2
An implementation of the new I/O lib as defined by PEP 3116 - "New I/O"
4
Classes defined here: UnsupportedOperation, BlockingIOError.
5
Functions defined here: open().
7
Mostly written by Amaury Forgeot d'Arc
10
#define PY_SSIZE_T_CLEAN
12
#include "structmember.h"
13
#include "_iomodule.h"
15
#ifdef HAVE_SYS_TYPES_H
16
#include <sys/types.h>
17
#endif /* HAVE_SYS_TYPES_H */
19
#ifdef HAVE_SYS_STAT_H
21
#endif /* HAVE_SYS_STAT_H */
24
/* Various interned strings */
26
PyObject *_PyIO_str_close;
27
PyObject *_PyIO_str_closed;
28
PyObject *_PyIO_str_decode;
29
PyObject *_PyIO_str_encode;
30
PyObject *_PyIO_str_fileno;
31
PyObject *_PyIO_str_flush;
32
PyObject *_PyIO_str_getstate;
33
PyObject *_PyIO_str_isatty;
34
PyObject *_PyIO_str_newlines;
35
PyObject *_PyIO_str_nl;
36
PyObject *_PyIO_str_read;
37
PyObject *_PyIO_str_read1;
38
PyObject *_PyIO_str_readable;
39
PyObject *_PyIO_str_readall;
40
PyObject *_PyIO_str_readinto;
41
PyObject *_PyIO_str_readline;
42
PyObject *_PyIO_str_reset;
43
PyObject *_PyIO_str_seek;
44
PyObject *_PyIO_str_seekable;
45
PyObject *_PyIO_str_setstate;
46
PyObject *_PyIO_str_tell;
47
PyObject *_PyIO_str_truncate;
48
PyObject *_PyIO_str_writable;
49
PyObject *_PyIO_str_write;
51
PyObject *_PyIO_empty_str;
52
PyObject *_PyIO_empty_bytes;
56
PyDoc_STRVAR(module_doc,
57
"The io module provides the Python interfaces to stream handling. The\n"
58
"builtin open function is defined in this module.\n"
60
"At the top of the I/O hierarchy is the abstract base class IOBase. It\n"
61
"defines the basic interface to a stream. Note, however, that there is no\n"
62
"separation between reading and writing to streams; implementations are\n"
63
"allowed to raise an IOError if they do not support a given operation.\n"
65
"Extending IOBase is RawIOBase which deals simply with the reading and\n"
66
"writing of raw bytes to a stream. FileIO subclasses RawIOBase to provide\n"
67
"an interface to OS files.\n"
69
"BufferedIOBase deals with buffering on a raw byte stream (RawIOBase). Its\n"
70
"subclasses, BufferedWriter, BufferedReader, and BufferedRWPair buffer\n"
71
"streams that are readable, writable, and both respectively.\n"
72
"BufferedRandom provides a buffered interface to random access\n"
73
"streams. BytesIO is a simple stream of in-memory bytes.\n"
75
"Another IOBase subclass, TextIOBase, deals with the encoding and decoding\n"
76
"of streams into text. TextIOWrapper, which extends it, is a buffered text\n"
77
"interface to a buffered raw stream (`BufferedIOBase`). Finally, StringIO\n"
78
"is a in-memory stream for text.\n"
80
"Argument names are not part of the specification, and only the arguments\n"
81
"of open() are intended to be used as keyword arguments.\n"
85
"DEFAULT_BUFFER_SIZE\n"
87
" An int containing the default buffer size used by the module's buffered\n"
88
" I/O classes. open() uses the file's blksize (as obtained by os.stat) if\n"
94
* The main open() function
96
PyDoc_STRVAR(open_doc,
97
"open(file, mode='r', buffering=-1, encoding=None,\n"
98
" errors=None, newline=None, closefd=True, opener=None) -> file object\n"
100
"Open file and return a stream. Raise IOError upon failure.\n"
102
"file is either a text or byte string giving the name (and the path\n"
103
"if the file isn't in the current working directory) of the file to\n"
104
"be opened or an integer file descriptor of the file to be\n"
105
"wrapped. (If a file descriptor is given, it is closed when the\n"
106
"returned I/O object is closed, unless closefd is set to False.)\n"
108
"mode is an optional string that specifies the mode in which the file\n"
109
"is opened. It defaults to 'r' which means open for reading in text\n"
110
"mode. Other common values are 'w' for writing (truncating the file if\n"
111
"it already exists), 'x' for creating and writing to a new file, and\n"
112
"'a' for appending (which on some Unix systems, means that all writes\n"
113
"append to the end of the file regardless of the current seek position).\n"
114
"In text mode, if encoding is not specified the encoding used is platform\n"
115
"dependent: locale.getpreferredencoding(False) is called to get the\n"
116
"current locale encoding. (For reading and writing raw bytes use binary\n"
117
"mode and leave encoding unspecified.) The available modes are:\n"
119
"========= ===============================================================\n"
120
"Character Meaning\n"
121
"--------- ---------------------------------------------------------------\n"
122
"'r' open for reading (default)\n"
123
"'w' open for writing, truncating the file first\n"
124
"'x' create a new file and open it for writing\n"
125
"'a' open for writing, appending to the end of the file if it exists\n"
127
"'t' text mode (default)\n"
128
"'+' open a disk file for updating (reading and writing)\n"
129
"'U' universal newline mode (deprecated)\n"
130
"========= ===============================================================\n"
132
"The default mode is 'rt' (open for reading text). For binary random\n"
133
"access, the mode 'w+b' opens and truncates the file to 0 bytes, while\n"
134
"'r+b' opens the file without truncation. The 'x' mode implies 'w' and\n"
135
"raises an `FileExistsError` if the file already exists.\n"
137
"Python distinguishes between files opened in binary and text modes,\n"
138
"even when the underlying operating system doesn't. Files opened in\n"
139
"binary mode (appending 'b' to the mode argument) return contents as\n"
140
"bytes objects without any decoding. In text mode (the default, or when\n"
141
"'t' is appended to the mode argument), the contents of the file are\n"
142
"returned as strings, the bytes having been first decoded using a\n"
143
"platform-dependent encoding or using the specified encoding if given.\n"
145
"'U' mode is deprecated and will raise an exception in future versions\n"
146
"of Python. It has no effect in Python 3. Use newline to control\n"
147
"universal newlines mode.\n"
149
"buffering is an optional integer used to set the buffering policy.\n"
150
"Pass 0 to switch buffering off (only allowed in binary mode), 1 to select\n"
151
"line buffering (only usable in text mode), and an integer > 1 to indicate\n"
152
"the size of a fixed-size chunk buffer. When no buffering argument is\n"
153
"given, the default buffering policy works as follows:\n"
155
"* Binary files are buffered in fixed-size chunks; the size of the buffer\n"
156
" is chosen using a heuristic trying to determine the underlying device's\n"
157
" \"block size\" and falling back on `io.DEFAULT_BUFFER_SIZE`.\n"
158
" On many systems, the buffer will typically be 4096 or 8192 bytes long.\n"
160
"* \"Interactive\" text files (files for which isatty() returns True)\n"
161
" use line buffering. Other text files use the policy described above\n"
162
" for binary files.\n"
164
"encoding is the name of the encoding used to decode or encode the\n"
165
"file. This should only be used in text mode. The default encoding is\n"
166
"platform dependent, but any encoding supported by Python can be\n"
167
"passed. See the codecs module for the list of supported encodings.\n"
169
"errors is an optional string that specifies how encoding errors are to\n"
170
"be handled---this argument should not be used in binary mode. Pass\n"
171
"'strict' to raise a ValueError exception if there is an encoding error\n"
172
"(the default of None has the same effect), or pass 'ignore' to ignore\n"
173
"errors. (Note that ignoring encoding errors can lead to data loss.)\n"
174
"See the documentation for codecs.register or run 'help(codecs.Codec)'\n"
175
"for a list of the permitted encoding error strings.\n"
177
"newline controls how universal newlines works (it only applies to text\n"
178
"mode). It can be None, '', '\\n', '\\r', and '\\r\\n'. It works as\n"
181
"* On input, if newline is None, universal newlines mode is\n"
182
" enabled. Lines in the input can end in '\\n', '\\r', or '\\r\\n', and\n"
183
" these are translated into '\\n' before being returned to the\n"
184
" caller. If it is '', universal newline mode is enabled, but line\n"
185
" endings are returned to the caller untranslated. If it has any of\n"
186
" the other legal values, input lines are only terminated by the given\n"
187
" string, and the line ending is returned to the caller untranslated.\n"
189
"* On output, if newline is None, any '\\n' characters written are\n"
190
" translated to the system default line separator, os.linesep. If\n"
191
" newline is '' or '\\n', no translation takes place. If newline is any\n"
192
" of the other legal values, any '\\n' characters written are translated\n"
193
" to the given string.\n"
195
"If closefd is False, the underlying file descriptor will be kept open\n"
196
"when the file is closed. This does not work when a file name is given\n"
197
"and must be True in that case.\n"
199
"A custom opener can be used by passing a callable as *opener*. The\n"
200
"underlying file descriptor for the file object is then obtained by\n"
201
"calling *opener* with (*file*, *flags*). *opener* must return an open\n"
202
"file descriptor (passing os.open as *opener* results in functionality\n"
203
"similar to passing None).\n"
205
"open() returns a file object whose type depends on the mode, and\n"
206
"through which the standard file operations such as reading and writing\n"
207
"are performed. When open() is used to open a file in a text mode ('w',\n"
208
"'r', 'wt', 'rt', etc.), it returns a TextIOWrapper. When used to open\n"
209
"a file in a binary mode, the returned class varies: in read binary\n"
210
"mode, it returns a BufferedReader; in write binary and append binary\n"
211
"modes, it returns a BufferedWriter, and in read/write mode, it returns\n"
212
"a BufferedRandom.\n"
214
"It is also possible to use a string or bytearray as a file for both\n"
215
"reading and writing. For strings StringIO can be used like a file\n"
216
"opened in a text mode, and for bytes a BytesIO can be used like a file\n"
217
"opened in a binary mode.\n"
221
io_open(PyObject *self, PyObject *args, PyObject *kwds)
223
char *kwlist[] = {"file", "mode", "buffering",
224
"encoding", "errors", "newline",
225
"closefd", "opener", NULL};
226
PyObject *file, *opener = Py_None;
228
int buffering = -1, closefd = 1;
229
char *encoding = NULL, *errors = NULL, *newline = NULL;
232
int creating = 0, reading = 0, writing = 0, appending = 0, updating = 0;
233
int text = 0, binary = 0, universal = 0;
236
int line_buffering, isatty;
238
PyObject *raw, *modeobj = NULL, *buffer = NULL, *wrapper = NULL;
240
_Py_IDENTIFIER(isatty);
241
_Py_IDENTIFIER(fileno);
242
_Py_IDENTIFIER(mode);
244
if (!PyArg_ParseTupleAndKeywords(args, kwds, "O|sizzziO:open", kwlist,
245
&file, &mode, &buffering,
246
&encoding, &errors, &newline,
247
&closefd, &opener)) {
251
if (!PyUnicode_Check(file) &&
252
!PyBytes_Check(file) &&
253
!PyNumber_Check(file)) {
254
PyErr_Format(PyExc_TypeError, "invalid file: %R", file);
259
for (i = 0; i < strlen(mode); i++) {
292
/* c must not be duplicated */
293
if (strchr(mode+i+1, c)) {
295
PyErr_Format(PyExc_ValueError, "invalid mode: '%s'", mode);
302
if (creating) *(m++) = 'x';
303
if (reading) *(m++) = 'r';
304
if (writing) *(m++) = 'w';
305
if (appending) *(m++) = 'a';
306
if (updating) *(m++) = '+';
309
/* Parameters validation */
311
if (writing || appending) {
312
PyErr_SetString(PyExc_ValueError,
313
"can't use U and writing mode at once");
316
if (PyErr_WarnEx(PyExc_DeprecationWarning,
317
"'U' mode is deprecated", 1) < 0)
322
if (text && binary) {
323
PyErr_SetString(PyExc_ValueError,
324
"can't have text and binary mode at once");
328
if (creating + reading + writing + appending > 1) {
329
PyErr_SetString(PyExc_ValueError,
330
"must have exactly one of create/read/write/append mode");
334
if (binary && encoding != NULL) {
335
PyErr_SetString(PyExc_ValueError,
336
"binary mode doesn't take an encoding argument");
340
if (binary && errors != NULL) {
341
PyErr_SetString(PyExc_ValueError,
342
"binary mode doesn't take an errors argument");
346
if (binary && newline != NULL) {
347
PyErr_SetString(PyExc_ValueError,
348
"binary mode doesn't take a newline argument");
352
/* Create the Raw file stream */
353
raw = PyObject_CallFunction((PyObject *)&PyFileIO_Type,
354
"OsiO", file, rawmode, closefd, opener);
358
modeobj = PyUnicode_FromString(mode);
364
PyObject *res = _PyObject_CallMethodId(raw, &PyId_isatty, NULL);
367
isatty = PyLong_AsLong(res);
369
if (isatty == -1 && PyErr_Occurred())
373
if (buffering == 1 || (buffering < 0 && isatty)) {
381
buffering = DEFAULT_BUFFER_SIZE;
382
#ifdef HAVE_STRUCT_STAT_ST_BLKSIZE
386
PyObject *res = _PyObject_CallMethodId(raw, &PyId_fileno, NULL);
390
fileno = PyLong_AsLong(res);
392
if (fileno == -1 && PyErr_Occurred())
395
if (fstat(fileno, &st) >= 0 && st.st_blksize > 1)
396
buffering = st.st_blksize;
401
PyErr_SetString(PyExc_ValueError,
402
"invalid buffering size");
406
/* if not buffering, returns the raw file object */
407
if (buffering == 0) {
409
PyErr_SetString(PyExc_ValueError,
410
"can't have unbuffered text I/O");
418
/* wraps into a buffered file */
420
PyObject *Buffered_class;
423
Buffered_class = (PyObject *)&PyBufferedRandom_Type;
424
else if (creating || writing || appending)
425
Buffered_class = (PyObject *)&PyBufferedWriter_Type;
427
Buffered_class = (PyObject *)&PyBufferedReader_Type;
429
PyErr_Format(PyExc_ValueError,
430
"unknown mode: '%s'", mode);
434
buffer = PyObject_CallFunction(Buffered_class, "Oi", raw, buffering);
441
/* if binary, returns the buffered file */
447
/* wraps into a TextIOWrapper */
448
wrapper = PyObject_CallFunction((PyObject *)&PyTextIOWrapper_Type,
451
encoding, errors, newline,
457
if (_PyObject_SetAttrId(wrapper, &PyId_mode, modeobj) < 0)
471
* Private helpers for the io module.
475
PyNumber_AsOff_t(PyObject *item, PyObject *err)
479
PyObject *value = PyNumber_Index(item);
483
/* We're done if PyLong_AsSsize_t() returns without error. */
484
result = PyLong_AsOff_t(value);
485
if (result != -1 || !(runerr = PyErr_Occurred()))
488
/* Error handling code -- only manage OverflowError differently */
489
if (!PyErr_GivenExceptionMatches(runerr, PyExc_OverflowError))
493
/* If no error-handling desired then the default clipping
497
assert(PyLong_Check(value));
498
/* Whether or not it is less than or equal to
499
zero is determined by the sign of ob_size
501
if (_PyLong_Sign(value) < 0)
502
result = PY_OFF_T_MIN;
504
result = PY_OFF_T_MAX;
507
/* Otherwise replace the error with caller's error object. */
509
"cannot fit '%.200s' into an offset-sized integer",
510
item->ob_type->tp_name);
519
/* Basically the "n" format code with the ability to turn None into -1. */
521
_PyIO_ConvertSsize_t(PyObject *obj, void *result) {
523
if (obj == Py_None) {
526
else if (PyNumber_Check(obj)) {
527
limit = PyNumber_AsSsize_t(obj, PyExc_OverflowError);
528
if (limit == -1 && PyErr_Occurred())
532
PyErr_Format(PyExc_TypeError,
533
"integer argument expected, got '%.200s'",
534
Py_TYPE(obj)->tp_name);
537
*((Py_ssize_t *)result) = limit;
543
_PyIO_get_locale_module(_PyIO_State *state)
546
if (state->locale_module != NULL) {
547
assert(PyWeakref_CheckRef(state->locale_module));
548
mod = PyWeakref_GET_OBJECT(state->locale_module);
549
if (mod != Py_None) {
553
Py_CLEAR(state->locale_module);
555
mod = PyImport_ImportModule("_bootlocale");
558
state->locale_module = PyWeakref_NewRef(mod, NULL);
559
if (state->locale_module == NULL) {
568
iomodule_traverse(PyObject *mod, visitproc visit, void *arg) {
569
_PyIO_State *state = IO_MOD_STATE(mod);
570
if (!state->initialized)
572
if (state->locale_module != NULL) {
573
Py_VISIT(state->locale_module);
575
Py_VISIT(state->unsupported_operation);
581
iomodule_clear(PyObject *mod) {
582
_PyIO_State *state = IO_MOD_STATE(mod);
583
if (!state->initialized)
585
if (state->locale_module != NULL)
586
Py_CLEAR(state->locale_module);
587
Py_CLEAR(state->unsupported_operation);
592
iomodule_free(PyObject *mod) {
601
static PyMethodDef module_methods[] = {
602
{"open", (PyCFunction)io_open, METH_VARARGS|METH_KEYWORDS, open_doc},
606
struct PyModuleDef _PyIO_Module = {
607
PyModuleDef_HEAD_INIT,
615
(freefunc)iomodule_free,
621
PyObject *m = PyModule_Create(&_PyIO_Module);
622
_PyIO_State *state = NULL;
625
state = IO_MOD_STATE(m);
626
state->initialized = 0;
628
#define ADD_TYPE(type, name) \
629
if (PyType_Ready(type) < 0) \
632
if (PyModule_AddObject(m, name, (PyObject *)type) < 0) { \
637
/* DEFAULT_BUFFER_SIZE */
638
if (PyModule_AddIntMacro(m, DEFAULT_BUFFER_SIZE) < 0)
641
/* UnsupportedOperation inherits from ValueError and IOError */
642
state->unsupported_operation = PyObject_CallFunction(
643
(PyObject *)&PyType_Type, "s(OO){}",
644
"UnsupportedOperation", PyExc_ValueError, PyExc_IOError);
645
if (state->unsupported_operation == NULL)
647
Py_INCREF(state->unsupported_operation);
648
if (PyModule_AddObject(m, "UnsupportedOperation",
649
state->unsupported_operation) < 0)
652
/* BlockingIOError, for compatibility */
653
Py_INCREF(PyExc_BlockingIOError);
654
if (PyModule_AddObject(m, "BlockingIOError",
655
(PyObject *) PyExc_BlockingIOError) < 0)
658
/* Concrete base types of the IO ABCs.
659
(the ABCs themselves are declared through inheritance in io.py)
661
ADD_TYPE(&PyIOBase_Type, "_IOBase");
662
ADD_TYPE(&PyRawIOBase_Type, "_RawIOBase");
663
ADD_TYPE(&PyBufferedIOBase_Type, "_BufferedIOBase");
664
ADD_TYPE(&PyTextIOBase_Type, "_TextIOBase");
666
/* Implementation of concrete IO objects. */
668
PyFileIO_Type.tp_base = &PyRawIOBase_Type;
669
ADD_TYPE(&PyFileIO_Type, "FileIO");
672
PyBytesIO_Type.tp_base = &PyBufferedIOBase_Type;
673
ADD_TYPE(&PyBytesIO_Type, "BytesIO");
674
if (PyType_Ready(&_PyBytesIOBuffer_Type) < 0)
678
PyStringIO_Type.tp_base = &PyTextIOBase_Type;
679
ADD_TYPE(&PyStringIO_Type, "StringIO");
682
PyBufferedReader_Type.tp_base = &PyBufferedIOBase_Type;
683
ADD_TYPE(&PyBufferedReader_Type, "BufferedReader");
686
PyBufferedWriter_Type.tp_base = &PyBufferedIOBase_Type;
687
ADD_TYPE(&PyBufferedWriter_Type, "BufferedWriter");
690
PyBufferedRWPair_Type.tp_base = &PyBufferedIOBase_Type;
691
ADD_TYPE(&PyBufferedRWPair_Type, "BufferedRWPair");
694
PyBufferedRandom_Type.tp_base = &PyBufferedIOBase_Type;
695
ADD_TYPE(&PyBufferedRandom_Type, "BufferedRandom");
698
PyTextIOWrapper_Type.tp_base = &PyTextIOBase_Type;
699
ADD_TYPE(&PyTextIOWrapper_Type, "TextIOWrapper");
701
/* IncrementalNewlineDecoder */
702
ADD_TYPE(&PyIncrementalNewlineDecoder_Type, "IncrementalNewlineDecoder");
704
/* Interned strings */
705
#define ADD_INTERNED(name) \
706
if (!_PyIO_str_ ## name && \
707
!(_PyIO_str_ ## name = PyUnicode_InternFromString(# name))) \
716
ADD_INTERNED(getstate)
718
ADD_INTERNED(newlines)
721
ADD_INTERNED(readable)
722
ADD_INTERNED(readall)
723
ADD_INTERNED(readinto)
724
ADD_INTERNED(readline)
727
ADD_INTERNED(seekable)
728
ADD_INTERNED(setstate)
730
ADD_INTERNED(truncate)
732
ADD_INTERNED(writable)
735
!(_PyIO_str_nl = PyUnicode_InternFromString("\n")))
738
if (!_PyIO_empty_str &&
739
!(_PyIO_empty_str = PyUnicode_FromStringAndSize(NULL, 0)))
741
if (!_PyIO_empty_bytes &&
742
!(_PyIO_empty_bytes = PyBytes_FromStringAndSize(NULL, 0)))
745
!(_PyIO_zero = PyLong_FromLong(0L)))
748
state->initialized = 1;
753
Py_XDECREF(state->unsupported_operation);