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
|
/* Python interpreter top-level routines, including init/exit */
#include "Python.h"
#include "Python-ast.h"
#undef Yield /* undefine macro conflicting with winbase.h */
#include "grammar.h"
#include "node.h"
#include "token.h"
#include "parsetok.h"
#include "errcode.h"
#include "code.h"
#include "compile.h"
#include "symtable.h"
#include "pyarena.h"
#include "ast.h"
#include "eval.h"
#include "marshal.h"
#include "abstract.h"
#ifdef HAVE_SIGNAL_H
#include <signal.h>
#endif
#ifdef MS_WINDOWS
#include "malloc.h" /* for alloca */
#endif
#ifdef HAVE_LANGINFO_H
#include <locale.h>
#include <langinfo.h>
#endif
#ifdef MS_WINDOWS
#undef BYTE
#include "windows.h"
#endif
#ifndef Py_REF_DEBUG
#define PRINT_TOTAL_REFS()
#else /* Py_REF_DEBUG */
#define PRINT_TOTAL_REFS() fprintf(stderr, \
"[%" PY_FORMAT_SIZE_T "d refs]\n", \
_Py_GetRefTotal())
#endif
#ifdef __cplusplus
extern "C" {
#endif
extern char *Py_GetPath(void);
extern grammar _PyParser_Grammar; /* From graminit.c */
/* Forward */
static void initmain(void);
static void initsite(void);
static PyObject *run_mod(mod_ty, const char *, PyObject *, PyObject *,
PyCompilerFlags *, PyArena *);
static PyObject *run_pyc_file(FILE *, const char *, PyObject *, PyObject *,
PyCompilerFlags *);
static void err_input(perrdetail *);
static void initsigs(void);
static void wait_for_thread_shutdown(void);
static void call_sys_exitfunc(void);
static void call_ll_exitfuncs(void);
extern void _PyUnicode_Init(void);
extern void _PyUnicode_Fini(void);
#ifdef WITH_THREAD
extern void _PyGILState_Init(PyInterpreterState *, PyThreadState *);
extern void _PyGILState_Fini(void);
#endif /* WITH_THREAD */
int Py_DebugFlag; /* Needed by parser.c */
int Py_VerboseFlag; /* Needed by import.c */
int Py_InteractiveFlag; /* Needed by Py_FdIsInteractive() below */
int Py_InspectFlag; /* Needed to determine whether to exit at SystemExit */
int Py_NoSiteFlag; /* Suppress 'import site' */
int Py_BytesWarningFlag; /* Warn on str(bytes) and str(buffer) */
int Py_DontWriteBytecodeFlag; /* Suppress writing bytecode files (*.py[co]) */
int Py_UseClassExceptionsFlag = 1; /* Needed by bltinmodule.c: deprecated */
int Py_FrozenFlag; /* Needed by getpath.c */
int Py_UnicodeFlag = 0; /* Needed by compile.c */
int Py_IgnoreEnvironmentFlag; /* e.g. PYTHONPATH, PYTHONHOME */
/* _XXX Py_QnewFlag should go away in 2.3. It's true iff -Qnew is passed,
on the command line, and is used in 2.2 by ceval.c to make all "/" divisions
true divisions (which they will be in 2.3). */
int _Py_QnewFlag = 0;
int Py_NoUserSiteDirectory = 0; /* for -s and site.py */
int Py_HashRandomizationFlag = 0; /* for -R and PYTHONHASHSEED */
/* PyModule_GetWarningsModule is no longer necessary as of 2.6
since _warnings is builtin. This API should not be used. */
PyObject *
PyModule_GetWarningsModule(void)
{
return PyImport_ImportModule("warnings");
}
static int initialized = 0;
/* API to access the initialized flag -- useful for esoteric use */
int
Py_IsInitialized(void)
{
return initialized;
}
/* Global initializations. Can be undone by Py_Finalize(). Don't
call this twice without an intervening Py_Finalize() call. When
initializations fail, a fatal error is issued and the function does
not return. On return, the first thread and interpreter state have
been created.
Locking: you must hold the interpreter lock while calling this.
(If the lock has not yet been initialized, that's equivalent to
having the lock, but you cannot use multiple threads.)
*/
static int
add_flag(int flag, const char *envs)
{
int env = atoi(envs);
if (flag < env)
flag = env;
if (flag < 1)
flag = 1;
return flag;
}
void
Py_InitializeEx(int install_sigs)
{
PyInterpreterState *interp;
PyThreadState *tstate;
PyObject *bimod, *sysmod;
char *p;
char *icodeset = NULL; /* On Windows, input codeset may theoretically
differ from output codeset. */
char *codeset = NULL;
char *errors = NULL;
int free_codeset = 0;
int overridden = 0;
PyObject *sys_stream, *sys_isatty;
#if defined(Py_USING_UNICODE) && defined(HAVE_LANGINFO_H) && defined(CODESET)
char *saved_locale, *loc_codeset;
#endif
#ifdef MS_WINDOWS
char ibuf[128];
char buf[128];
#endif
extern void _Py_ReadyTypes(void);
if (initialized)
return;
initialized = 1;
if ((p = Py_GETENV("PYTHONDEBUG")) && *p != '\0')
Py_DebugFlag = add_flag(Py_DebugFlag, p);
if ((p = Py_GETENV("PYTHONVERBOSE")) && *p != '\0')
Py_VerboseFlag = add_flag(Py_VerboseFlag, p);
if ((p = Py_GETENV("PYTHONOPTIMIZE")) && *p != '\0')
Py_OptimizeFlag = add_flag(Py_OptimizeFlag, p);
if ((p = Py_GETENV("PYTHONDONTWRITEBYTECODE")) && *p != '\0')
Py_DontWriteBytecodeFlag = add_flag(Py_DontWriteBytecodeFlag, p);
/* The variable is only tested for existence here; _PyRandom_Init will
check its value further. */
if ((p = Py_GETENV("PYTHONHASHSEED")) && *p != '\0')
Py_HashRandomizationFlag = add_flag(Py_HashRandomizationFlag, p);
_PyRandom_Init();
interp = PyInterpreterState_New();
if (interp == NULL)
Py_FatalError("Py_Initialize: can't make first interpreter");
tstate = PyThreadState_New(interp);
if (tstate == NULL)
Py_FatalError("Py_Initialize: can't make first thread");
(void) PyThreadState_Swap(tstate);
_Py_ReadyTypes();
if (!_PyFrame_Init())
Py_FatalError("Py_Initialize: can't init frames");
if (!_PyInt_Init())
Py_FatalError("Py_Initialize: can't init ints");
if (!_PyLong_Init())
Py_FatalError("Py_Initialize: can't init longs");
if (!PyByteArray_Init())
Py_FatalError("Py_Initialize: can't init bytearray");
_PyFloat_Init();
interp->modules = PyDict_New();
if (interp->modules == NULL)
Py_FatalError("Py_Initialize: can't make modules dictionary");
interp->modules_reloading = PyDict_New();
if (interp->modules_reloading == NULL)
Py_FatalError("Py_Initialize: can't make modules_reloading dictionary");
#ifdef Py_USING_UNICODE
/* Init Unicode implementation; relies on the codec registry */
_PyUnicode_Init();
#endif
bimod = _PyBuiltin_Init();
if (bimod == NULL)
Py_FatalError("Py_Initialize: can't initialize __builtin__");
interp->builtins = PyModule_GetDict(bimod);
if (interp->builtins == NULL)
Py_FatalError("Py_Initialize: can't initialize builtins dict");
Py_INCREF(interp->builtins);
sysmod = _PySys_Init();
if (sysmod == NULL)
Py_FatalError("Py_Initialize: can't initialize sys");
interp->sysdict = PyModule_GetDict(sysmod);
if (interp->sysdict == NULL)
Py_FatalError("Py_Initialize: can't initialize sys dict");
Py_INCREF(interp->sysdict);
_PyImport_FixupExtension("sys", "sys");
PySys_SetPath(Py_GetPath());
PyDict_SetItemString(interp->sysdict, "modules",
interp->modules);
_PyImport_Init();
/* initialize builtin exceptions */
_PyExc_Init();
_PyImport_FixupExtension("exceptions", "exceptions");
/* phase 2 of builtins */
_PyImport_FixupExtension("__builtin__", "__builtin__");
_PyImportHooks_Init();
if (install_sigs)
initsigs(); /* Signal handling stuff, including initintr() */
/* Initialize warnings. */
_PyWarnings_Init();
if (PySys_HasWarnOptions()) {
PyObject *warnings_module = PyImport_ImportModule("warnings");
if (!warnings_module)
PyErr_Clear();
Py_XDECREF(warnings_module);
}
initmain(); /* Module __main__ */
/* auto-thread-state API, if available */
#ifdef WITH_THREAD
_PyGILState_Init(interp, tstate);
#endif /* WITH_THREAD */
if (!Py_NoSiteFlag)
initsite(); /* Module site */
if ((p = Py_GETENV("PYTHONIOENCODING")) && *p != '\0') {
p = icodeset = codeset = strdup(p);
free_codeset = 1;
errors = strchr(p, ':');
if (errors) {
*errors = '\0';
errors++;
}
overridden = 1;
}
#if defined(Py_USING_UNICODE) && defined(HAVE_LANGINFO_H) && defined(CODESET)
/* On Unix, set the file system encoding according to the
user's preference, if the CODESET names a well-known
Python codec, and Py_FileSystemDefaultEncoding isn't
initialized by other means. Also set the encoding of
stdin and stdout if these are terminals, unless overridden. */
if (!overridden || !Py_FileSystemDefaultEncoding) {
saved_locale = strdup(setlocale(LC_CTYPE, NULL));
setlocale(LC_CTYPE, "");
loc_codeset = nl_langinfo(CODESET);
if (loc_codeset && *loc_codeset) {
PyObject *enc = PyCodec_Encoder(loc_codeset);
if (enc) {
loc_codeset = strdup(loc_codeset);
Py_DECREF(enc);
} else {
if (PyErr_ExceptionMatches(PyExc_LookupError)) {
PyErr_Clear();
loc_codeset = NULL;
} else {
PyErr_Print();
exit(1);
}
}
} else
loc_codeset = NULL;
setlocale(LC_CTYPE, saved_locale);
free(saved_locale);
if (!overridden) {
codeset = icodeset = loc_codeset;
free_codeset = 1;
}
/* Initialize Py_FileSystemDefaultEncoding from
locale even if PYTHONIOENCODING is set. */
if (!Py_FileSystemDefaultEncoding) {
Py_FileSystemDefaultEncoding = loc_codeset;
if (!overridden)
free_codeset = 0;
}
}
#endif
#ifdef MS_WINDOWS
if (!overridden) {
icodeset = ibuf;
codeset = buf;
sprintf(ibuf, "cp%d", GetConsoleCP());
sprintf(buf, "cp%d", GetConsoleOutputCP());
}
#endif
if (codeset) {
sys_stream = PySys_GetObject("stdin");
sys_isatty = PyObject_CallMethod(sys_stream, "isatty", "");
if (!sys_isatty)
PyErr_Clear();
if ((overridden ||
(sys_isatty && PyObject_IsTrue(sys_isatty))) &&
PyFile_Check(sys_stream)) {
if (!PyFile_SetEncodingAndErrors(sys_stream, icodeset, errors))
Py_FatalError("Cannot set codeset of stdin");
}
Py_XDECREF(sys_isatty);
sys_stream = PySys_GetObject("stdout");
sys_isatty = PyObject_CallMethod(sys_stream, "isatty", "");
if (!sys_isatty)
PyErr_Clear();
if ((overridden ||
(sys_isatty && PyObject_IsTrue(sys_isatty))) &&
PyFile_Check(sys_stream)) {
if (!PyFile_SetEncodingAndErrors(sys_stream, codeset, errors))
Py_FatalError("Cannot set codeset of stdout");
}
Py_XDECREF(sys_isatty);
sys_stream = PySys_GetObject("stderr");
sys_isatty = PyObject_CallMethod(sys_stream, "isatty", "");
if (!sys_isatty)
PyErr_Clear();
if((overridden ||
(sys_isatty && PyObject_IsTrue(sys_isatty))) &&
PyFile_Check(sys_stream)) {
if (!PyFile_SetEncodingAndErrors(sys_stream, codeset, errors))
Py_FatalError("Cannot set codeset of stderr");
}
Py_XDECREF(sys_isatty);
if (free_codeset)
free(codeset);
}
}
void
Py_Initialize(void)
{
Py_InitializeEx(1);
}
#ifdef COUNT_ALLOCS
extern void dump_counts(FILE*);
#endif
/* Undo the effect of Py_Initialize().
Beware: if multiple interpreter and/or thread states exist, these
are not wiped out; only the current thread and interpreter state
are deleted. But since everything else is deleted, those other
interpreter and thread states should no longer be used.
(XXX We should do better, e.g. wipe out all interpreters and
threads.)
Locking: as above.
*/
void
Py_Finalize(void)
{
PyInterpreterState *interp;
PyThreadState *tstate;
if (!initialized)
return;
wait_for_thread_shutdown();
/* The interpreter is still entirely intact at this point, and the
* exit funcs may be relying on that. In particular, if some thread
* or exit func is still waiting to do an import, the import machinery
* expects Py_IsInitialized() to return true. So don't say the
* interpreter is uninitialized until after the exit funcs have run.
* Note that Threading.py uses an exit func to do a join on all the
* threads created thru it, so this also protects pending imports in
* the threads created via Threading.
*/
call_sys_exitfunc();
initialized = 0;
/* Get current thread state and interpreter pointer */
tstate = PyThreadState_GET();
interp = tstate->interp;
/* Disable signal handling */
PyOS_FiniInterrupts();
/* Clear type lookup cache */
PyType_ClearCache();
/* Collect garbage. This may call finalizers; it's nice to call these
* before all modules are destroyed.
* XXX If a __del__ or weakref callback is triggered here, and tries to
* XXX import a module, bad things can happen, because Python no
* XXX longer believes it's initialized.
* XXX Fatal Python error: Interpreter not initialized (version mismatch?)
* XXX is easy to provoke that way. I've also seen, e.g.,
* XXX Exception exceptions.ImportError: 'No module named sha'
* XXX in <function callback at 0x008F5718> ignored
* XXX but I'm unclear on exactly how that one happens. In any case,
* XXX I haven't seen a real-life report of either of these.
*/
PyGC_Collect();
#ifdef COUNT_ALLOCS
/* With COUNT_ALLOCS, it helps to run GC multiple times:
each collection might release some types from the type
list, so they become garbage. */
while (PyGC_Collect() > 0)
/* nothing */;
#endif
/* Destroy all modules */
PyImport_Cleanup();
/* Collect final garbage. This disposes of cycles created by
* new-style class definitions, for example.
* XXX This is disabled because it caused too many problems. If
* XXX a __del__ or weakref callback triggers here, Python code has
* XXX a hard time running, because even the sys module has been
* XXX cleared out (sys.stdout is gone, sys.excepthook is gone, etc).
* XXX One symptom is a sequence of information-free messages
* XXX coming from threads (if a __del__ or callback is invoked,
* XXX other threads can execute too, and any exception they encounter
* XXX triggers a comedy of errors as subsystem after subsystem
* XXX fails to find what it *expects* to find in sys to help report
* XXX the exception and consequent unexpected failures). I've also
* XXX seen segfaults then, after adding print statements to the
* XXX Python code getting called.
*/
#if 0
PyGC_Collect();
#endif
/* Destroy the database used by _PyImport_{Fixup,Find}Extension */
_PyImport_Fini();
/* Debugging stuff */
#ifdef COUNT_ALLOCS
dump_counts(stdout);
#endif
PRINT_TOTAL_REFS();
#ifdef Py_TRACE_REFS
/* Display all objects still alive -- this can invoke arbitrary
* __repr__ overrides, so requires a mostly-intact interpreter.
* Alas, a lot of stuff may still be alive now that will be cleaned
* up later.
*/
if (Py_GETENV("PYTHONDUMPREFS"))
_Py_PrintReferences(stderr);
#endif /* Py_TRACE_REFS */
/* Clear interpreter state */
PyInterpreterState_Clear(interp);
/* Now we decref the exception classes. After this point nothing
can raise an exception. That's okay, because each Fini() method
below has been checked to make sure no exceptions are ever
raised.
*/
_PyExc_Fini();
/* Cleanup auto-thread-state */
#ifdef WITH_THREAD
_PyGILState_Fini();
#endif /* WITH_THREAD */
/* Delete current thread */
PyThreadState_Swap(NULL);
PyInterpreterState_Delete(interp);
/* Sundry finalizers */
PyMethod_Fini();
PyFrame_Fini();
PyCFunction_Fini();
PyTuple_Fini();
PyList_Fini();
PySet_Fini();
PyString_Fini();
PyByteArray_Fini();
PyInt_Fini();
PyFloat_Fini();
PyDict_Fini();
#ifdef Py_USING_UNICODE
/* Cleanup Unicode implementation */
_PyUnicode_Fini();
#endif
/* XXX Still allocated:
- various static ad-hoc pointers to interned strings kwa">if (strcmp(loc, "C") != 0 && strcmp(loc, "POSIX") != 0) {
/* the LC_CTYPE locale is different than C and POSIX */
return 0;
}
#if defined(HAVE_LANGINFO_H) && defined(CODESET)
const char *codeset = nl_langinfo(CODESET);
if (!codeset || codeset[0] == '\0') {
/* CODESET is not set or empty */
goto error;
}
char encoding[20]; /* longest name: "iso_646.irv_1991\0" */
if (!_Py_normalize_encoding(codeset, encoding, sizeof(encoding))) {
goto error;
}
#ifdef __hpux
if (strcmp(encoding, "roman8") == 0) {
unsigned char ch;
wchar_t wch;
size_t res;
ch = (unsigned char)0xA7;
res = _Py_mbstowcs(&wch, (char*)&ch, 1);
if (res != DECODE_ERROR && wch == L'\xA7') {
/* On HP-UX with C locale or the POSIX locale,
nl_langinfo(CODESET) announces "roman8", whereas mbstowcs() uses
Latin1 encoding in practice. Force ASCII in this case.
Roman8 decodes 0xA7 to U+00CF. Latin1 decodes 0xA7 to U+00A7. */
return 1;
}
}
#else
const char* ascii_aliases[] = {
"ascii",
/* Aliases from Lib/encodings/aliases.py */
"646",
"ansi_x3.4_1968",
"ansi_x3.4_1986",
"ansi_x3_4_1968",
"cp367",
"csascii",
"ibm367",
"iso646_us",
"iso_646.irv_1991",
"iso_ir_6",
"us",
"us_ascii",
NULL
};
int is_ascii = 0;
for (const char **alias=ascii_aliases; *alias != NULL; alias++) {
if (strcmp(encoding, *alias) == 0) {
is_ascii = 1;
break;
}
}
if (!is_ascii) {
/* nl_langinfo(CODESET) is not "ascii" or an alias of ASCII */
return 0;
}
for (unsigned int i=0x80; i<=0xff; i++) {
char ch[1];
wchar_t wch[1];
size_t res;
unsigned uch = (unsigned char)i;
ch[0] = (char)uch;
res = _Py_mbstowcs(wch, ch, 1);
if (res != DECODE_ERROR) {
/* decoding a non-ASCII character from the locale encoding succeed:
the locale encoding is not ASCII, force ASCII */
return 1;
}
}
/* None of the bytes in the range 0x80-0xff can be decoded from the locale
encoding: the locale encoding is really ASCII */
#endif /* !defined(__hpux) */
return 0;
#else
/* nl_langinfo(CODESET) is not available: always force ASCII */
return 1;
#endif /* defined(HAVE_LANGINFO_H) && defined(CODESET) */
error:
/* if an error occurred, force the ASCII encoding */
return 1;
}
int
_Py_GetForceASCII(void)
{
if (force_ascii == -1) {
force_ascii = check_force_ascii();
}
return force_ascii;
}
void
_Py_ResetForceASCII(void)
{
force_ascii = -1;
}
static int
encode_ascii(const wchar_t *text, char **str,
size_t *error_pos, const char **reason,
int raw_malloc, _Py_error_handler errors)
{
char *result = NULL, *out;
size_t len, i;
wchar_t ch;
int surrogateescape;
if (get_surrogateescape(errors, &surrogateescape) < 0) {
return -3;
}
len = wcslen(text);
/* +1 for NULL byte */
if (raw_malloc) {
result = PyMem_RawMalloc(len + 1);
}
else {
result = PyMem_Malloc(len + 1);
}
if (result == NULL) {
return -1;
}
out = result;
for (i=0; i<len; i++) {
ch = text[i];
if (ch <= 0x7f) {
/* ASCII character */
*out++ = (char)ch;
}
else if (surrogateescape && 0xdc80 <= ch && ch <= 0xdcff) {
/* UTF-8b surrogate */
*out++ = (char)(ch - 0xdc00);
}
else {
if (raw_malloc) {
PyMem_RawFree(result);
}
else {
PyMem_Free(result);
}
if (error_pos != NULL) {
*error_pos = i;
}
if (reason) {
*reason = "encoding error";
}
return -2;
}
}
*out = '\0';
*str = result;
return 0;
}
#else
int
_Py_GetForceASCII(void)
{
return 0;
}
void
_Py_ResetForceASCII(void)
{
/* nothing to do */
}
#endif /* !defined(_Py_FORCE_UTF8_FS_ENCODING) && !defined(MS_WINDOWS) */
#if !defined(HAVE_MBRTOWC) || defined(USE_FORCE_ASCII)
static int
decode_ascii(const char *arg, wchar_t **wstr, size_t *wlen,
const char **reason, _Py_error_handler errors)
{
wchar_t *res;
unsigned char *in;
wchar_t *out;
size_t argsize = strlen(arg) + 1;
int surrogateescape;
if (get_surrogateescape(errors, &surrogateescape) < 0) {
return -3;
}
if (argsize > PY_SSIZE_T_MAX / sizeof(wchar_t)) {
return -1;
}
res = PyMem_RawMalloc(argsize * sizeof(wchar_t));
if (!res) {
return -1;
}
out = res;
for (in = (unsigned char*)arg; *in; in++) {
unsigned char ch = *in;
if (ch < 128) {
*out++ = ch;
}
else {
if (!surrogateescape) {
PyMem_RawFree(res);
if (wlen) {
*wlen = in - (unsigned char*)arg;
}
if (reason) {
*reason = "decoding error";
}
return -2;
}
*out++ = 0xdc00 + ch;
}
}
*out = 0;
if (wlen != NULL) {
*wlen = out - res;
}
*wstr = res;
return 0;
}
#endif /* !HAVE_MBRTOWC */
static int
decode_current_locale(const char* arg, wchar_t **wstr, size_t *wlen,
const char **reason, _Py_error_handler errors)
{
wchar_t *res;
size_t argsize;
size_t count;
#ifdef HAVE_MBRTOWC
unsigned char *in;
wchar_t *out;
mbstate_t mbs;
#endif
int surrogateescape;
if (get_surrogateescape(errors, &surrogateescape) < 0) {
return -3;
}
#ifdef HAVE_BROKEN_MBSTOWCS
/* Some platforms have a broken implementation of
* mbstowcs which does not count the characters that
* would result from conversion. Use an upper bound.
*/
argsize = strlen(arg);
#else
argsize = _Py_mbstowcs(NULL, arg, 0);
#endif
if (argsize != DECODE_ERROR) {
if (argsize > PY_SSIZE_T_MAX / sizeof(wchar_t) - 1) {
return -1;
}
res = (wchar_t *)PyMem_RawMalloc((argsize + 1) * sizeof(wchar_t));
if (!res) {
return -1;
}
count = _Py_mbstowcs(res, arg, argsize + 1);
if (count != DECODE_ERROR) {
*wstr = res;
if (wlen != NULL) {
*wlen = count;
}
return 0;
}
PyMem_RawFree(res);
}
/* Conversion failed. Fall back to escaping with surrogateescape. */
#ifdef HAVE_MBRTOWC
/* Try conversion with mbrtwoc (C99), and escape non-decodable bytes. */
/* Overallocate; as multi-byte characters are in the argument, the
actual output could use less memory. */
argsize = strlen(arg) + 1;
if (argsize > PY_SSIZE_T_MAX / sizeof(wchar_t)) {
return -1;
}
res = (wchar_t*)PyMem_RawMalloc(argsize * sizeof(wchar_t));
if (!res) {
return -1;
}
in = (unsigned char*)arg;
out = res;
memset(&mbs, 0, sizeof mbs);
while (argsize) {
size_t converted = _Py_mbrtowc(out, (char*)in, argsize, &mbs);
if (converted == 0) {
/* Reached end of string; null char stored. */
break;
}
if (converted == INCOMPLETE_CHARACTER) {
/* Incomplete character. This should never happen,
since we provide everything that we have -
unless there is a bug in the C library, or I
misunderstood how mbrtowc works. */
goto decode_error;
}
if (converted == DECODE_ERROR) {
if (!surrogateescape) {
goto decode_error;
}
/* Decoding error. Escape as UTF-8b, and start over in the initial
shift state. */
*out++ = 0xdc00 + *in++;
argsize--;
memset(&mbs, 0, sizeof mbs);
continue;
}
// _Py_mbrtowc() reject lone surrogate characters
assert(!Py_UNICODE_IS_SURROGATE(*out));
/* successfully converted some bytes */
in += converted;
argsize -= converted;
out++;
}
if (wlen != NULL) {
*wlen = out - res;
}
*wstr = res;
return 0;
decode_error:
PyMem_RawFree(res);
if (wlen) {
*wlen = in - (unsigned char*)arg;
}
if (reason) {
*reason = "decoding error";
}
return -2;
#else /* HAVE_MBRTOWC */
/* Cannot use C locale for escaping; manually escape as if charset
is ASCII (i.e. escape all bytes > 128. This will still roundtrip
correctly in the locale's charset, which must be an ASCII superset. */
return decode_ascii(arg, wstr, wlen, reason, errors);
#endif /* HAVE_MBRTOWC */
}
/* Decode a byte string from the locale encoding.
Use the strict error handler if 'surrogateescape' is zero. Use the
surrogateescape error handler if 'surrogateescape' is non-zero: undecodable
bytes are decoded as characters in range U+DC80..U+DCFF. If a byte sequence
can be decoded as a surrogate character, escape the bytes using the
surrogateescape error handler instead of decoding them.
On success, return 0 and write the newly allocated wide character string into
*wstr (use PyMem_RawFree() to free the memory). If wlen is not NULL, write
the number of wide characters excluding the null character into *wlen.
On memory allocation failure, return -1.
On decoding error, return -2. If wlen is not NULL, write the start of
invalid byte sequence in the input string into *wlen. If reason is not NULL,
write the decoding error message into *reason.
Return -3 if the error handler 'errors' is not supported.
Use the Py_EncodeLocaleEx() function to encode the character string back to
a byte string. */
int
_Py_DecodeLocaleEx(const char* arg, wchar_t **wstr, size_t *wlen,
const char **reason,
int current_locale, _Py_error_handler errors)
{
if (current_locale) {
#ifdef _Py_FORCE_UTF8_LOCALE
return _Py_DecodeUTF8Ex(arg, strlen(arg), wstr, wlen, reason,
errors);
#else
return decode_current_locale(arg, wstr, wlen, reason, errors);
#endif
}
#ifdef _Py_FORCE_UTF8_FS_ENCODING
return _Py_DecodeUTF8Ex(arg, strlen(arg), wstr, wlen, reason,
errors);
#else
int use_utf8 = (_PyRuntime.preconfig.utf8_mode >= 1);
#ifdef MS_WINDOWS
use_utf8 |= (_PyRuntime.preconfig.legacy_windows_fs_encoding == 0);
#endif
if (use_utf8) {
return _Py_DecodeUTF8Ex(arg, strlen(arg), wstr, wlen, reason,
errors);
}
#ifdef USE_FORCE_ASCII
if (force_ascii == -1) {
force_ascii = check_force_ascii();
}
if (force_ascii) {
/* force ASCII encoding to workaround mbstowcs() issue */
return decode_ascii(arg, wstr, wlen, reason, errors);
}
#endif
return decode_current_locale(arg, wstr, wlen, reason, errors);
#endif /* !_Py_FORCE_UTF8_FS_ENCODING */
}
/* Decode a byte string from the locale encoding with the
surrogateescape error handler: undecodable bytes are decoded as characters
in range U+DC80..U+DCFF. If a byte sequence can be decoded as a surrogate
character, escape the bytes using the surrogateescape error handler instead
of decoding them.
Return a pointer to a newly allocated wide character string, use
PyMem_RawFree() to free the memory. If size is not NULL, write the number of
wide characters excluding the null character into *size
Return NULL on decoding error or memory allocation error. If *size* is not
NULL, *size is set to (size_t)-1 on memory error or set to (size_t)-2 on
decoding error.
Decoding errors should never happen, unless there is a bug in the C
library.
Use the Py_EncodeLocale() function to encode the character string back to a
byte string. */
wchar_t*
Py_DecodeLocale(const char* arg, size_t *wlen)
{
wchar_t *wstr;
int res = _Py_DecodeLocaleEx(arg, &wstr, wlen,
NULL, 0,
_Py_ERROR_SURROGATEESCAPE);
if (res != 0) {
assert(res != -3);
if (wlen != NULL) {
*wlen = (size_t)res;
}
return NULL;
}
return wstr;
}
static int
encode_current_locale(const wchar_t *text, char **str,
size_t *error_pos, const char **reason,
int raw_malloc, _Py_error_handler errors)
{
const size_t len = wcslen(text);
char *result = NULL, *bytes = NULL;
size_t i, size, converted;
wchar_t c, buf[2];
int surrogateescape;
if (get_surrogateescape(errors, &surrogateescape) < 0) {
return -3;
}
/* The function works in two steps:
1. compute the length of the output buffer in bytes (size)
2. outputs the bytes */
size = 0;
buf[1] = 0;
while (1) {
for (i=0; i < len; i++) {
c = text[i];
if (c >= 0xdc80 && c <= 0xdcff) {
if (!surrogateescape) {
goto encode_error;
}
/* UTF-8b surrogate */
if (bytes != NULL) {
*bytes++ = c - 0xdc00;
size--;
}
else {
size++;
}
continue;
}
else {
buf[0] = c;
if (bytes != NULL) {
converted = wcstombs(bytes, buf, size);
}
else {
converted = wcstombs(NULL, buf, 0);
}
if (converted == DECODE_ERROR) {
goto encode_error;
}
if (bytes != NULL) {
bytes += converted;
size -= converted;
}
else {
size += converted;
}
}
}
if (result != NULL) {
*bytes = '\0';
break;
}
size += 1; /* nul byte at the end */
if (raw_malloc) {
result = PyMem_RawMalloc(size);
}
else {
result = PyMem_Malloc(size);
}
if (result == NULL) {
return -1;
}
bytes = result;
}
*str = result;
return 0;
encode_error:
if (raw_malloc) {
PyMem_RawFree(result);
}
else {
PyMem_Free(result);
}
if (error_pos != NULL) {
*error_pos = i;
}
if (reason) {
*reason = "encoding error";
}
return -2;
}
/* Encode a string to the locale encoding.
Parameters:
* raw_malloc: if non-zero, allocate memory using PyMem_RawMalloc() instead
of PyMem_Malloc().
* current_locale: if non-zero, use the current LC_CTYPE, otherwise use
Python filesystem encoding.
* errors: error handler like "strict" or "surrogateescape".
Return value:
0: success, *str is set to a newly allocated decoded string.
-1: memory allocation failure
-2: encoding error, set *error_pos and *reason (if set).
-3: the error handler 'errors' is not supported.
*/
static int
encode_locale_ex(const wchar_t *text, char **str, size_t *error_pos,
const char **reason,
int raw_malloc, int current_locale, _Py_error_handler errors)
{
if (current_locale) {
#ifdef _Py_FORCE_UTF8_LOCALE
return _Py_EncodeUTF8Ex(text, str, error_pos, reason,
raw_malloc, errors);
#else
return encode_current_locale(text, str, error_pos, reason,
raw_malloc, errors);
#endif
}
#ifdef _Py_FORCE_UTF8_FS_ENCODING
return _Py_EncodeUTF8Ex(text, str, error_pos, reason,
raw_malloc, errors);
#else
int use_utf8 = (_PyRuntime.preconfig.utf8_mode >= 1);
#ifdef MS_WINDOWS
use_utf8 |= (_PyRuntime.preconfig.legacy_windows_fs_encoding == 0);
#endif
if (use_utf8) {
return _Py_EncodeUTF8Ex(text, str, error_pos, reason,
raw_malloc, errors);
}
#ifdef USE_FORCE_ASCII
if (force_ascii == -1) {
force_ascii = check_force_ascii();
}
if (force_ascii) {
return encode_ascii(text, str, error_pos, reason,
raw_malloc, errors);
}
#endif
return encode_current_locale(text, str, error_pos, reason,
raw_malloc, errors);
#endif /* _Py_FORCE_UTF8_FS_ENCODING */
}
static char*
encode_locale(const wchar_t *text, size_t *error_pos,
int raw_malloc, int current_locale)
{
char *str;
int res = encode_locale_ex(text, &str, error_pos, NULL,
raw_malloc, current_locale,
_Py_ERROR_SURROGATEESCAPE);
if (res != -2 && error_pos) {
*error_pos = (size_t)-1;
}
if (res != 0) {
return NULL;
}
return str;
}
/* Encode a wide character string to the locale encoding with the
surrogateescape error handler: surrogate characters in the range
U+DC80..U+DCFF are converted to bytes 0x80..0xFF.
Return a pointer to a newly allocated byte string, use PyMem_Free() to free
the memory. Return NULL on encoding or memory allocation error.
If error_pos is not NULL, *error_pos is set to (size_t)-1 on success, or set
to the index of the invalid character on encoding error.
Use the Py_DecodeLocale() function to decode the bytes string back to a wide
character string. */
char*
Py_EncodeLocale(const wchar_t *text, size_t *error_pos)
{
return encode_locale(text, error_pos, 0, 0);
}
/* Similar to Py_EncodeLocale(), but result must be freed by PyMem_RawFree()
instead of PyMem_Free(). */
char*
_Py_EncodeLocaleRaw(const wchar_t *text, size_t *error_pos)
{
return encode_locale(text, error_pos, 1, 0);
}
int
_Py_EncodeLocaleEx(const wchar_t *text, char **str,
size_t *error_pos, const char **reason,
int current_locale, _Py_error_handler errors)
{
return encode_locale_ex(text, str, error_pos, reason, 1,
current_locale, errors);
}
// Get the current locale encoding name:
//
// - Return "utf-8" if _Py_FORCE_UTF8_LOCALE macro is defined (ex: on Android)
// - Return "utf-8" if the UTF-8 Mode is enabled
// - On Windows, return the ANSI code page (ex: "cp1250")
// - Return "utf-8" if nl_langinfo(CODESET) returns an empty string.
// - Otherwise, return nl_langinfo(CODESET).
//
// Return NULL on memory allocation failure.
//
// See also config_get_locale_encoding()
wchar_t*
_Py_GetLocaleEncoding(void)
{
#ifdef _Py_FORCE_UTF8_LOCALE
// On Android langinfo.h and CODESET are missing,
// and UTF-8 is always used in mbstowcs() and wcstombs().
return _PyMem_RawWcsdup(L"utf-8");
#else
#ifdef MS_WINDOWS
wchar_t encoding[23];
unsigned int ansi_codepage = GetACP();
swprintf(encoding, Py_ARRAY_LENGTH(encoding), L"cp%u", ansi_codepage);
encoding[Py_ARRAY_LENGTH(encoding) - 1] = 0;
return _PyMem_RawWcsdup(encoding);
#else
const char *encoding = nl_langinfo(CODESET);
if (!encoding || encoding[0] == '\0') {
// Use UTF-8 if nl_langinfo() returns an empty string. It can happen on
// macOS if the LC_CTYPE locale is not supported.
return _PyMem_RawWcsdup(L"utf-8");
}
wchar_t *wstr;
int res = decode_current_locale(encoding, &wstr, NULL,
NULL, _Py_ERROR_SURROGATEESCAPE);
if (res < 0) {
return NULL;
}
return wstr;
#endif // !MS_WINDOWS
#endif // !_Py_FORCE_UTF8_LOCALE
}
PyObject *
_Py_GetLocaleEncodingObject(void)
{
wchar_t *encoding = _Py_GetLocaleEncoding();
if (encoding == NULL) {
PyErr_NoMemory();
return NULL;
}
PyObject *str = PyUnicode_FromWideChar(encoding, -1);
PyMem_RawFree(encoding);
return str;
}
#ifdef HAVE_NON_UNICODE_WCHAR_T_REPRESENTATION
/* Check whether current locale uses Unicode as internal wchar_t form. */
int
_Py_LocaleUsesNonUnicodeWchar(void)
{
/* Oracle Solaris uses non-Unicode internal wchar_t form for
non-Unicode locales and hence needs conversion to UTF first. */
char* codeset = nl_langinfo(CODESET);
if (!codeset) {
return 0;
}
/* 646 refers to ISO/IEC 646 standard that corresponds to ASCII encoding */
return (strcmp(codeset, "UTF-8") != 0 && strcmp(codeset, "646") != 0);
}
static wchar_t *
_Py_ConvertWCharForm(const wchar_t *source, Py_ssize_t size,
const char *tocode, const char *fromcode)
{
static_assert(sizeof(wchar_t) == 4, "wchar_t must be 32-bit");
/* Ensure we won't overflow the size. */
if (size > (PY_SSIZE_T_MAX / (Py_ssize_t)sizeof(wchar_t))) {
PyErr_NoMemory();
return NULL;
}
/* the string doesn't have to be NULL terminated */
wchar_t* target = PyMem_Malloc(size * sizeof(wchar_t));
if (target == NULL) {
PyErr_NoMemory();
return NULL;
}
iconv_t cd = iconv_open(tocode, fromcode);
if (cd == (iconv_t)-1) {
PyErr_Format(PyExc_ValueError, "iconv_open() failed");
PyMem_Free(target);
return NULL;
}
char *inbuf = (char *) source;
char *outbuf = (char *) target;
size_t inbytesleft = sizeof(wchar_t) * size;
size_t outbytesleft = inbytesleft;
size_t ret = iconv(cd, &inbuf, &inbytesleft, &outbuf, &outbytesleft);
if (ret == DECODE_ERROR) {
PyErr_Format(PyExc_ValueError, "iconv() failed");
PyMem_Free(target);
iconv_close(cd);
return NULL;
}
iconv_close(cd);
return target;
}
/* Convert a wide character string to the UCS-4 encoded string. This
is necessary on systems where internal form of wchar_t are not Unicode
code points (e.g. Oracle Solaris).
Return a pointer to a newly allocated string, use PyMem_Free() to free
the memory. Return NULL and raise exception on conversion or memory
allocation error. */
wchar_t *
_Py_DecodeNonUnicodeWchar(const wchar_t *native, Py_ssize_t size)
{
return _Py_ConvertWCharForm(native, size, "UCS-4-INTERNAL", "wchar_t");
}
/* Convert a UCS-4 encoded string to native wide character string. This
is necessary on systems where internal form of wchar_t are not Unicode
code points (e.g. Oracle Solaris).
The conversion is done in place. This can be done because both wchar_t
and UCS-4 use 4-byte encoding, and one wchar_t symbol always correspond
to a single UCS-4 symbol and vice versa. (This is true for Oracle Solaris,
which is currently the only system using these functions; it doesn't have
to be for other systems).
Return 0 on success. Return -1 and raise exception on conversion
or memory allocation error. */
int
_Py_EncodeNonUnicodeWchar_InPlace(wchar_t *unicode, Py_ssize_t size)
{
wchar_t* result = _Py_ConvertWCharForm(unicode, size, "wchar_t", "UCS-4-INTERNAL");
if (!result) {
return -1;
}
memcpy(unicode, result, size * sizeof(wchar_t));
PyMem_Free(result);
return 0;
}
#endif /* HAVE_NON_UNICODE_WCHAR_T_REPRESENTATION */
#ifdef MS_WINDOWS
static __int64 secs_between_epochs = 11644473600; /* Seconds between 1.1.1601 and 1.1.1970 */
static void
FILE_TIME_to_time_t_nsec(FILETIME *in_ptr, time_t *time_out, int* nsec_out)
{
/* XXX endianness. Shouldn't matter, as all Windows implementations are little-endian */
/* Cannot simply cast and dereference in_ptr,
since it might not be aligned properly */
__int64 in;
memcpy(&in, in_ptr, sizeof(in));
*nsec_out = (int)(in % 10000000) * 100; /* FILETIME is in units of 100 nsec. */
*time_out = Py_SAFE_DOWNCAST((in / 10000000) - secs_between_epochs, __int64, time_t);
}
static void
LARGE_INTEGER_to_time_t_nsec(LARGE_INTEGER *in_ptr, time_t *time_out, int* nsec_out)
{
*nsec_out = (int)(in_ptr->QuadPart % 10000000) * 100; /* FILETIME is in units of 100 nsec. */
*time_out = Py_SAFE_DOWNCAST((in_ptr->QuadPart / 10000000) - secs_between_epochs, __int64, time_t);
}
void
_Py_time_t_to_FILE_TIME(time_t time_in, int nsec_in, FILETIME *out_ptr)
{
/* XXX endianness */
__int64 out;
out = time_in + secs_between_epochs;
out = out * 10000000 + nsec_in / 100;
memcpy(out_ptr, &out, sizeof(out));
}
/* Below, we *know* that ugo+r is 0444 */
#if _S_IREAD != 0400
#error Unsupported C library
#endif
static int
attributes_to_mode(DWORD attr)
{
int m = 0;
if (attr & FILE_ATTRIBUTE_DIRECTORY)
m |= _S_IFDIR | 0111; /* IFEXEC for user,group,other */
else
m |= _S_IFREG;
if (attr & FILE_ATTRIBUTE_READONLY)
m |= 0444;
else
m |= 0666;
return m;
}
typedef union {
FILE_ID_128 id;
struct {
uint64_t st_ino;
uint64_t st_ino_high;
};
} id_128_to_ino;
void
_Py_attribute_data_to_stat(BY_HANDLE_FILE_INFORMATION *info, ULONG reparse_tag,
FILE_BASIC_INFO *basic_info, FILE_ID_INFO *id_info,
struct _Py_stat_struct *result)
{
memset(result, 0, sizeof(*result));
result->st_mode = attributes_to_mode(info->dwFileAttributes);
result->st_size = (((__int64)info->nFileSizeHigh)<<32) + info->nFileSizeLow;
result->st_dev = id_info ? id_info->VolumeSerialNumber : info->dwVolumeSerialNumber;
result->st_rdev = 0;
/* st_ctime is deprecated, but we preserve the legacy value in our caller, not here */
if (basic_info) {
LARGE_INTEGER_to_time_t_nsec(&basic_info->CreationTime, &result->st_birthtime, &result->st_birthtime_nsec);
LARGE_INTEGER_to_time_t_nsec(&basic_info->ChangeTime, &result->st_ctime, &result->st_ctime_nsec);
LARGE_INTEGER_to_time_t_nsec(&basic_info->LastWriteTime, &result->st_mtime, &result->st_mtime_nsec);
LARGE_INTEGER_to_time_t_nsec(&basic_info->LastAccessTime, &result->st_atime, &result->st_atime_nsec);
} else {
FILE_TIME_to_time_t_nsec(&info->ftCreationTime, &result->st_birthtime, &result->st_birthtime_nsec);
FILE_TIME_to_time_t_nsec(&info->ftLastWriteTime, &result->st_mtime, &result->st_mtime_nsec);
FILE_TIME_to_time_t_nsec(&info->ftLastAccessTime, &result->st_atime, &result->st_atime_nsec);
}
result->st_nlink = info->nNumberOfLinks;
if (id_info) {
id_128_to_ino file_id;
file_id.id = id_info->FileId;
result->st_ino = file_id.st_ino;
result->st_ino_high = file_id.st_ino_high;
}
if (!result->st_ino && !result->st_ino_high) {
/* should only occur for DirEntry_from_find_data, in which case the
index is likely to be zero anyway. */
result->st_ino = (((uint64_t)info->nFileIndexHigh) << 32) + info->nFileIndexLow;
}
/* bpo-37834: Only actual symlinks set the S_IFLNK flag. But lstat() will
open other name surrogate reparse points without traversing them. To
detect/handle these, check st_file_attributes and st_reparse_tag. */
result->st_reparse_tag = reparse_tag;
if (info->dwFileAttributes & FILE_ATTRIBUTE_REPARSE_POINT &&
reparse_tag == IO_REPARSE_TAG_SYMLINK) {
/* set the bits that make this a symlink */
result->st_mode = (result->st_mode & ~S_IFMT) | S_IFLNK;
}
result->st_file_attributes = info->dwFileAttributes;
}
void
_Py_stat_basic_info_to_stat(FILE_STAT_BASIC_INFORMATION *info,
struct _Py_stat_struct *result)
{
memset(result, 0, sizeof(*result));
result->st_mode = attributes_to_mode(info->FileAttributes);
result->st_size = info->EndOfFile.QuadPart;
LARGE_INTEGER_to_time_t_nsec(&info->CreationTime, &result->st_birthtime, &result->st_birthtime_nsec);
LARGE_INTEGER_to_time_t_nsec(&info->ChangeTime, &result->st_ctime, &result->st_ctime_nsec);
LARGE_INTEGER_to_time_t_nsec(&info->LastWriteTime, &result->st_mtime, &result->st_mtime_nsec);
LARGE_INTEGER_to_time_t_nsec(&info->LastAccessTime, &result->st_atime, &result->st_atime_nsec);
result->st_nlink = info->NumberOfLinks;
result->st_dev = info->VolumeSerialNumber.QuadPart;
/* File systems with less than 128-bits zero pad into this field */
id_128_to_ino file_id;
file_id.id = info->FileId128;
result->st_ino = file_id.st_ino;
result->st_ino_high = file_id.st_ino_high;
/* bpo-37834: Only actual symlinks set the S_IFLNK flag. But lstat() will
open other name surrogate reparse points without traversing them. To
detect/handle these, check st_file_attributes and st_reparse_tag. */
result->st_reparse_tag = info->ReparseTag;
if (info->FileAttributes & FILE_ATTRIBUTE_REPARSE_POINT &&
info->ReparseTag == IO_REPARSE_TAG_SYMLINK) {
/* set the bits that make this a symlink */
result->st_mode = (result->st_mode & ~S_IFMT) | S_IFLNK;
}
result->st_file_attributes = info->FileAttributes;
switch (info->DeviceType) {
case FILE_DEVICE_DISK:
case FILE_DEVICE_VIRTUAL_DISK:
case FILE_DEVICE_DFS:
case FILE_DEVICE_CD_ROM:
case FILE_DEVICE_CONTROLLER:
case FILE_DEVICE_DATALINK:
break;
case FILE_DEVICE_DISK_FILE_SYSTEM:
case FILE_DEVICE_CD_ROM_FILE_SYSTEM:
case FILE_DEVICE_NETWORK_FILE_SYSTEM:
result->st_mode = (result->st_mode & ~S_IFMT) | 0x6000; /* _S_IFBLK */
break;
case FILE_DEVICE_CONSOLE:
case FILE_DEVICE_NULL:
case FILE_DEVICE_KEYBOARD:
case FILE_DEVICE_MODEM:
case FILE_DEVICE_MOUSE:
case FILE_DEVICE_PARALLEL_PORT:
case FILE_DEVICE_PRINTER:
case FILE_DEVICE_SCREEN:
case FILE_DEVICE_SERIAL_PORT:
case FILE_DEVICE_SOUND:
result->st_mode = (result->st_mode & ~S_IFMT) | _S_IFCHR;
break;
case FILE_DEVICE_NAMED_PIPE:
result->st_mode = (result->st_mode & ~S_IFMT) | _S_IFIFO;
break;
default:
if (info->FileAttributes & FILE_ATTRIBUTE_DIRECTORY) {
result->st_mode = (result->st_mode & ~S_IFMT) | _S_IFDIR;
}
break;
}
}
#endif
/* Return information about a file.
On POSIX, use fstat().
On Windows, use GetFileType() and GetFileInformationByHandle() which support
files larger than 2 GiB. fstat() may fail with EOVERFLOW on files larger
than 2 GiB because the file size type is a signed 32-bit integer: see issue
#23152.
On Windows, set the last Windows error and return nonzero on error. On
POSIX, set errno and return nonzero on error. Fill status and return 0 on
success. */
int
_Py_fstat_noraise(int fd, struct _Py_stat_struct *status)
{
#ifdef MS_WINDOWS
BY_HANDLE_FILE_INFORMATION info;
FILE_BASIC_INFO basicInfo;
FILE_ID_INFO idInfo;
HANDLE h;
int type;
h = _Py_get_osfhandle_noraise(fd);
if (h == INVALID_HANDLE_VALUE) {
/* errno is already set by _get_osfhandle, but we also set
the Win32 error for callers who expect that */
SetLastError(ERROR_INVALID_HANDLE);
return -1;
}
memset(status, 0, sizeof(*status));
type = GetFileType(h);
if (type == FILE_TYPE_UNKNOWN) {
DWORD error = GetLastError();
if (error != 0) {
errno = winerror_to_errno(error);
return -1;
}
/* else: valid but unknown file */
}
if (type != FILE_TYPE_DISK) {
if (type == FILE_TYPE_CHAR)
status->st_mode = _S_IFCHR;
else if (type == FILE_TYPE_PIPE)
status->st_mode = _S_IFIFO;
return 0;
}
if (!GetFileInformationByHandle(h, &info) ||
!GetFileInformationByHandleEx(h, FileBasicInfo, &basicInfo, sizeof(basicInfo)) ||
!GetFileInformationByHandleEx(h, FileIdInfo, &idInfo, sizeof(idInfo))) {
/* The Win32 error is already set, but we also set errno for
callers who expect it */
errno = winerror_to_errno(GetLastError());
return -1;
}
_Py_attribute_data_to_stat(&info, 0, &basicInfo, &idInfo, status);
return 0;
#else
return fstat(fd, status);
#endif
}
/* Return information about a file.
On POSIX, use fstat().
On Windows, use GetFileType() and GetFileInformationByHandle() which support
files larger than 2 GiB. fstat() may fail with EOVERFLOW on files larger
than 2 GiB because the file size type is a signed 32-bit integer: see issue
#23152.
Raise an exception and return -1 on error. On Windows, set the last Windows
error on error. On POSIX, set errno on error. Fill status and return 0 on
success.
Release the GIL to call GetFileType() and GetFileInformationByHandle(), or
to call fstat(). The caller must hold the GIL. */
int
_Py_fstat(int fd, struct _Py_stat_struct *status)
{
int res;
assert(PyGILState_Check());
Py_BEGIN_ALLOW_THREADS
res = _Py_fstat_noraise(fd, status);
Py_END_ALLOW_THREADS
if (res != 0) {
#ifdef MS_WINDOWS
PyErr_SetFromWindowsErr(0);
#else
PyErr_SetFromErrno(PyExc_OSError);
#endif
return -1;
}
return 0;
}
/* Like _Py_stat() but with a raw filename. */
int
_Py_wstat(const wchar_t* path, struct stat *buf)
{
int err;
#ifdef MS_WINDOWS
struct _stat wstatbuf;
err = _wstat(path, &wstatbuf);
if (!err) {
buf->st_mode = wstatbuf.st_mode;
}
#else
char *fname;
fname = _Py_EncodeLocaleRaw(path, NULL);
if (fname == NULL) {
errno = EINVAL;
return -1;
}
err = stat(fname, buf);
PyMem_RawFree(fname);
#endif
return err;
}
/* Call _wstat() on Windows, or encode the path to the filesystem encoding and
call stat() otherwise. Only fill st_mode attribute on Windows.
Return 0 on success, -1 on _wstat() / stat() error, -2 if an exception was
raised. */
int
_Py_stat(PyObject *path, struct stat *statbuf)
{
#ifdef MS_WINDOWS
int err;
wchar_t *wpath = PyUnicode_AsWideCharString(path, NULL);
if (wpath == NULL)
return -2;
err = _Py_wstat(wpath, statbuf);
PyMem_Free(wpath);
return err;
#else
int ret;
PyObject *bytes;
char *cpath;
bytes = PyUnicode_EncodeFSDefault(path);
if (bytes == NULL)
return -2;
/* check for embedded null bytes */
if (PyBytes_AsStringAndSize(bytes, &cpath, NULL) == -1) {
Py_DECREF(bytes);
return -2;
}
ret = stat(cpath, statbuf);
Py_DECREF(bytes);
return ret;
#endif
}
#ifdef MS_WINDOWS
// For some Windows API partitions, SetHandleInformation() is declared
// but none of the handle flags are defined.
#ifndef HANDLE_FLAG_INHERIT
#define HANDLE_FLAG_INHERIT 0x00000001
#endif
#endif
/* This function MUST be kept async-signal-safe on POSIX when raise=0. */
static int
get_inheritable(int fd, int raise)
{
#ifdef MS_WINDOWS
HANDLE handle;
DWORD flags;
handle = _Py_get_osfhandle_noraise(fd);
if (handle == INVALID_HANDLE_VALUE) {
if (raise)
PyErr_SetFromErrno(PyExc_OSError);
return -1;
}
if (!GetHandleInformation(handle, &flags)) {
if (raise)
PyErr_SetFromWindowsErr(0);
return -1;
}
return (flags & HANDLE_FLAG_INHERIT);
#else
int flags;
flags = fcntl(fd, F_GETFD, 0);
if (flags == -1) {
if (raise)
PyErr_SetFromErrno(PyExc_OSError);
return -1;
}
return !(flags & FD_CLOEXEC);
#endif
}
/* Get the inheritable flag of the specified file descriptor.
Return 1 if the file descriptor can be inherited, 0 if it cannot,
raise an exception and return -1 on error. */
int
_Py_get_inheritable(int fd)
{
return get_inheritable(fd, 1);
}
/* This function MUST be kept async-signal-safe on POSIX when raise=0. */
static int
set_inheritable(int fd, int inheritable, int raise, int *atomic_flag_works)
{
#ifdef MS_WINDOWS
HANDLE handle;
DWORD flags;
#else
#if defined(HAVE_SYS_IOCTL_H) && defined(FIOCLEX) && defined(FIONCLEX)
static int ioctl_works = -1;
int request;
int err;
#endif
int flags, new_flags;
int res;
#endif
/* atomic_flag_works can only be used to make the file descriptor
non-inheritable */
assert(!(atomic_flag_works != NULL && inheritable));
if (atomic_flag_works != NULL && !inheritable) {
if (*atomic_flag_works == -1) {
int isInheritable = get_inheritable(fd, raise);
if (isInheritable == -1)
return -1;
*atomic_flag_works = !isInheritable;
}
if (*atomic_flag_works)
return 0;
}
#ifdef MS_WINDOWS
handle = _Py_get_osfhandle_noraise(fd);
if (handle == INVALID_HANDLE_VALUE) {
if (raise)
PyErr_SetFromErrno(PyExc_OSError);
return -1;
}
if (inheritable)
flags = HANDLE_FLAG_INHERIT;
else
flags = 0;
if (!SetHandleInformation(handle, HANDLE_FLAG_INHERIT, flags)) {
if (raise)
PyErr_SetFromWindowsErr(0);
return -1;
}
return 0;
#else
#if defined(HAVE_SYS_IOCTL_H) && defined(FIOCLEX) && defined(FIONCLEX)
if (ioctl_works != 0 && raise != 0) {
/* fast-path: ioctl() only requires one syscall */
/* caveat: raise=0 is an indicator that we must be async-signal-safe
* thus avoid using ioctl() so we skip the fast-path. */
if (inheritable)
request = FIONCLEX;
else
request = FIOCLEX;
err = ioctl(fd, request, NULL);
if (!err) {
ioctl_works = 1;
return 0;
}
#ifdef O_PATH
if (errno == EBADF) {
// bpo-44849: On Linux and FreeBSD, ioctl(FIOCLEX) fails with EBADF
// on O_PATH file descriptors. Fall through to the fcntl()
// implementation.
}
else
#endif
if (errno != ENOTTY && errno != EACCES) {
if (raise)
PyErr_SetFromErrno(PyExc_OSError);
return -1;
}
else {
/* Issue #22258: Here, ENOTTY means "Inappropriate ioctl for
device". The ioctl is declared but not supported by the kernel.
Remember that ioctl() doesn't work. It is the case on
Illumos-based OS for example.
Issue #27057: When SELinux policy disallows ioctl it will fail
with EACCES. While FIOCLEX is safe operation it may be
unavailable because ioctl was denied altogether.
This can be the case on Android. */
ioctl_works = 0;
}
/* fallback to fcntl() if ioctl() does not work */
}
#endif
/* slow-path: fcntl() requires two syscalls */
flags = fcntl(fd, F_GETFD);
if (flags < 0) {
if (raise)
PyErr_SetFromErrno(PyExc_OSError);
return -1;
}
if (inheritable) {
new_flags = flags & ~FD_CLOEXEC;
}
else {
new_flags = flags | FD_CLOEXEC;
}
if (new_flags == flags) {
/* FD_CLOEXEC flag already set/cleared: nothing to do */
return 0;
}
res = fcntl(fd, F_SETFD, new_flags);
if (res < 0) {
if (raise)
PyErr_SetFromErrno(PyExc_OSError);
return -1;
}
return 0;
#endif
}
/* Make the file descriptor non-inheritable.
Return 0 on success, set errno and return -1 on error. */
static int
make_non_inheritable(int fd)
{
return set_inheritable(fd, 0, 0, NULL);
}
/* Set the inheritable flag of the specified file descriptor.
On success: return 0, on error: raise an exception and return -1.
If atomic_flag_works is not NULL:
* if *atomic_flag_works==-1, check if the inheritable is set on the file
descriptor: if yes, set *atomic_flag_works to 1, otherwise set to 0 and
set the inheritable flag
* if *atomic_flag_works==1: do nothing
* if *atomic_flag_works==0: set inheritable flag to False
Set atomic_flag_works to NULL if no atomic flag was used to create the
file descriptor.
atomic_flag_works can only be used to make a file descriptor
non-inheritable: atomic_flag_works must be NULL if inheritable=1. */
int
_Py_set_inheritable(int fd, int inheritable, int *atomic_flag_works)
{
return set_inheritable(fd, inheritable, 1, atomic_flag_works);
}
/* Same as _Py_set_inheritable() but on error, set errno and
don't raise an exception.
This function is async-signal-safe. */
int
_Py_set_inheritable_async_safe(int fd, int inheritable, int *atomic_flag_works)
{
return set_inheritable(fd, inheritable, 0, atomic_flag_works);
}
static int
_Py_open_impl(const char *pathname, int flags, int gil_held)
{
int fd;
int async_err = 0;
#ifndef MS_WINDOWS
int *atomic_flag_works;
#endif
#ifdef MS_WINDOWS
flags |= O_NOINHERIT;
#elif defined(O_CLOEXEC)
atomic_flag_works = &_Py_open_cloexec_works;
flags |= O_CLOEXEC;
#else
atomic_flag_works = NULL;
#endif
if (gil_held) {
PyObject *pathname_obj = PyUnicode_DecodeFSDefault(pathname);
if (pathname_obj == NULL) {
return -1;
}
if (PySys_Audit("open", "OOi", pathname_obj, Py_None, flags) < 0) {
Py_DECREF(pathname_obj);
return -1;
}
do {
Py_BEGIN_ALLOW_THREADS
fd = open(pathname, flags);
Py_END_ALLOW_THREADS
} while (fd < 0
&& errno == EINTR && !(async_err = PyErr_CheckSignals()));
if (async_err) {
Py_DECREF(pathname_obj);
return -1;
}
if (fd < 0) {
PyErr_SetFromErrnoWithFilenameObjects(PyExc_OSError, pathname_obj, NULL);
Py_DECREF(pathname_obj);
return -1;
}
Py_DECREF(pathname_obj);
}
else {
fd = open(pathname, flags);
if (fd < 0)
return -1;
}
#ifndef MS_WINDOWS
if (set_inheritable(fd, 0, gil_held, atomic_flag_works) < 0) {
close(fd);
return -1;
}
#endif
return fd;
}
/* Open a file with the specified flags (wrapper to open() function).
Return a file descriptor on success. Raise an exception and return -1 on
error.
The file descriptor is created non-inheritable.
When interrupted by a signal (open() fails with EINTR), retry the syscall,
except if the Python signal handler raises an exception.
Release the GIL to call open(). The caller must hold the GIL. */
int
_Py_open(const char *pathname, int flags)
{
/* _Py_open() must be called with the GIL held. */
assert(PyGILState_Check());
return _Py_open_impl(pathname, flags, 1);
}
/* Open a file with the specified flags (wrapper to open() function).
Return a file descriptor on success. Set errno and return -1 on error.
The file descriptor is created non-inheritable.
If interrupted by a signal, fail with EINTR. */
int
_Py_open_noraise(const char *pathname, int flags)
{
return _Py_open_impl(pathname, flags, 0);
}
/* Open a file. Use _wfopen() on Windows, encode the path to the locale
encoding and use fopen() otherwise.
The file descriptor is created non-inheritable.
If interrupted by a signal, fail with EINTR. */
FILE *
_Py_wfopen(const wchar_t *path, const wchar_t *mode)
{
FILE *f;
if (PySys_Audit("open", "uui", path, mode, 0) < 0) {
return NULL;
}
#ifndef MS_WINDOWS
char *cpath;
char cmode[10];
size_t r;
r = wcstombs(cmode, mode, 10);
if (r == DECODE_ERROR || r >= 10) {
errno = EINVAL;
return NULL;
}
cpath = _Py_EncodeLocaleRaw(path, NULL);
if (cpath == NULL) {
return NULL;
}
f = fopen(cpath, cmode);
PyMem_RawFree(cpath);
#else
f = _wfopen(path, mode);
#endif
if (f == NULL)
return NULL;
if (make_non_inheritable(fileno(f)) < 0) {
fclose(f);
return NULL;
}
return f;
}
/* Open a file. Call _wfopen() on Windows, or encode the path to the filesystem
encoding and call fopen() otherwise.
Return the new file object on success. Raise an exception and return NULL
on error.
The file descriptor is created non-inheritable.
When interrupted by a signal (open() fails with EINTR), retry the syscall,
except if the Python signal handler raises an exception.
Release the GIL to call _wfopen() or fopen(). The caller must hold
the GIL. */
FILE*
_Py_fopen_obj(PyObject *path, const char *mode)
{
FILE *f;
int async_err = 0;
#ifdef MS_WINDOWS
wchar_t wmode[10];
int usize;
assert(PyGILState_Check());
if (PySys_Audit("open", "Osi", path, mode, 0) < 0) {
return NULL;
}
if (!PyUnicode_Check(path)) {
PyErr_Format(PyExc_TypeError,
"str file path expected under Windows, got %R",
Py_TYPE(path));
return NULL;
}
wchar_t *wpath = PyUnicode_AsWideCharString(path, NULL);
if (wpath == NULL)
return NULL;
usize = MultiByteToWideChar(CP_ACP, 0, mode, -1,
wmode, Py_ARRAY_LENGTH(wmode));
if (usize == 0) {
PyErr_SetFromWindowsErr(0);
PyMem_Free(wpath);
return NULL;
}
do {
Py_BEGIN_ALLOW_THREADS
f = _wfopen(wpath, wmode);
Py_END_ALLOW_THREADS
} while (f == NULL
&& errno == EINTR && !(async_err = PyErr_CheckSignals()));
int saved_errno = errno;
PyMem_Free(wpath);
#else
PyObject *bytes;
const char *path_bytes;
assert(PyGILState_Check());
if (!PyUnicode_FSConverter(path, &bytes))
return NULL;
path_bytes = PyBytes_AS_STRING(bytes);
if (PySys_Audit("open", "Osi", path, mode, 0) < 0) {
Py_DECREF(bytes);
return NULL;
}
do {
Py_BEGIN_ALLOW_THREADS
f = fopen(path_bytes, mode);
Py_END_ALLOW_THREADS
} while (f == NULL
&& errno == EINTR && !(async_err = PyErr_CheckSignals()));
int saved_errno = errno;
Py_DECREF(bytes);
#endif
if (async_err)
return NULL;
if (f == NULL) {
errno = saved_errno;
PyErr_SetFromErrnoWithFilenameObject(PyExc_OSError, path);
return NULL;
}
if (set_inheritable(fileno(f), 0, 1, NULL) < 0) {
fclose(f);
return NULL;
}
return f;
}
/* Read count bytes from fd into buf.
On success, return the number of read bytes, it can be lower than count.
If the current file offset is at or past the end of file, no bytes are read,
and read() returns zero.
On error, raise an exception, set errno and return -1.
When interrupted by a signal (read() fails with EINTR), retry the syscall.
If the Python signal handler raises an exception, the function returns -1
(the syscall is not retried).
Release the GIL to call read(). The caller must hold the GIL. */
Py_ssize_t
_Py_read(int fd, void *buf, size_t count)
{
Py_ssize_t n;
int err;
int async_err = 0;
assert(PyGILState_Check());
/* _Py_read() must not be called with an exception set, otherwise the
* caller may think that read() was interrupted by a signal and the signal
* handler raised an exception. */
assert(!PyErr_Occurred());
if (count > _PY_READ_MAX) {
count = _PY_READ_MAX;
}
_Py_BEGIN_SUPPRESS_IPH
do {
Py_BEGIN_ALLOW_THREADS
errno = 0;
#ifdef MS_WINDOWS
_doserrno = 0;
n = read(fd, buf, (int)count);
// read() on a non-blocking empty pipe fails with EINVAL, which is
// mapped from the Windows error code ERROR_NO_DATA.
if (n < 0 && errno == EINVAL) {
if (_doserrno == ERROR_NO_DATA) {
errno = EAGAIN;
}
}
#else
n = read(fd, buf, count);
#endif
/* save/restore errno because PyErr_CheckSignals()
* and PyErr_SetFromErrno() can modify it */
err = errno;
Py_END_ALLOW_THREADS
} while (n < 0 && err == EINTR &&
!(async_err = PyErr_CheckSignals()));
_Py_END_SUPPRESS_IPH
if (async_err) {
/* read() was interrupted by a signal (failed with EINTR)
* and the Python signal handler raised an exception */
errno = err;
assert(errno == EINTR && PyErr_Occurred());
return -1;
}
if (n < 0) {
PyErr_SetFromErrno(PyExc_OSError);
errno = err;
return -1;
}
return n;
}
static Py_ssize_t
_Py_write_impl(int fd, const void *buf, size_t count, int gil_held)
{
Py_ssize_t n;
int err;
int async_err = 0;
_Py_BEGIN_SUPPRESS_IPH
#ifdef MS_WINDOWS
if (count > 32767) {
/* Issue #11395: the Windows console returns an error (12: not
enough space error) on writing into stdout if stdout mode is
binary and the length is greater than 66,000 bytes (or less,
depending on heap usage). */
if (gil_held) {
Py_BEGIN_ALLOW_THREADS
if (isatty(fd)) {
count = 32767;
}
Py_END_ALLOW_THREADS
} else {
if (isatty(fd)) {
count = 32767;
}
}
}
#endif
if (count > _PY_WRITE_MAX) {
count = _PY_WRITE_MAX;
}
if (gil_held) {
do {
Py_BEGIN_ALLOW_THREADS
errno = 0;
#ifdef MS_WINDOWS
// write() on a non-blocking pipe fails with ENOSPC on Windows if
// the pipe lacks available space for the entire buffer.
int c = (int)count;
do {
_doserrno = 0;
n = write(fd, buf, c);
if (n >= 0 || errno != ENOSPC || _doserrno != 0) {
break;
}
errno = EAGAIN;
c /= 2;
} while (c > 0);
#else
n = write(fd, buf, count);
#endif
/* save/restore errno because PyErr_CheckSignals()
* and PyErr_SetFromErrno() can modify it */
err = errno;
Py_END_ALLOW_THREADS
} while (n < 0 && err == EINTR &&
!(async_err = PyErr_CheckSignals()));
}
else {
do {
errno = 0;
#ifdef MS_WINDOWS
// write() on a non-blocking pipe fails with ENOSPC on Windows if
// the pipe lacks available space for the entire buffer.
int c = (int)count;
do {
_doserrno = 0;
n = write(fd, buf, c);
if (n >= 0 || errno != ENOSPC || _doserrno != 0) {
break;
}
errno = EAGAIN;
c /= 2;
} while (c > 0);
#else
n = write(fd, buf, count);
#endif
err = errno;
} while (n < 0 && err == EINTR);
}
_Py_END_SUPPRESS_IPH
if (async_err) {
/* write() was interrupted by a signal (failed with EINTR)
and the Python signal handler raised an exception (if gil_held is
nonzero). */
errno = err;
assert(errno == EINTR && (!gil_held || PyErr_Occurred()));
return -1;
}
if (n < 0) {
if (gil_held)
PyErr_SetFromErrno(PyExc_OSError);
errno = err;
return -1;
}
return n;
}
/* Write count bytes of buf into fd.
On success, return the number of written bytes, it can be lower than count
including 0. On error, raise an exception, set errno and return -1.
When interrupted by a signal (write() fails with EINTR), retry the syscall.
If the Python signal handler raises an exception, the function returns -1
(the syscall is not retried).
Release the GIL to call write(). The caller must hold the GIL. */
Py_ssize_t
_Py_write(int fd, const void *buf, size_t count)
{
assert(PyGILState_Check());
/* _Py_write() must not be called with an exception set, otherwise the
* caller may think that write() was interrupted by a signal and the signal
* handler raised an exception. */
assert(!PyErr_Occurred());
return _Py_write_impl(fd, buf, count, 1);
}
/* Write count bytes of buf into fd.
*
* On success, return the number of written bytes, it can be lower than count
* including 0. On error, set errno and return -1.
*
* When interrupted by a signal (write() fails with EINTR), retry the syscall
* without calling the Python signal handler. */
Py_ssize_t
_Py_write_noraise(int fd, const void *buf, size_t count)
{
return _Py_write_impl(fd, buf, count, 0);
}
#ifdef HAVE_READLINK
/* Read value of symbolic link. Encode the path to the locale encoding, decode
the result from the locale encoding.
Return -1 on encoding error, on readlink() error, if the internal buffer is
too short, on decoding error, or if 'buf' is too short. */
int
_Py_wreadlink(const wchar_t *path, wchar_t *buf, size_t buflen)
{
char *cpath;
char cbuf[MAXPATHLEN];
size_t cbuf_len = Py_ARRAY_LENGTH(cbuf);
wchar_t *wbuf;
Py_ssize_t res;
size_t r1;
cpath = _Py_EncodeLocaleRaw(path, NULL);
if (cpath == NULL) {
errno = EINVAL;
return -1;
}
res = readlink(cpath, cbuf, cbuf_len);
PyMem_RawFree(cpath);
if (res == -1) {
return -1;
}
if ((size_t)res == cbuf_len) {
errno = EINVAL;
return -1;
}
cbuf[res] = '\0'; /* buf will be null terminated */
wbuf = Py_DecodeLocale(cbuf, &r1);
if (wbuf == NULL) {
errno = EINVAL;
return -1;
}
/* wbuf must have space to store the trailing NUL character */
if (buflen <= r1) {
PyMem_RawFree(wbuf);
errno = EINVAL;
return -1;
}
wcsncpy(buf, wbuf, buflen);
PyMem_RawFree(wbuf);
return (int)r1;
}
#endif
#ifdef HAVE_REALPATH
/* Return the canonicalized absolute pathname. Encode path to the locale
encoding, decode the result from the locale encoding.
Return NULL on encoding error, realpath() error, decoding error
or if 'resolved_path' is too short. */
wchar_t*
_Py_wrealpath(const wchar_t *path,
wchar_t *resolved_path, size_t resolved_path_len)
{
char *cpath;
char cresolved_path[MAXPATHLEN];
wchar_t *wresolved_path;
char *res;
size_t r;
cpath = _Py_EncodeLocaleRaw(path, NULL);
if (cpath == NULL) {
errno = EINVAL;
return NULL;
}
res = realpath(cpath, cresolved_path);
PyMem_RawFree(cpath);
if (res == NULL)
return NULL;
wresolved_path = Py_DecodeLocale(cresolved_path, &r);
if (wresolved_path == NULL) {
errno = EINVAL;
return NULL;
}
/* wresolved_path must have space to store the trailing NUL character */
if (resolved_path_len <= r) {
PyMem_RawFree(wresolved_path);
errno = EINVAL;
return NULL;
}
wcsncpy(resolved_path, wresolved_path, resolved_path_len);
PyMem_RawFree(wresolved_path);
return resolved_path;
}
#endif
int
_Py_isabs(const wchar_t *path)
{
#ifdef MS_WINDOWS
const wchar_t *tail;
HRESULT hr = PathCchSkipRoot(path, &tail);
if (FAILED(hr) || path == tail) {
return 0;
}
if (tail == &path[1] && (path[0] == SEP || path[0] == ALTSEP)) {
// Exclude paths with leading SEP
return 0;
}
if (tail == &path[2] && path[1] == L':') {
// Exclude drive-relative paths (e.g. C:filename.ext)
return 0;
}
return 1;
#else
return (path[0] == SEP);
#endif
}
/* Get an absolute path.
On error (ex: fail to get the current directory), return -1.
On memory allocation failure, set *abspath_p to NULL and return 0.
On success, return a newly allocated to *abspath_p to and return 0.
The string must be freed by PyMem_RawFree(). */
int
_Py_abspath(const wchar_t *path, wchar_t **abspath_p)
{
if (path[0] == '\0' || !wcscmp(path, L".")) {
wchar_t cwd[MAXPATHLEN + 1];
cwd[Py_ARRAY_LENGTH(cwd) - 1] = 0;
if (!_Py_wgetcwd(cwd, Py_ARRAY_LENGTH(cwd) - 1)) {
/* unable to get the current directory */
return -1;
}
*abspath_p = _PyMem_RawWcsdup(cwd);
return 0;
}
if (_Py_isabs(path)) {
*abspath_p = _PyMem_RawWcsdup(path);
return 0;
}
#ifdef MS_WINDOWS
return _PyOS_getfullpathname(path, abspath_p);
#else
wchar_t cwd[MAXPATHLEN + 1];
cwd[Py_ARRAY_LENGTH(cwd) - 1] = 0;
if (!_Py_wgetcwd(cwd, Py_ARRAY_LENGTH(cwd) - 1)) {
/* unable to get the current directory */
return -1;
}
size_t cwd_len = wcslen(cwd);
size_t path_len = wcslen(path);
size_t len = cwd_len + 1 + path_len + 1;
if (len <= (size_t)PY_SSIZE_T_MAX / sizeof(wchar_t)) {
*abspath_p = PyMem_RawMalloc(len * sizeof(wchar_t));
}
else {
*abspath_p = NULL;
}
if (*abspath_p == NULL) {
return 0;
}
wchar_t *abspath = *abspath_p;
memcpy(abspath, cwd, cwd_len * sizeof(wchar_t));
abspath += cwd_len;
*abspath = (wchar_t)SEP;
abspath++;
memcpy(abspath, path, path_len * sizeof(wchar_t));
abspath += path_len;
*abspath = 0;
return 0;
#endif
}
// The Windows Games API family implements the PathCch* APIs in the Xbox OS,
// but does not expose them yet. Load them dynamically until
// 1) they are officially exposed
// 2) we stop supporting older versions of the GDK which do not expose them
#if defined(MS_WINDOWS_GAMES) && !defined(MS_WINDOWS_DESKTOP)
HRESULT
PathCchSkipRoot(const wchar_t *path, const wchar_t **rootEnd)
{
static int initialized = 0;
typedef HRESULT(__stdcall *PPathCchSkipRoot) (PCWSTR pszPath,
PCWSTR *ppszRootEnd);
static PPathCchSkipRoot _PathCchSkipRoot;
if (initialized == 0) {
HMODULE pathapi = LoadLibraryExW(L"api-ms-win-core-path-l1-1-0.dll", NULL,
LOAD_LIBRARY_SEARCH_SYSTEM32);
if (pathapi) {
_PathCchSkipRoot = (PPathCchSkipRoot)GetProcAddress(
pathapi, "PathCchSkipRoot");
}
else {
_PathCchSkipRoot = NULL;
}
initialized = 1;
}
if (!_PathCchSkipRoot) {
return E_NOINTERFACE;
}
return _PathCchSkipRoot(path, rootEnd);
}
static HRESULT
PathCchCombineEx(wchar_t *buffer, size_t bufsize, const wchar_t *dirname,
const wchar_t *relfile, unsigned long flags)
{
static int initialized = 0;
typedef HRESULT(__stdcall *PPathCchCombineEx) (PWSTR pszPathOut,
size_t cchPathOut,
PCWSTR pszPathIn,
PCWSTR pszMore,
unsigned long dwFlags);
static PPathCchCombineEx _PathCchCombineEx;
if (initialized == 0) {
HMODULE pathapi = LoadLibraryExW(L"api-ms-win-core-path-l1-1-0.dll", NULL,
LOAD_LIBRARY_SEARCH_SYSTEM32);
if (pathapi) {
_PathCchCombineEx = (PPathCchCombineEx)GetProcAddress(
pathapi, "PathCchCombineEx");
}
else {
_PathCchCombineEx = NULL;
}
initialized = 1;
}
if (!_PathCchCombineEx) {
return E_NOINTERFACE;
}
return _PathCchCombineEx(buffer, bufsize, dirname, relfile, flags);
}
#endif /* defined(MS_WINDOWS_GAMES) && !defined(MS_WINDOWS_DESKTOP) */
// The caller must ensure "buffer" is big enough.
static int
join_relfile(wchar_t *buffer, size_t bufsize,
const wchar_t *dirname, const wchar_t *relfile)
{
#ifdef MS_WINDOWS
if (FAILED(PathCchCombineEx(buffer, bufsize, dirname, relfile,
PATHCCH_ALLOW_LONG_PATHS))) {
return -1;
}
#else
assert(!_Py_isabs(relfile));
size_t dirlen = wcslen(dirname);
size_t rellen = wcslen(relfile);
size_t maxlen = bufsize - 1;
if (maxlen > MAXPATHLEN || dirlen >= maxlen || rellen >= maxlen - dirlen) {
return -1;
}
if (dirlen == 0) {
// We do not add a leading separator.
wcscpy(buffer, relfile);
}
else {
if (dirname != buffer) {
wcscpy(buffer, dirname);
}
size_t relstart = dirlen;
if (dirlen > 1 && dirname[dirlen - 1] != SEP) {
buffer[dirlen] = SEP;
relstart += 1;
}
wcscpy(&buffer[relstart], relfile);
}
#endif
return 0;
}
/* Join the two paths together, like os.path.join(). Return NULL
if memory could not be allocated. The caller is responsible
for calling PyMem_RawFree() on the result. */
wchar_t *
_Py_join_relfile(const wchar_t *dirname, const wchar_t *relfile)
{
assert(dirname != NULL && relfile != NULL);
#ifndef MS_WINDOWS
assert(!_Py_isabs(relfile));
#endif
size_t maxlen = wcslen(dirname) + 1 + wcslen(relfile);
size_t bufsize = maxlen + 1;
wchar_t *filename = PyMem_RawMalloc(bufsize * sizeof(wchar_t));
if (filename == NULL) {
return NULL;
}
assert(wcslen(dirname) < MAXPATHLEN);
assert(wcslen(relfile) < MAXPATHLEN - wcslen(dirname));
if (join_relfile(filename, bufsize, dirname, relfile) < 0) {
PyMem_RawFree(filename);
return NULL;
}
return filename;
}
/* Join the two paths together, like os.path.join().
dirname: the target buffer with the dirname already in place,
including trailing NUL
relfile: this must be a relative path
bufsize: total allocated size of the buffer
Return -1 if anything is wrong with the path lengths. */
int
_Py_add_relfile(wchar_t *dirname, const wchar_t *relfile, size_t bufsize)
{
assert(dirname != NULL && relfile != NULL);
assert(bufsize > 0);
return join_relfile(dirname, bufsize, dirname, relfile);
}
size_t
_Py_find_basename(const wchar_t *filename)
{
for (size_t i = wcslen(filename); i > 0; --i) {
if (filename[i] == SEP) {
return i + 1;
}
}
return 0;
}
/* In-place path normalisation. Returns the start of the normalized
path, which will be within the original buffer. Guaranteed to not
make the path longer, and will not fail. 'size' is the length of
the path, if known. If -1, the first null character will be assumed
to be the end of the path. 'normsize' will be set to contain the
length of the resulting normalized path. */
wchar_t *
_Py_normpath_and_size(wchar_t *path, Py_ssize_t size, Py_ssize_t *normsize)
{
assert(path != NULL);
if ((size < 0 && !path[0]) || size == 0) {
*normsize = 0;
return path;
}
wchar_t *pEnd = size >= 0 ? &path[size] : NULL;
wchar_t *p1 = path; // sequentially scanned address in the path
wchar_t *p2 = path; // destination of a scanned character to be ljusted
wchar_t *minP2 = path; // the beginning of the destination range
wchar_t lastC = L'\0'; // the last ljusted character, p2[-1] in most cases
#define IS_END(x) (pEnd ? (x) == pEnd : !*(x))
#ifdef ALTSEP
#define IS_SEP(x) (*(x) == SEP || *(x) == ALTSEP)
#else
#define IS_SEP(x) (*(x) == SEP)
#endif
#define SEP_OR_END(x) (IS_SEP(x) || IS_END(x))
// Skip leading '.\'
if (p1[0] == L'.' && IS_SEP(&p1[1])) {
path = &path[2];
while (IS_SEP(path) && !IS_END(path)) {
path++;
}
p1 = p2 = minP2 = path;
lastC = SEP;
}
#ifdef MS_WINDOWS
// Skip past drive segment and update minP2
else if (p1[0] && p1[1] == L':') {
*p2++ = *p1++;
*p2++ = *p1++;
minP2 = p2;
lastC = L':';
}
// Skip past all \\-prefixed paths, including \\?\, \\.\,
// and network paths, including the first segment.
else if (IS_SEP(&p1[0]) && IS_SEP(&p1[1])) {
int sepCount = 2;
*p2++ = SEP;
*p2++ = SEP;
p1 += 2;
for (; !IS_END(p1) && sepCount; ++p1) {
if (IS_SEP(p1)) {
--sepCount;
*p2++ = lastC = SEP;
} else {
*p2++ = lastC = *p1;
}
}
minP2 = p2 - 1;
}
#else
// Skip past two leading SEPs
else if (IS_SEP(&p1[0]) && IS_SEP(&p1[1]) && !IS_SEP(&p1[2])) {
*p2++ = *p1++;
*p2++ = *p1++;
minP2 = p2 - 1; // Absolute path has SEP at minP2
lastC = SEP;
}
#endif /* MS_WINDOWS */
/* if pEnd is specified, check that. Else, check for null terminator */
for (; !IS_END(p1); ++p1) {
wchar_t c = *p1;
#ifdef ALTSEP
if (c == ALTSEP) {
c = SEP;
}
#endif
if (lastC == SEP) {
if (c == L'.') {
int sep_at_1 = SEP_OR_END(&p1[1]);
int sep_at_2 = !sep_at_1 && SEP_OR_END(&p1[2]);
if (sep_at_2 && p1[1] == L'.') {
wchar_t *p3 = p2;
while (p3 != minP2 && *--p3 == SEP) { }
while (p3 != minP2 && *(p3 - 1) != SEP) { --p3; }
if (p2 == minP2
|| (p3[0] == L'.' && p3[1] == L'.' && IS_SEP(&p3[2])))
{
// Previous segment is also ../, so append instead.
// Relative path does not absorb ../ at minP2 as well.
*p2++ = L'.';
*p2++ = L'.';
lastC = L'.';
} else if (p3[0] == SEP) {
// Absolute path, so absorb segment
p2 = p3 + 1;
} else {
p2 = p3;
}
p1 += 1;
} else if (sep_at_1) {
} else {
*p2++ = lastC = c;
}
} else if (c == SEP) {
} else {
*p2++ = lastC = c;
}
} else {
*p2++ = lastC = c;
}
}
*p2 = L'\0';
if (p2 != minP2) {
while (--p2 != minP2 && *p2 == SEP) {
*p2 = L'\0';
}
} else {
--p2;
}
*normsize = p2 - path + 1;
#undef SEP_OR_END
#undef IS_SEP
#undef IS_END
return path;
}
/* In-place path normalisation. Returns the start of the normalized
path, which will be within the original buffer. Guaranteed to not
make the path longer, and will not fail. 'size' is the length of
the path, if known. If -1, the first null character will be assumed
to be the end of the path. */
wchar_t *
_Py_normpath(wchar_t *path, Py_ssize_t size)
{
Py_ssize_t norm_length;
return _Py_normpath_and_size(path, size, &norm_length);
}
/* Get the current directory. buflen is the buffer size in wide characters
including the null character. Decode the path from the locale encoding.
Return NULL on getcwd() error, on decoding error, or if 'buf' is
too short. */
wchar_t*
_Py_wgetcwd(wchar_t *buf, size_t buflen)
{
#ifdef MS_WINDOWS
int ibuflen = (int)Py_MIN(buflen, INT_MAX);
return _wgetcwd(buf, ibuflen);
#else
char fname[MAXPATHLEN];
wchar_t *wname;
size_t len;
if (getcwd(fname, Py_ARRAY_LENGTH(fname)) == NULL)
return NULL;
wname = Py_DecodeLocale(fname, &len);
if (wname == NULL)
return NULL;
/* wname must have space to store the trailing NUL character */
if (buflen <= len) {
PyMem_RawFree(wname);
return NULL;
}
wcsncpy(buf, wname, buflen);
PyMem_RawFree(wname);
return buf;
#endif
}
/* Duplicate a file descriptor. The new file descriptor is created as
non-inheritable. Return a new file descriptor on success, raise an OSError
exception and return -1 on error.
The GIL is released to call dup(). The caller must hold the GIL. */
int
_Py_dup(int fd)
{
#ifdef MS_WINDOWS
HANDLE handle;
#endif
assert(PyGILState_Check());
#ifdef MS_WINDOWS
handle = _Py_get_osfhandle(fd);
if (handle == INVALID_HANDLE_VALUE)
return -1;
Py_BEGIN_ALLOW_THREADS
_Py_BEGIN_SUPPRESS_IPH
fd = dup(fd);
_Py_END_SUPPRESS_IPH
Py_END_ALLOW_THREADS
if (fd < 0) {
PyErr_SetFromErrno(PyExc_OSError);
return -1;
}
if (_Py_set_inheritable(fd, 0, NULL) < 0) {
_Py_BEGIN_SUPPRESS_IPH
close(fd);
_Py_END_SUPPRESS_IPH
return -1;
}
#elif defined(HAVE_FCNTL_H) && defined(F_DUPFD_CLOEXEC)
Py_BEGIN_ALLOW_THREADS
_Py_BEGIN_SUPPRESS_IPH
fd = fcntl(fd, F_DUPFD_CLOEXEC, 0);
_Py_END_SUPPRESS_IPH
Py_END_ALLOW_THREADS
if (fd < 0) {
PyErr_SetFromErrno(PyExc_OSError);
return -1;
}
#elif HAVE_DUP
Py_BEGIN_ALLOW_THREADS
_Py_BEGIN_SUPPRESS_IPH
fd = dup(fd);
_Py_END_SUPPRESS_IPH
Py_END_ALLOW_THREADS
if (fd < 0) {
PyErr_SetFromErrno(PyExc_OSError);
return -1;
}
if (_Py_set_inheritable(fd, 0, NULL) < 0) {
_Py_BEGIN_SUPPRESS_IPH
close(fd);
_Py_END_SUPPRESS_IPH
return -1;
}
#else
errno = ENOTSUP;
PyErr_SetFromErrno(PyExc_OSError);
return -1;
#endif
return fd;
}
#ifndef MS_WINDOWS
/* Get the blocking mode of the file descriptor.
Return 0 if the O_NONBLOCK flag is set, 1 if the flag is cleared,
raise an exception and return -1 on error. */
int
_Py_get_blocking(int fd)
{
int flags;
_Py_BEGIN_SUPPRESS_IPH
flags = fcntl(fd, F_GETFL, 0);
_Py_END_SUPPRESS_IPH
if (flags < 0) {
PyErr_SetFromErrno(PyExc_OSError);
return -1;
}
return !(flags & O_NONBLOCK);
}
/* Set the blocking mode of the specified file descriptor.
Set the O_NONBLOCK flag if blocking is False, clear the O_NONBLOCK flag
otherwise.
Return 0 on success, raise an exception and return -1 on error. */
int
_Py_set_blocking(int fd, int blocking)
{
/* bpo-41462: On VxWorks, ioctl(FIONBIO) only works on sockets.
Use fcntl() instead. */
#if defined(HAVE_SYS_IOCTL_H) && defined(FIONBIO) && !defined(__VXWORKS__)
int arg = !blocking;
if (ioctl(fd, FIONBIO, &arg) < 0)
goto error;
#else
int flags, res;
_Py_BEGIN_SUPPRESS_IPH
flags = fcntl(fd, F_GETFL, 0);
if (flags >= 0) {
if (blocking)
flags = flags & (~O_NONBLOCK);
else
flags = flags | O_NONBLOCK;
res = fcntl(fd, F_SETFL, flags);
} else {
res = -1;
}
_Py_END_SUPPRESS_IPH
if (res < 0)
goto error;
#endif
return 0;
error:
PyErr_SetFromErrno(PyExc_OSError);
return -1;
}
#else /* MS_WINDOWS */
int
_Py_get_blocking(int fd)
{
HANDLE handle;
DWORD mode;
BOOL success;
handle = _Py_get_osfhandle(fd);
if (handle == INVALID_HANDLE_VALUE) {
return -1;
}
Py_BEGIN_ALLOW_THREADS
success = GetNamedPipeHandleStateW(handle, &mode,
NULL, NULL, NULL, NULL, 0);
Py_END_ALLOW_THREADS
if (!success) {
PyErr_SetFromWindowsErr(0);
return -1;
}
return !(mode & PIPE_NOWAIT);
}
int
_Py_set_blocking(int fd, int blocking)
{
HANDLE handle;
DWORD mode;
BOOL success;
handle = _Py_get_osfhandle(fd);
if (handle == INVALID_HANDLE_VALUE) {
return -1;
}
Py_BEGIN_ALLOW_THREADS
success = GetNamedPipeHandleStateW(handle, &mode,
NULL, NULL, NULL, NULL, 0);
if (success) {
if (blocking) {
mode &= ~PIPE_NOWAIT;
}
else {
mode |= PIPE_NOWAIT;
}
success = SetNamedPipeHandleState(handle, &mode, NULL, NULL);
}
Py_END_ALLOW_THREADS
if (!success) {
PyErr_SetFromWindowsErr(0);
return -1;
}
return 0;
}
void*
_Py_get_osfhandle_noraise(int fd)
{
void *handle;
_Py_BEGIN_SUPPRESS_IPH
handle = (void*)_get_osfhandle(fd);
_Py_END_SUPPRESS_IPH
return handle;
}
void*
_Py_get_osfhandle(int fd)
{
void *handle = _Py_get_osfhandle_noraise(fd);
if (handle == INVALID_HANDLE_VALUE)
PyErr_SetFromErrno(PyExc_OSError);
return handle;
}
int
_Py_open_osfhandle_noraise(void *handle, int flags)
{
int fd;
_Py_BEGIN_SUPPRESS_IPH
fd = _open_osfhandle((intptr_t)handle, flags);
_Py_END_SUPPRESS_IPH
return fd;
}
int
_Py_open_osfhandle(void *handle, int flags)
{
int fd = _Py_open_osfhandle_noraise(handle, flags);
if (fd == -1)
PyErr_SetFromErrno(PyExc_OSError);
return fd;
}
#endif /* MS_WINDOWS */
int
_Py_GetLocaleconvNumeric(struct lconv *lc,
PyObject **decimal_point, PyObject **thousands_sep)
{
assert(decimal_point != NULL);
assert(thousands_sep != NULL);
#ifndef MS_WINDOWS
int change_locale = 0;
if ((strlen(lc->decimal_point) > 1 || ((unsigned char)lc->decimal_point[0]) > 127)) {
change_locale = 1;
}
if ((strlen(lc->thousands_sep) > 1 || ((unsigned char)lc->thousands_sep[0]) > 127)) {
change_locale = 1;
}
/* Keep a copy of the LC_CTYPE locale */
char *oldloc = NULL, *loc = NULL;
if (change_locale) {
oldloc = setlocale(LC_CTYPE, NULL);
if (!oldloc) {
PyErr_SetString(PyExc_RuntimeWarning,
"failed to get LC_CTYPE locale");
return -1;
}
oldloc = _PyMem_Strdup(oldloc);
if (!oldloc) {
PyErr_NoMemory();
return -1;
}
loc = setlocale(LC_NUMERIC, NULL);
if (loc != NULL && strcmp(loc, oldloc) == 0) {
loc = NULL;
}
if (loc != NULL) {
/* Only set the locale temporarily the LC_CTYPE locale
if LC_NUMERIC locale is different than LC_CTYPE locale and
decimal_point and/or thousands_sep are non-ASCII or longer than
1 byte */
setlocale(LC_CTYPE, loc);
}
}
#define GET_LOCALE_STRING(ATTR) PyUnicode_DecodeLocale(lc->ATTR, NULL)
#else /* MS_WINDOWS */
/* Use _W_* fields of Windows strcut lconv */
#define GET_LOCALE_STRING(ATTR) PyUnicode_FromWideChar(lc->_W_ ## ATTR, -1)
#endif /* MS_WINDOWS */
int res = -1;
*decimal_point = GET_LOCALE_STRING(decimal_point);
if (*decimal_point == NULL) {
goto done;
}
*thousands_sep = GET_LOCALE_STRING(thousands_sep);
if (*thousands_sep == NULL) {
goto done;
}
res = 0;
done:
#ifndef MS_WINDOWS
if (loc != NULL) {
setlocale(LC_CTYPE, oldloc);
}
PyMem_Free(oldloc);
#endif
return res;
#undef GET_LOCALE_STRING
}
/* Our selection logic for which function to use is as follows:
* 1. If close_range(2) is available, always prefer that; it's better for
* contiguous ranges like this than fdwalk(3) which entails iterating over
* the entire fd space and simply doing nothing for those outside the range.
* 2. If closefrom(2) is available, we'll attempt to use that next if we're
* closing up to sysconf(_SC_OPEN_MAX).
* 2a. Fallback to fdwalk(3) if we're not closing up to sysconf(_SC_OPEN_MAX),
* as that will be more performant if the range happens to have any chunk of
* non-opened fd in the middle.
* 2b. If fdwalk(3) isn't available, just do a plain close(2) loop.
*/
#ifdef __FreeBSD__
# define USE_CLOSEFROM
#endif /* __FreeBSD__ */
#ifdef HAVE_FDWALK
# define USE_FDWALK
#endif /* HAVE_FDWALK */
#ifdef USE_FDWALK
static int
_fdwalk_close_func(void *lohi, int fd)
{
int lo = ((int *)lohi)[0];
int hi = ((int *)lohi)[1];
if (fd >= hi) {
return 1;
}
else if (fd >= lo) {
/* Ignore errors */
(void)close(fd);
}
return 0;
}
#endif /* USE_FDWALK */
/* Closes all file descriptors in [first, last], ignoring errors. */
void
_Py_closerange(int first, int last)
{
first = Py_MAX(first, 0);
_Py_BEGIN_SUPPRESS_IPH
#ifdef HAVE_CLOSE_RANGE
if (close_range(first, last, 0) == 0) {
/* close_range() ignores errors when it closes file descriptors.
* Possible reasons of an error return are lack of kernel support
* or denial of the underlying syscall by a seccomp sandbox on Linux.
* Fallback to other methods in case of any error. */
}
else
#endif /* HAVE_CLOSE_RANGE */
#ifdef USE_CLOSEFROM
if (last >= sysconf(_SC_OPEN_MAX)) {
/* Any errors encountered while closing file descriptors are ignored */
closefrom(first);
}
else
#endif /* USE_CLOSEFROM */
#ifdef USE_FDWALK
{
int lohi[2];
lohi[0] = first;
lohi[1] = last + 1;
fdwalk(_fdwalk_close_func, lohi);
}
#else
{
for (int i = first; i <= last; i++) {
/* Ignore errors */
(void)close(i);
}
}
#endif /* USE_FDWALK */
_Py_END_SUPPRESS_IPH
}
|