 303 and 307 have
been added for servers that wish to make unambiguously clear which
kind of reaction is expected of the client.
(From 10.3.4)
Note: Many pre-HTTP/1.1 user agents do not understand the 303
status. When interoperability with such clients is a concern, the
302 status code may be used instead, since most user agents react
to a 302 response as described here for 303.
*/
case 303: /* See Other */
/* Disable both types of POSTs, since doing a second POST when
* following isn't what anyone would want! */
if(data->set.httpreq != HTTPREQ_GET) {
data->set.httpreq = HTTPREQ_GET; /* enforce GET request */
infof(data, "Disables POST, goes with %s\n",
data->set.opt_no_body?"HEAD":"GET");
}
break;
case 304: /* Not Modified */
/* 304 means we did a conditional request and it was "Not modified".
* We shouldn't get any Location: header in this response!
*/
break;
case 305: /* Use Proxy */
/* (quote from RFC2616, section 10.3.6):
* "The requested resource MUST be accessed through the proxy given
* by the Location field. The Location field gives the URI of the
* proxy. The recipient is expected to repeat this single request
* via the proxy. 305 responses MUST only be generated by origin
* servers."
*/
break;
}
Curl_pgrsTime(data, TIMER_REDIRECT);
Curl_pgrsResetTimes(data);
return CURLE_OK;
}
static CURLcode
Curl_connect_host(struct SessionHandle *data,
struct connectdata **conn)
{
CURLcode res = CURLE_OK;
int urlchanged = FALSE;
do {
bool async;
bool protocol_done=TRUE; /* will be TRUE always since this is only used
within the easy interface */
Curl_pgrsTime(data, TIMER_STARTSINGLE);
data->change.url_changed = FALSE;
res = Curl_connect(data, conn, &async, &protocol_done);
if((CURLE_OK == res) && async) {
/* Now, if async is TRUE here, we need to wait for the name
to resolve */
res = Curl_wait_for_resolv(*conn, NULL);
if(CURLE_OK == res)
/* Resolved, continue with the connection */
res = Curl_async_resolved(*conn, &protocol_done);
else
/* if we can't resolve, we kill this "connection" now */
(void)Curl_disconnect(*conn);
}
if(res)
break;
/* If a callback (or something) has altered the URL we should use within
the Curl_connect(), we detect it here and act as if we are redirected
to the new URL */
urlchanged = data->change.url_changed;
if ((CURLE_OK == res) && urlchanged) {
res = Curl_done(conn, res, FALSE);
if(CURLE_OK == res) {
char *gotourl = strdup(data->change.url);
res = Curl_follow(data, gotourl, FALSE);
if(res)
free(gotourl);
}
}
} while (urlchanged && res == CURLE_OK);
return res;
}
/* Returns TRUE and sets '*url' if a request retry is wanted */
bool Curl_retry_request(struct connectdata *conn,
char **url)
{
bool retry = FALSE;
struct SessionHandle *data = conn->data;
if((data->reqdata.keep.bytecount+conn->headerbytecount == 0) &&
conn->bits.reuse &&
!conn->bits.no_body) {
/* We got no data, we attempted to re-use a connection and yet we want a
"body". This might happen if the connection was left alive when we were
done using it before, but that was closed when we wanted to read from
it again. Bad luck. Retry the same request on a fresh connect! */
infof(conn->data, "Connection died, retrying a fresh connect\n");
*url = strdup(conn->data->change.url);
conn->bits.close = TRUE; /* close this connection */
conn->bits.retry = TRUE; /* mark this as a connection we're about
to retry. Marking it this way should
prevent i.e HTTP transfers to return
error just because nothing has been
transfered! */
retry = TRUE;
}
return retry;
}
/*
* Curl_perform() is the internal high-level function that gets called by the
* external curl_easy_perform() function. It inits, performs and cleans up a
* single file transfer.
*/
CURLcode Curl_perform(struct SessionHandle *data)
{
CURLcode res;
CURLcode res2;
struct connectdata *conn=NULL;
char *newurl = NULL; /* possibly a new URL to follow to! */
bool retry = FALSE;
data->state.used_interface = Curl_if_easy;
res = Curl_pretransfer(data);
if(res)
return res;
/*
* It is important that there is NO 'return' from this function at any other
* place than falling down to the end of the function! This is because we
* have cleanup stuff that must be done before we get back, and that is only
* performed after this do-while loop.
*/
do {
res = Curl_connect_host(data, &conn); /* primary connection */
if(res == CURLE_OK) {
bool do_done;
if(data->set.connect_only) {
/* keep connection open for application to use the socket */
conn->bits.close = FALSE;
res = Curl_done(&conn, CURLE_OK, FALSE);
break;
}
res = Curl_do(&conn, &do_done);
if(res == CURLE_OK) {
res = Transfer(conn); /* now fetch that URL please */
if(res == CURLE_OK) {
retry = Curl_retry_request(conn, &newurl);
if(!retry)
/*
* We must duplicate the new URL here as the connection data may
* be free()ed in the Curl_done() function.
*/
newurl = data->reqdata.newurl?strdup(data->reqdata.newurl):NULL;
}
else {
/* The transfer phase returned error, we mark the connection to get
* closed to prevent being re-used. This is becasue we can't
* possibly know if the connection is in a good shape or not now. */
conn->bits.close = TRUE;
if(CURL_SOCKET_BAD != conn->sock[SECONDARYSOCKET]) {
/* if we failed anywhere, we must clean up the secondary socket if
it was used */
sclose(conn->sock[SECONDARYSOCKET]);
conn->sock[SECONDARYSOCKET] = CURL_SOCKET_BAD;
}
}
/* Always run Curl_done(), even if some of the previous calls
failed, but return the previous (original) error code */
res2 = Curl_done(&conn, res, FALSE);
if(CURLE_OK == res)
res = res2;
}
else
/* Curl_do() failed, clean up left-overs in the done-call */
res2 = Curl_done(&conn, res, FALSE);
/*
* Important: 'conn' cannot be used here, since it may have been closed
* in 'Curl_done' or other functions.
*/
if((res == CURLE_OK) && newurl) {
res = Curl_follow(data, newurl, retry);
if(CURLE_OK == res) {
newurl = NULL;
continue;
}
}
}
break; /* it only reaches here when this shouldn't loop */
} while(1); /* loop if Location: */
if(newurl)
free(newurl);
if(res && !data->state.errorbuf) {
/*
* As an extra precaution: if no error string has been set and there was
* an error, use the strerror() string or if things are so bad that not
* even that is good, set a bad string that mentions the error code.
*/
const char *str = curl_easy_strerror(res);
if(!str)
failf(data, "unspecified error %d", (int)res);
else
failf(data, "%s", str);
}
/* run post-transfer uncondionally, but don't clobber the return code if
we already have an error code recorder */
res2 = Curl_posttransfer(data);
if(!res && res2)
res = res2;
return res;
}
/*
* Curl_setup_transfer() is called to setup some basic properties for the
* upcoming transfer.
*/
CURLcode
Curl_setup_transfer(
struct connectdata *c_conn, /* connection data */
int sockindex, /* socket index to read from or -1 */
curl_off_t size, /* -1 if unknown at this point */
bool getheader, /* TRUE if header parsing is wanted */
curl_off_t *bytecountp, /* return number of bytes read or NULL */
int writesockindex, /* socket index to write to, it may very
well be the same we read from. -1
disables */
curl_off_t *writecountp /* return number of bytes written or
NULL */
)
{
struct connectdata *conn = (struct connectdata *)c_conn;
struct SessionHandle *data = conn->data;
if(!conn)
return CURLE_BAD_FUNCTION_ARGUMENT;
DEBUGASSERT((sockindex <= 1) && (sockindex >= -1));
/* now copy all input parameters */
conn->sockfd = sockindex == -1 ?
CURL_SOCKET_BAD : conn->sock[sockindex];
conn->writesockfd = writesockindex == -1 ?
CURL_SOCKET_BAD:conn->sock[writesockindex];
conn->bits.getheader = getheader;
data->reqdata.size = size;
data->reqdata.bytecountp = bytecountp;
data->reqdata.writebytecountp = writecountp;
return CURLE_OK;
}
|