forked from python/cpython
-
Notifications
You must be signed in to change notification settings - Fork 0
/
fileobject.c
2313 lines (2143 loc) · 56.7 KB
/
fileobject.c
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
/* File object implementation */
#include "Python.h"
#include "structmember.h"
#ifndef DONT_HAVE_SYS_TYPES_H
#include <sys/types.h>
#endif /* DONT_HAVE_SYS_TYPES_H */
#ifdef MS_WINDOWS
#define fileno _fileno
/* can simulate truncate with Win32 API functions; see file_truncate */
#define HAVE_FTRUNCATE
#define WIN32_LEAN_AND_MEAN
#include <windows.h>
#endif
#ifdef _MSC_VER
/* Need GetVersion to see if on NT so safe to use _wfopen */
#define WIN32_LEAN_AND_MEAN
#include <windows.h>
#endif /* _MSC_VER */
#ifdef macintosh
#ifdef USE_GUSI
#define HAVE_FTRUNCATE
#endif
#endif
#ifdef __MWERKS__
/* Mwerks fopen() doesn't always set errno */
#define NO_FOPEN_ERRNO
#endif
#if defined(PYOS_OS2) && defined(PYCC_GCC)
#include <io.h>
#endif
#define BUF(v) PyString_AS_STRING((PyStringObject *)v)
#ifndef DONT_HAVE_ERRNO_H
#include <errno.h>
#endif
#ifdef HAVE_GETC_UNLOCKED
#define GETC(f) getc_unlocked(f)
#define FLOCKFILE(f) flockfile(f)
#define FUNLOCKFILE(f) funlockfile(f)
#else
#define GETC(f) getc(f)
#define FLOCKFILE(f)
#define FUNLOCKFILE(f)
#endif
#ifdef WITH_UNIVERSAL_NEWLINES
/* Bits in f_newlinetypes */
#define NEWLINE_UNKNOWN 0 /* No newline seen, yet */
#define NEWLINE_CR 1 /* \r newline seen */
#define NEWLINE_LF 2 /* \n newline seen */
#define NEWLINE_CRLF 4 /* \r\n newline seen */
#endif
FILE *
PyFile_AsFile(PyObject *f)
{
if (f == NULL || !PyFile_Check(f))
return NULL;
else
return ((PyFileObject *)f)->f_fp;
}
PyObject *
PyFile_Name(PyObject *f)
{
if (f == NULL || !PyFile_Check(f))
return NULL;
else
return ((PyFileObject *)f)->f_name;
}
/* On Unix, fopen will succeed for directories.
In Python, there should be no file objects referring to
directories, so we need a check. */
static PyFileObject*
dircheck(PyFileObject* f)
{
#if defined(HAVE_FSTAT) && defined(S_IFDIR) && defined(EISDIR)
struct stat buf;
if (f->f_fp == NULL)
return f;
if (fstat(fileno(f->f_fp), &buf) == 0 &&
S_ISDIR(buf.st_mode)) {
#ifdef HAVE_STRERROR
char *msg = strerror(EISDIR);
#else
char *msg = "Is a directory";
#endif
PyObject *exc = PyObject_CallFunction(PyExc_IOError, "(is)",
EISDIR, msg);
PyErr_SetObject(PyExc_IOError, exc);
return NULL;
}
#endif
return f;
}
static PyObject *
fill_file_fields(PyFileObject *f, FILE *fp, char *name, char *mode,
int (*close)(FILE *), PyObject *wname)
{
assert(f != NULL);
assert(PyFile_Check(f));
assert(f->f_fp == NULL);
Py_DECREF(f->f_name);
Py_DECREF(f->f_mode);
#ifdef Py_USING_UNICODE
if (wname)
f->f_name = PyUnicode_FromObject(wname);
else
#endif
f->f_name = PyString_FromString(name);
f->f_mode = PyString_FromString(mode);
f->f_close = close;
f->f_softspace = 0;
f->f_binary = strchr(mode,'b') != NULL;
f->f_buf = NULL;
#ifdef WITH_UNIVERSAL_NEWLINES
f->f_univ_newline = (strchr(mode, 'U') != NULL);
f->f_newlinetypes = NEWLINE_UNKNOWN;
f->f_skipnextlf = 0;
#endif
if (f->f_name == NULL || f->f_mode == NULL)
return NULL;
f->f_fp = fp;
f = dircheck(f);
return (PyObject *) f;
}
static PyObject *
open_the_file(PyFileObject *f, char *name, char *mode)
{
assert(f != NULL);
assert(PyFile_Check(f));
#ifdef MS_WINDOWS
/* windows ignores the passed name in order to support Unicode */
assert(f->f_name != NULL);
#else
assert(name != NULL);
#endif
assert(mode != NULL);
assert(f->f_fp == NULL);
/* rexec.py can't stop a user from getting the file() constructor --
all they have to do is get *any* file object f, and then do
type(f). Here we prevent them from doing damage with it. */
if (PyEval_GetRestricted()) {
PyErr_SetString(PyExc_IOError,
"file() constructor not accessible in restricted mode");
return NULL;
}
errno = 0;
#ifdef HAVE_FOPENRF
if (*mode == '*') {
FILE *fopenRF();
f->f_fp = fopenRF(name, mode+1);
}
else
#endif
{
#ifdef WITH_UNIVERSAL_NEWLINES
if (strcmp(mode, "U") == 0 || strcmp(mode, "rU") == 0)
mode = "rb";
#else
/* Compatibility: specifying U in a Python without universal
** newlines is allowed, and the file is opened as a normal text
** file.
*/
if (strcmp(mode, "U") == 0 || strcmp(mode, "rU") == 0)
mode = "r";
#endif
#ifdef MS_WINDOWS
if (PyUnicode_Check(f->f_name)) {
PyObject *wmode;
wmode = PyUnicode_DecodeASCII(mode, strlen(mode), NULL);
if (f->f_name && wmode) {
Py_BEGIN_ALLOW_THREADS
/* PyUnicode_AS_UNICODE OK without thread
lock as it is a simple dereference. */
f->f_fp = _wfopen(PyUnicode_AS_UNICODE(f->f_name),
PyUnicode_AS_UNICODE(wmode));
Py_END_ALLOW_THREADS
}
Py_XDECREF(wmode);
}
#endif
if (NULL == f->f_fp && NULL != name) {
Py_BEGIN_ALLOW_THREADS
f->f_fp = fopen(name, mode);
Py_END_ALLOW_THREADS
}
}
if (f->f_fp == NULL) {
#ifdef NO_FOPEN_ERRNO
/* Metroworks only, wich does not always sets errno */
if (errno == 0) {
PyObject *v;
v = Py_BuildValue("(is)", 0, "Cannot open file");
if (v != NULL) {
PyErr_SetObject(PyExc_IOError, v);
Py_DECREF(v);
}
return NULL;
}
#endif
#ifdef _MSC_VER
/* MSVC 6 (Microsoft) leaves errno at 0 for bad mode strings,
* across all Windows flavors. When it sets EINVAL varies
* across Windows flavors, the exact conditions aren't
* documented, and the answer lies in the OS's implementation
* of Win32's CreateFile function (whose source is secret).
* Seems the best we can do is map EINVAL to ENOENT.
*/
if (errno == 0) /* bad mode string */
errno = EINVAL;
else if (errno == EINVAL) /* unknown, but not a mode string */
errno = ENOENT;
#endif
if (errno == EINVAL)
PyErr_Format(PyExc_IOError, "invalid mode: %s",
mode);
else
#ifdef MS_WINDOWS
PyErr_SetFromErrnoWithFilenameObject(PyExc_IOError, f->f_name);
#else
PyErr_SetFromErrnoWithFilename(PyExc_IOError, name);
#endif /* MS_WINDOWS */
f = NULL;
}
if (f != NULL)
f = dircheck(f);
return (PyObject *)f;
}
PyObject *
PyFile_FromFile(FILE *fp, char *name, char *mode, int (*close)(FILE *))
{
PyFileObject *f = (PyFileObject *)PyFile_Type.tp_new(&PyFile_Type,
NULL, NULL);
if (f != NULL) {
if (fill_file_fields(f, fp, name, mode, close, NULL) == NULL) {
Py_DECREF(f);
f = NULL;
}
}
return (PyObject *) f;
}
PyObject *
PyFile_FromString(char *name, char *mode)
{
extern int fclose(FILE *);
PyFileObject *f;
f = (PyFileObject *)PyFile_FromFile((FILE *)NULL, name, mode, fclose);
if (f != NULL) {
if (open_the_file(f, name, mode) == NULL) {
Py_DECREF(f);
f = NULL;
}
}
return (PyObject *)f;
}
void
PyFile_SetBufSize(PyObject *f, int bufsize)
{
if (bufsize >= 0) {
#ifdef HAVE_SETVBUF
int type;
switch (bufsize) {
case 0:
type = _IONBF;
break;
case 1:
type = _IOLBF;
bufsize = BUFSIZ;
break;
default:
type = _IOFBF;
}
setvbuf(((PyFileObject *)f)->f_fp, (char *)NULL,
type, bufsize);
#else /* !HAVE_SETVBUF */
if (bufsize <= 1)
setbuf(((PyFileObject *)f)->f_fp, (char *)NULL);
#endif /* !HAVE_SETVBUF */
}
}
static PyObject *
err_closed(void)
{
PyErr_SetString(PyExc_ValueError, "I/O operation on closed file");
return NULL;
}
static void drop_readahead(PyFileObject *);
/* Methods */
static void
file_dealloc(PyFileObject *f)
{
if (f->f_fp != NULL && f->f_close != NULL) {
Py_BEGIN_ALLOW_THREADS
(*f->f_close)(f->f_fp);
Py_END_ALLOW_THREADS
}
Py_XDECREF(f->f_name);
Py_XDECREF(f->f_mode);
drop_readahead(f);
f->ob_type->tp_free((PyObject *)f);
}
static PyObject *
file_repr(PyFileObject *f)
{
if (PyUnicode_Check(f->f_name)) {
#ifdef Py_USING_UNICODE
PyObject *ret = NULL;
PyObject *name;
name = PyUnicode_AsUnicodeEscapeString(f->f_name);
ret = PyString_FromFormat("<%s file u'%s', mode '%s' at %p>",
f->f_fp == NULL ? "closed" : "open",
PyString_AsString(name),
PyString_AsString(f->f_mode),
f);
Py_XDECREF(name);
return ret;
#endif
} else {
return PyString_FromFormat("<%s file '%s', mode '%s' at %p>",
f->f_fp == NULL ? "closed" : "open",
PyString_AsString(f->f_name),
PyString_AsString(f->f_mode),
f);
}
}
static PyObject *
file_close(PyFileObject *f)
{
int sts = 0;
if (f->f_fp != NULL) {
if (f->f_close != NULL) {
Py_BEGIN_ALLOW_THREADS
errno = 0;
sts = (*f->f_close)(f->f_fp);
Py_END_ALLOW_THREADS
}
f->f_fp = NULL;
}
if (sts == EOF)
return PyErr_SetFromErrno(PyExc_IOError);
if (sts != 0)
return PyInt_FromLong((long)sts);
Py_INCREF(Py_None);
return Py_None;
}
/* Our very own off_t-like type, 64-bit if possible */
#if !defined(HAVE_LARGEFILE_SUPPORT)
typedef off_t Py_off_t;
#elif SIZEOF_OFF_T >= 8
typedef off_t Py_off_t;
#elif SIZEOF_FPOS_T >= 8
typedef fpos_t Py_off_t;
#else
#error "Large file support, but neither off_t nor fpos_t is large enough."
#endif
/* a portable fseek() function
return 0 on success, non-zero on failure (with errno set) */
static int
_portable_fseek(FILE *fp, Py_off_t offset, int whence)
{
#if !defined(HAVE_LARGEFILE_SUPPORT)
return fseek(fp, offset, whence);
#elif defined(HAVE_FSEEKO) && SIZEOF_OFF_T >= 8
return fseeko(fp, offset, whence);
#elif defined(HAVE_FSEEK64)
return fseek64(fp, offset, whence);
#elif defined(__BEOS__)
return _fseek(fp, offset, whence);
#elif SIZEOF_FPOS_T >= 8
/* lacking a 64-bit capable fseek(), use a 64-bit capable fsetpos()
and fgetpos() to implement fseek()*/
fpos_t pos;
switch (whence) {
case SEEK_END:
#ifdef MS_WINDOWS
fflush(fp);
if (_lseeki64(fileno(fp), 0, 2) == -1)
return -1;
#else
if (fseek(fp, 0, SEEK_END) != 0)
return -1;
#endif
/* fall through */
case SEEK_CUR:
if (fgetpos(fp, &pos) != 0)
return -1;
offset += pos;
break;
/* case SEEK_SET: break; */
}
return fsetpos(fp, &offset);
#else
#error "Large file support, but no way to fseek."
#endif
}
/* a portable ftell() function
Return -1 on failure with errno set appropriately, current file
position on success */
static Py_off_t
_portable_ftell(FILE* fp)
{
#if !defined(HAVE_LARGEFILE_SUPPORT)
return ftell(fp);
#elif defined(HAVE_FTELLO) && SIZEOF_OFF_T >= 8
return ftello(fp);
#elif defined(HAVE_FTELL64)
return ftell64(fp);
#elif SIZEOF_FPOS_T >= 8
fpos_t pos;
if (fgetpos(fp, &pos) != 0)
return -1;
return pos;
#else
#error "Large file support, but no way to ftell."
#endif
}
static PyObject *
file_seek(PyFileObject *f, PyObject *args)
{
int whence;
int ret;
Py_off_t offset;
PyObject *offobj;
if (f->f_fp == NULL)
return err_closed();
drop_readahead(f);
whence = 0;
if (!PyArg_ParseTuple(args, "O|i:seek", &offobj, &whence))
return NULL;
#if !defined(HAVE_LARGEFILE_SUPPORT)
offset = PyInt_AsLong(offobj);
#else
offset = PyLong_Check(offobj) ?
PyLong_AsLongLong(offobj) : PyInt_AsLong(offobj);
#endif
if (PyErr_Occurred())
return NULL;
Py_BEGIN_ALLOW_THREADS
errno = 0;
ret = _portable_fseek(f->f_fp, offset, whence);
Py_END_ALLOW_THREADS
if (ret != 0) {
PyErr_SetFromErrno(PyExc_IOError);
clearerr(f->f_fp);
return NULL;
}
#ifdef WITH_UNIVERSAL_NEWLINES
f->f_skipnextlf = 0;
#endif
Py_INCREF(Py_None);
return Py_None;
}
#ifdef HAVE_FTRUNCATE
static PyObject *
file_truncate(PyFileObject *f, PyObject *args)
{
int ret;
Py_off_t newsize;
PyObject *newsizeobj;
if (f->f_fp == NULL)
return err_closed();
newsizeobj = NULL;
if (!PyArg_ParseTuple(args, "|O:truncate", &newsizeobj))
return NULL;
/* Set newsize to current postion if newsizeobj NULL, else to the
specified value. */
if (newsizeobj != NULL) {
#if !defined(HAVE_LARGEFILE_SUPPORT)
newsize = PyInt_AsLong(newsizeobj);
#else
newsize = PyLong_Check(newsizeobj) ?
PyLong_AsLongLong(newsizeobj) :
PyInt_AsLong(newsizeobj);
#endif
if (PyErr_Occurred())
return NULL;
}
else {
/* Default to current position. */
Py_BEGIN_ALLOW_THREADS
errno = 0;
newsize = _portable_ftell(f->f_fp);
Py_END_ALLOW_THREADS
if (newsize == -1)
goto onioerror;
}
/* Flush the file. */
Py_BEGIN_ALLOW_THREADS
errno = 0;
ret = fflush(f->f_fp);
Py_END_ALLOW_THREADS
if (ret != 0)
goto onioerror;
#ifdef MS_WINDOWS
/* MS _chsize doesn't work if newsize doesn't fit in 32 bits,
so don't even try using it. */
{
Py_off_t current; /* current file position */
HANDLE hFile;
int error;
/* current <- current file postion. */
if (newsizeobj == NULL)
current = newsize;
else {
Py_BEGIN_ALLOW_THREADS
errno = 0;
current = _portable_ftell(f->f_fp);
Py_END_ALLOW_THREADS
if (current == -1)
goto onioerror;
}
/* Move to newsize. */
if (current != newsize) {
Py_BEGIN_ALLOW_THREADS
errno = 0;
error = _portable_fseek(f->f_fp, newsize, SEEK_SET)
!= 0;
Py_END_ALLOW_THREADS
if (error)
goto onioerror;
}
/* Truncate. Note that this may grow the file! */
Py_BEGIN_ALLOW_THREADS
errno = 0;
hFile = (HANDLE)_get_osfhandle(fileno(f->f_fp));
error = hFile == (HANDLE)-1;
if (!error) {
error = SetEndOfFile(hFile) == 0;
if (error)
errno = EACCES;
}
Py_END_ALLOW_THREADS
if (error)
goto onioerror;
/* Restore original file position. */
if (current != newsize) {
Py_BEGIN_ALLOW_THREADS
errno = 0;
error = _portable_fseek(f->f_fp, current, SEEK_SET)
!= 0;
Py_END_ALLOW_THREADS
if (error)
goto onioerror;
}
}
#else
Py_BEGIN_ALLOW_THREADS
errno = 0;
ret = ftruncate(fileno(f->f_fp), newsize);
Py_END_ALLOW_THREADS
if (ret != 0) goto onioerror;
#endif /* !MS_WINDOWS */
Py_INCREF(Py_None);
return Py_None;
onioerror:
PyErr_SetFromErrno(PyExc_IOError);
clearerr(f->f_fp);
return NULL;
}
#endif /* HAVE_FTRUNCATE */
static PyObject *
file_tell(PyFileObject *f)
{
Py_off_t pos;
if (f->f_fp == NULL)
return err_closed();
Py_BEGIN_ALLOW_THREADS
errno = 0;
pos = _portable_ftell(f->f_fp);
Py_END_ALLOW_THREADS
if (pos == -1) {
PyErr_SetFromErrno(PyExc_IOError);
clearerr(f->f_fp);
return NULL;
}
#ifdef WITH_UNIVERSAL_NEWLINES
if (f->f_skipnextlf) {
int c;
c = GETC(f->f_fp);
if (c == '\n') {
pos++;
f->f_skipnextlf = 0;
} else if (c != EOF) ungetc(c, f->f_fp);
}
#endif
#if !defined(HAVE_LARGEFILE_SUPPORT)
return PyInt_FromLong(pos);
#else
return PyLong_FromLongLong(pos);
#endif
}
static PyObject *
file_fileno(PyFileObject *f)
{
if (f->f_fp == NULL)
return err_closed();
return PyInt_FromLong((long) fileno(f->f_fp));
}
static PyObject *
file_flush(PyFileObject *f)
{
int res;
if (f->f_fp == NULL)
return err_closed();
Py_BEGIN_ALLOW_THREADS
errno = 0;
res = fflush(f->f_fp);
Py_END_ALLOW_THREADS
if (res != 0) {
PyErr_SetFromErrno(PyExc_IOError);
clearerr(f->f_fp);
return NULL;
}
Py_INCREF(Py_None);
return Py_None;
}
static PyObject *
file_isatty(PyFileObject *f)
{
long res;
if (f->f_fp == NULL)
return err_closed();
Py_BEGIN_ALLOW_THREADS
res = isatty((int)fileno(f->f_fp));
Py_END_ALLOW_THREADS
return PyBool_FromLong(res);
}
#if BUFSIZ < 8192
#define SMALLCHUNK 8192
#else
#define SMALLCHUNK BUFSIZ
#endif
#if SIZEOF_INT < 4
#define BIGCHUNK (512 * 32)
#else
#define BIGCHUNK (512 * 1024)
#endif
static size_t
new_buffersize(PyFileObject *f, size_t currentsize)
{
#ifdef HAVE_FSTAT
off_t pos, end;
struct stat st;
if (fstat(fileno(f->f_fp), &st) == 0) {
end = st.st_size;
/* The following is not a bug: we really need to call lseek()
*and* ftell(). The reason is that some stdio libraries
mistakenly flush their buffer when ftell() is called and
the lseek() call it makes fails, thereby throwing away
data that cannot be recovered in any way. To avoid this,
we first test lseek(), and only call ftell() if lseek()
works. We can't use the lseek() value either, because we
need to take the amount of buffered data into account.
(Yet another reason why stdio stinks. :-) */
#ifdef USE_GUSI2
pos = lseek(fileno(f->f_fp), 1L, SEEK_CUR);
pos = lseek(fileno(f->f_fp), -1L, SEEK_CUR);
#else
pos = lseek(fileno(f->f_fp), 0L, SEEK_CUR);
#endif
if (pos >= 0) {
pos = ftell(f->f_fp);
}
if (pos < 0)
clearerr(f->f_fp);
if (end > pos && pos >= 0)
return currentsize + end - pos + 1;
/* Add 1 so if the file were to grow we'd notice. */
}
#endif
if (currentsize > SMALLCHUNK) {
/* Keep doubling until we reach BIGCHUNK;
then keep adding BIGCHUNK. */
if (currentsize <= BIGCHUNK)
return currentsize + currentsize;
else
return currentsize + BIGCHUNK;
}
return currentsize + SMALLCHUNK;
}
#if defined(EWOULDBLOCK) && defined(EAGAIN) && EWOULDBLOCK != EAGAIN
#define BLOCKED_ERRNO(x) ((x) == EWOULDBLOCK || (x) == EAGAIN)
#else
#ifdef EWOULDBLOCK
#define BLOCKED_ERRNO(x) ((x) == EWOULDBLOCK)
#else
#ifdef EAGAIN
#define BLOCKED_ERRNO(x) ((x) == EAGAIN)
#else
#define BLOCKED_ERRNO(x) 0
#endif
#endif
#endif
static PyObject *
file_read(PyFileObject *f, PyObject *args)
{
long bytesrequested = -1;
size_t bytesread, buffersize, chunksize;
PyObject *v;
if (f->f_fp == NULL)
return err_closed();
if (!PyArg_ParseTuple(args, "|l:read", &bytesrequested))
return NULL;
if (bytesrequested < 0)
buffersize = new_buffersize(f, (size_t)0);
else
buffersize = bytesrequested;
if (buffersize > INT_MAX) {
PyErr_SetString(PyExc_OverflowError,
"requested number of bytes is more than a Python string can hold");
return NULL;
}
v = PyString_FromStringAndSize((char *)NULL, buffersize);
if (v == NULL)
return NULL;
bytesread = 0;
for (;;) {
Py_BEGIN_ALLOW_THREADS
errno = 0;
chunksize = Py_UniversalNewlineFread(BUF(v) + bytesread,
buffersize - bytesread, f->f_fp, (PyObject *)f);
Py_END_ALLOW_THREADS
if (chunksize == 0) {
if (!ferror(f->f_fp))
break;
clearerr(f->f_fp);
/* When in non-blocking mode, data shouldn't
* be discarded if a blocking signal was
* received. That will also happen if
* chunksize != 0, but bytesread < buffersize. */
if (bytesread > 0 && BLOCKED_ERRNO(errno))
break;
PyErr_SetFromErrno(PyExc_IOError);
Py_DECREF(v);
return NULL;
}
bytesread += chunksize;
if (bytesread < buffersize) {
clearerr(f->f_fp);
break;
}
if (bytesrequested < 0) {
buffersize = new_buffersize(f, buffersize);
if (_PyString_Resize(&v, buffersize) < 0)
return NULL;
} else {
/* Got what was requested. */
break;
}
}
if (bytesread != buffersize)
_PyString_Resize(&v, bytesread);
return v;
}
static PyObject *
file_readinto(PyFileObject *f, PyObject *args)
{
char *ptr;
int ntodo;
size_t ndone, nnow;
if (f->f_fp == NULL)
return err_closed();
if (!PyArg_ParseTuple(args, "w#", &ptr, &ntodo))
return NULL;
ndone = 0;
while (ntodo > 0) {
Py_BEGIN_ALLOW_THREADS
errno = 0;
nnow = Py_UniversalNewlineFread(ptr+ndone, ntodo, f->f_fp,
(PyObject *)f);
Py_END_ALLOW_THREADS
if (nnow == 0) {
if (!ferror(f->f_fp))
break;
PyErr_SetFromErrno(PyExc_IOError);
clearerr(f->f_fp);
return NULL;
}
ndone += nnow;
ntodo -= nnow;
}
return PyInt_FromLong((long)ndone);
}
/**************************************************************************
Routine to get next line using platform fgets().
Under MSVC 6:
+ MS threadsafe getc is very slow (multiple layers of function calls before+
after each character, to lock+unlock the stream).
+ The stream-locking functions are MS-internal -- can't access them from user
code.
+ There's nothing Tim could find in the MS C or platform SDK libraries that
can worm around this.
+ MS fgets locks/unlocks only once per line; it's the only hook we have.
So we use fgets for speed(!), despite that it's painful.
MS realloc is also slow.
Reports from other platforms on this method vs getc_unlocked (which MS doesn't
have):
Linux a wash
Solaris a wash
Tru64 Unix getline_via_fgets significantly faster
CAUTION: The C std isn't clear about this: in those cases where fgets
writes something into the buffer, can it write into any position beyond the
required trailing null byte? MSVC 6 fgets does not, and no platform is (yet)
known on which it does; and it would be a strange way to code fgets. Still,
getline_via_fgets may not work correctly if it does. The std test
test_bufio.py should fail if platform fgets() routinely writes beyond the
trailing null byte. #define DONT_USE_FGETS_IN_GETLINE to disable this code.
**************************************************************************/
/* Use this routine if told to, or by default on non-get_unlocked()
* platforms unless told not to. Yikes! Let's spell that out:
* On a platform with getc_unlocked():
* By default, use getc_unlocked().
* If you want to use fgets() instead, #define USE_FGETS_IN_GETLINE.
* On a platform without getc_unlocked():
* By default, use fgets().
* If you don't want to use fgets(), #define DONT_USE_FGETS_IN_GETLINE.
*/
#if !defined(USE_FGETS_IN_GETLINE) && !defined(HAVE_GETC_UNLOCKED)
#define USE_FGETS_IN_GETLINE
#endif
#if defined(DONT_USE_FGETS_IN_GETLINE) && defined(USE_FGETS_IN_GETLINE)
#undef USE_FGETS_IN_GETLINE
#endif
#ifdef USE_FGETS_IN_GETLINE
static PyObject*
getline_via_fgets(FILE *fp)
{
/* INITBUFSIZE is the maximum line length that lets us get away with the fast
* no-realloc, one-fgets()-call path. Boosting it isn't free, because we have
* to fill this much of the buffer with a known value in order to figure out
* how much of the buffer fgets() overwrites. So if INITBUFSIZE is larger
* than "most" lines, we waste time filling unused buffer slots. 100 is
* surely adequate for most peoples' email archives, chewing over source code,
* etc -- "regular old text files".
* MAXBUFSIZE is the maximum line length that lets us get away with the less
* fast (but still zippy) no-realloc, two-fgets()-call path. See above for
* cautions about boosting that. 300 was chosen because the worst real-life
* text-crunching job reported on Python-Dev was a mail-log crawler where over
* half the lines were 254 chars.
*/
#define INITBUFSIZE 100
#define MAXBUFSIZE 300
char* p; /* temp */
char buf[MAXBUFSIZE];
PyObject* v; /* the string object result */
char* pvfree; /* address of next free slot */
char* pvend; /* address one beyond last free slot */
size_t nfree; /* # of free buffer slots; pvend-pvfree */
size_t total_v_size; /* total # of slots in buffer */
size_t increment; /* amount to increment the buffer */
/* Optimize for normal case: avoid _PyString_Resize if at all
* possible via first reading into stack buffer "buf".
*/
total_v_size = INITBUFSIZE; /* start small and pray */
pvfree = buf;
for (;;) {
Py_BEGIN_ALLOW_THREADS
pvend = buf + total_v_size;
nfree = pvend - pvfree;
memset(pvfree, '\n', nfree);
p = fgets(pvfree, nfree, fp);
Py_END_ALLOW_THREADS
if (p == NULL) {
clearerr(fp);
if (PyErr_CheckSignals())
return NULL;
v = PyString_FromStringAndSize(buf, pvfree - buf);
return v;
}
/* fgets read *something* */
p = memchr(pvfree, '\n', nfree);
if (p != NULL) {
/* Did the \n come from fgets or from us?
* Since fgets stops at the first \n, and then writes
* \0, if it's from fgets a \0 must be next. But if
* that's so, it could not have come from us, since
* the \n's we filled the buffer with have only more
* \n's to the right.
*/
if (p+1 < pvend && *(p+1) == '\0') {
/* It's from fgets: we win! In particular,
* we haven't done any mallocs yet, and can
* build the final result on the first try.
*/
++p; /* include \n from fgets */
}
else {
/* Must be from us: fgets didn't fill the
* buffer and didn't find a newline, so it
* must be the last and newline-free line of
* the file.
*/
assert(p > pvfree && *(p-1) == '\0');
--p; /* don't include \0 from fgets */
}
v = PyString_FromStringAndSize(buf, p - buf);
return v;
}
/* yuck: fgets overwrote all the newlines, i.e. the entire
* buffer. So this line isn't over yet, or maybe it is but
* we're exactly at EOF. If we haven't already, try using the
* rest of the stack buffer.
*/
assert(*(pvend-1) == '\0');
if (pvfree == buf) {
pvfree = pvend - 1; /* overwrite trailing null */
total_v_size = MAXBUFSIZE;
}
else
break;
}
/* The stack buffer isn't big enough; malloc a string object and read
* into its buffer.
*/
total_v_size = MAXBUFSIZE << 1;
v = PyString_FromStringAndSize((char*)NULL, (int)total_v_size);
if (v == NULL)
return v;
/* copy over everything except the last null byte */
memcpy(BUF(v), buf, MAXBUFSIZE-1);