blob: 1dbd563e2c69e959f784230e67e0d9fa3752c2bc [file] [log] [blame]
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +00001/*
2 An implementation of the new I/O lib as defined by PEP 3116 - "New I/O"
Brett Cannonefb00c02012-02-29 18:31:31 -05003
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +00004 Classes defined here: UnsupportedOperation, BlockingIOError.
5 Functions defined here: open().
Brett Cannonefb00c02012-02-29 18:31:31 -05006
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +00007 Mostly written by Amaury Forgeot d'Arc
8*/
9
10#define PY_SSIZE_T_CLEAN
11#include "Python.h"
12#include "structmember.h"
13#include "_iomodule.h"
14
15#ifdef HAVE_SYS_TYPES_H
16#include <sys/types.h>
17#endif /* HAVE_SYS_TYPES_H */
18
19#ifdef HAVE_SYS_STAT_H
20#include <sys/stat.h>
21#endif /* HAVE_SYS_STAT_H */
22
23
24/* Various interned strings */
25
26PyObject *_PyIO_str_close;
27PyObject *_PyIO_str_closed;
28PyObject *_PyIO_str_decode;
29PyObject *_PyIO_str_encode;
30PyObject *_PyIO_str_fileno;
31PyObject *_PyIO_str_flush;
32PyObject *_PyIO_str_getstate;
33PyObject *_PyIO_str_isatty;
34PyObject *_PyIO_str_newlines;
35PyObject *_PyIO_str_nl;
36PyObject *_PyIO_str_read;
37PyObject *_PyIO_str_read1;
38PyObject *_PyIO_str_readable;
Victor Stinnerb57f1082011-05-26 00:19:38 +020039PyObject *_PyIO_str_readall;
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +000040PyObject *_PyIO_str_readinto;
41PyObject *_PyIO_str_readline;
42PyObject *_PyIO_str_reset;
43PyObject *_PyIO_str_seek;
44PyObject *_PyIO_str_seekable;
Antoine Pitroue4501852009-05-14 18:55:55 +000045PyObject *_PyIO_str_setstate;
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +000046PyObject *_PyIO_str_tell;
47PyObject *_PyIO_str_truncate;
48PyObject *_PyIO_str_writable;
49PyObject *_PyIO_str_write;
50
51PyObject *_PyIO_empty_str;
52PyObject *_PyIO_empty_bytes;
Antoine Pitroue4501852009-05-14 18:55:55 +000053PyObject *_PyIO_zero;
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +000054
55
56PyDoc_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"
59"\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"
Éric Araujofab97662012-02-26 02:14:08 +010062"separation between reading and writing to streams; implementations are\n"
Andrew Svetlov737fb892012-12-18 21:14:22 +020063"allowed to raise an IOError if they do not support a given operation.\n"
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +000064"\n"
65"Extending IOBase is RawIOBase which deals simply with the reading and\n"
Benjamin Peterson8f2b6652009-04-05 00:46:27 +000066"writing of raw bytes to a stream. FileIO subclasses RawIOBase to provide\n"
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +000067"an interface to OS files.\n"
68"\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"
74"\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"
79"\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"
82"\n"
83"data:\n"
84"\n"
85"DEFAULT_BUFFER_SIZE\n"
86"\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"
89" possible.\n"
90 );
91
92
93/*
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +000094 * The main open() function
95 */
96PyDoc_STRVAR(open_doc,
Georg Brandle40ee502010-07-11 09:33:39 +000097"open(file, mode='r', buffering=-1, encoding=None,\n"
Ross Lagerwall59142db2011-10-31 20:34:46 +020098" errors=None, newline=None, closefd=True, opener=None) -> file object\n"
Georg Brandl5e8f6d12009-12-23 10:30:45 +000099"\n"
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000100"Open file and return a stream. Raise IOError upon failure.\n"
101"\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"
107"\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"
Charles-François Natalidc3044c2012-01-09 22:40:02 +0100111"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"
Victor Stinnerf86a5e82012-06-05 13:43:22 +0200115"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"
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000118"\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"
Charles-François Natalidc3044c2012-01-09 22:40:02 +0100124"'x' create a new file and open it for writing\n"
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000125"'a' open for writing, appending to the end of the file if it exists\n"
126"'b' binary mode\n"
127"'t' text mode (default)\n"
128"'+' open a disk file for updating (reading and writing)\n"
Serhiy Storchaka6787a382013-11-23 22:12:06 +0200129"'U' universal newline mode (deprecated)\n"
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000130"========= ===============================================================\n"
131"\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"
Charles-François Natalidc3044c2012-01-09 22:40:02 +0100134"'r+b' opens the file without truncation. The 'x' mode implies 'w' and\n"
135"raises an `FileExistsError` if the file already exists.\n"
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000136"\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"
144"\n"
Serhiy Storchaka6787a382013-11-23 22:12:06 +0200145"'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"
148"\n"
Antoine Pitroud5587bc2009-12-19 21:08:31 +0000149"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"
154"\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"
159"\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"
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000163"\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"
168"\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"
Andrew Kuchlingc7b6c502013-06-16 12:58:48 -0400174"See the documentation for codecs.register or run 'help(codecs.Codec)'\n"
175"for a list of the permitted encoding error strings.\n"
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000176"\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"
179"follows:\n"
180"\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"
188"\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"
Ezio Melotti16d2b472012-09-18 07:20:18 +0300191" newline is '' or '\\n', no translation takes place. If newline is any\n"
Victor Stinner401e17d2012-08-04 01:18:56 +0200192" of the other legal values, any '\\n' characters written are translated\n"
193" to the given string.\n"
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000194"\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"
198"\n"
Ross Lagerwall59142db2011-10-31 20:34:46 +0200199"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"
204"\n"
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000205"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"
213"\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"
218 );
219
220static PyObject *
221io_open(PyObject *self, PyObject *args, PyObject *kwds)
222{
223 char *kwlist[] = {"file", "mode", "buffering",
224 "encoding", "errors", "newline",
Ross Lagerwall59142db2011-10-31 20:34:46 +0200225 "closefd", "opener", NULL};
226 PyObject *file, *opener = Py_None;
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000227 char *mode = "r";
228 int buffering = -1, closefd = 1;
229 char *encoding = NULL, *errors = NULL, *newline = NULL;
230 unsigned i;
231
Charles-François Natalidc3044c2012-01-09 22:40:02 +0100232 int creating = 0, reading = 0, writing = 0, appending = 0, updating = 0;
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000233 int text = 0, binary = 0, universal = 0;
234
Christian Heimes89ff3c72012-09-10 03:50:48 +0200235 char rawmode[6], *m;
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000236 int line_buffering, isatty;
237
Serhiy Storchakaf10063e2014-06-09 13:32:34 +0300238 PyObject *raw, *modeobj = NULL, *buffer, *wrapper, *result = NULL;
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000239
Martin v. Löwisbd928fe2011-10-14 10:20:37 +0200240 _Py_IDENTIFIER(isatty);
241 _Py_IDENTIFIER(fileno);
Martin v. Löwis767046a2011-10-14 15:35:36 +0200242 _Py_IDENTIFIER(mode);
Serhiy Storchakaf10063e2014-06-09 13:32:34 +0300243 _Py_IDENTIFIER(close);
Martin v. Löwisafe55bb2011-10-09 10:38:36 +0200244
Ross Lagerwall59142db2011-10-31 20:34:46 +0200245 if (!PyArg_ParseTupleAndKeywords(args, kwds, "O|sizzziO:open", kwlist,
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000246 &file, &mode, &buffering,
247 &encoding, &errors, &newline,
Ross Lagerwall59142db2011-10-31 20:34:46 +0200248 &closefd, &opener)) {
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000249 return NULL;
250 }
251
252 if (!PyUnicode_Check(file) &&
253 !PyBytes_Check(file) &&
254 !PyNumber_Check(file)) {
255 PyErr_Format(PyExc_TypeError, "invalid file: %R", file);
256 return NULL;
257 }
258
259 /* Decode mode */
260 for (i = 0; i < strlen(mode); i++) {
261 char c = mode[i];
262
263 switch (c) {
Charles-François Natalidc3044c2012-01-09 22:40:02 +0100264 case 'x':
265 creating = 1;
266 break;
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000267 case 'r':
268 reading = 1;
269 break;
270 case 'w':
271 writing = 1;
272 break;
273 case 'a':
274 appending = 1;
275 break;
276 case '+':
277 updating = 1;
278 break;
279 case 't':
280 text = 1;
281 break;
282 case 'b':
283 binary = 1;
284 break;
285 case 'U':
286 universal = 1;
287 reading = 1;
288 break;
289 default:
290 goto invalid_mode;
291 }
292
293 /* c must not be duplicated */
294 if (strchr(mode+i+1, c)) {
295 invalid_mode:
296 PyErr_Format(PyExc_ValueError, "invalid mode: '%s'", mode);
297 return NULL;
298 }
299
300 }
301
302 m = rawmode;
Charles-François Natalidc3044c2012-01-09 22:40:02 +0100303 if (creating) *(m++) = 'x';
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000304 if (reading) *(m++) = 'r';
305 if (writing) *(m++) = 'w';
306 if (appending) *(m++) = 'a';
307 if (updating) *(m++) = '+';
308 *m = '\0';
309
310 /* Parameters validation */
311 if (universal) {
312 if (writing || appending) {
313 PyErr_SetString(PyExc_ValueError,
314 "can't use U and writing mode at once");
315 return NULL;
316 }
Serhiy Storchaka6787a382013-11-23 22:12:06 +0200317 if (PyErr_WarnEx(PyExc_DeprecationWarning,
318 "'U' mode is deprecated", 1) < 0)
319 return NULL;
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000320 reading = 1;
321 }
322
323 if (text && binary) {
324 PyErr_SetString(PyExc_ValueError,
325 "can't have text and binary mode at once");
326 return NULL;
327 }
328
Charles-François Natalidc3044c2012-01-09 22:40:02 +0100329 if (creating + reading + writing + appending > 1) {
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000330 PyErr_SetString(PyExc_ValueError,
Charles-François Natalidc3044c2012-01-09 22:40:02 +0100331 "must have exactly one of create/read/write/append mode");
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000332 return NULL;
333 }
334
335 if (binary && encoding != NULL) {
336 PyErr_SetString(PyExc_ValueError,
337 "binary mode doesn't take an encoding argument");
338 return NULL;
339 }
340
341 if (binary && errors != NULL) {
342 PyErr_SetString(PyExc_ValueError,
343 "binary mode doesn't take an errors argument");
344 return NULL;
345 }
346
347 if (binary && newline != NULL) {
348 PyErr_SetString(PyExc_ValueError,
349 "binary mode doesn't take a newline argument");
350 return NULL;
351 }
352
353 /* Create the Raw file stream */
354 raw = PyObject_CallFunction((PyObject *)&PyFileIO_Type,
Ross Lagerwall59142db2011-10-31 20:34:46 +0200355 "OsiO", file, rawmode, closefd, opener);
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000356 if (raw == NULL)
357 return NULL;
Serhiy Storchakaf10063e2014-06-09 13:32:34 +0300358 result = raw;
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000359
360 modeobj = PyUnicode_FromString(mode);
361 if (modeobj == NULL)
362 goto error;
363
364 /* buffering */
365 {
Martin v. Löwisafe55bb2011-10-09 10:38:36 +0200366 PyObject *res = _PyObject_CallMethodId(raw, &PyId_isatty, NULL);
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000367 if (res == NULL)
368 goto error;
369 isatty = PyLong_AsLong(res);
370 Py_DECREF(res);
371 if (isatty == -1 && PyErr_Occurred())
372 goto error;
373 }
374
375 if (buffering == 1 || (buffering < 0 && isatty)) {
376 buffering = -1;
377 line_buffering = 1;
378 }
379 else
380 line_buffering = 0;
381
382 if (buffering < 0) {
383 buffering = DEFAULT_BUFFER_SIZE;
384#ifdef HAVE_STRUCT_STAT_ST_BLKSIZE
385 {
386 struct stat st;
387 long fileno;
Martin v. Löwisafe55bb2011-10-09 10:38:36 +0200388 PyObject *res = _PyObject_CallMethodId(raw, &PyId_fileno, NULL);
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000389 if (res == NULL)
390 goto error;
391
392 fileno = PyLong_AsLong(res);
393 Py_DECREF(res);
394 if (fileno == -1 && PyErr_Occurred())
395 goto error;
396
Antoine Pitrouea5d17d2010-10-27 19:45:43 +0000397 if (fstat(fileno, &st) >= 0 && st.st_blksize > 1)
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000398 buffering = st.st_blksize;
399 }
400#endif
401 }
402 if (buffering < 0) {
403 PyErr_SetString(PyExc_ValueError,
404 "invalid buffering size");
405 goto error;
406 }
407
408 /* if not buffering, returns the raw file object */
409 if (buffering == 0) {
410 if (!binary) {
411 PyErr_SetString(PyExc_ValueError,
412 "can't have unbuffered text I/O");
413 goto error;
414 }
415
416 Py_DECREF(modeobj);
Serhiy Storchakaf10063e2014-06-09 13:32:34 +0300417 return result;
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000418 }
419
420 /* wraps into a buffered file */
421 {
422 PyObject *Buffered_class;
423
424 if (updating)
425 Buffered_class = (PyObject *)&PyBufferedRandom_Type;
Charles-François Natalidc3044c2012-01-09 22:40:02 +0100426 else if (creating || writing || appending)
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000427 Buffered_class = (PyObject *)&PyBufferedWriter_Type;
428 else if (reading)
429 Buffered_class = (PyObject *)&PyBufferedReader_Type;
430 else {
431 PyErr_Format(PyExc_ValueError,
432 "unknown mode: '%s'", mode);
433 goto error;
434 }
435
436 buffer = PyObject_CallFunction(Buffered_class, "Oi", raw, buffering);
437 }
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000438 if (buffer == NULL)
439 goto error;
Serhiy Storchakaf10063e2014-06-09 13:32:34 +0300440 result = buffer;
441 Py_DECREF(raw);
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000442
443
444 /* if binary, returns the buffered file */
445 if (binary) {
446 Py_DECREF(modeobj);
Serhiy Storchakaf10063e2014-06-09 13:32:34 +0300447 return result;
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000448 }
449
450 /* wraps into a TextIOWrapper */
451 wrapper = PyObject_CallFunction((PyObject *)&PyTextIOWrapper_Type,
452 "Osssi",
453 buffer,
454 encoding, errors, newline,
455 line_buffering);
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000456 if (wrapper == NULL)
457 goto error;
Serhiy Storchakaf10063e2014-06-09 13:32:34 +0300458 result = wrapper;
459 Py_DECREF(buffer);
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000460
Martin v. Löwis767046a2011-10-14 15:35:36 +0200461 if (_PyObject_SetAttrId(wrapper, &PyId_mode, modeobj) < 0)
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000462 goto error;
463 Py_DECREF(modeobj);
Serhiy Storchakaf10063e2014-06-09 13:32:34 +0300464 return result;
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000465
466 error:
Serhiy Storchakaf10063e2014-06-09 13:32:34 +0300467 if (result != NULL) {
468 PyObject *exc, *val, *tb;
469 PyErr_Fetch(&exc, &val, &tb);
470 if (_PyObject_CallMethodId(result, &PyId_close, NULL) != NULL)
471 PyErr_Restore(exc, val, tb);
472 else {
473 PyObject *val2;
474 PyErr_NormalizeException(&exc, &val, &tb);
475 Py_XDECREF(exc);
476 Py_XDECREF(tb);
477 PyErr_Fetch(&exc, &val2, &tb);
478 PyErr_NormalizeException(&exc, &val2, &tb);
479 PyException_SetContext(val2, val);
480 PyErr_Restore(exc, val2, tb);
481 }
482 Py_DECREF(result);
483 }
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000484 Py_XDECREF(modeobj);
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000485 return NULL;
486}
487
488/*
489 * Private helpers for the io module.
490 */
491
492Py_off_t
493PyNumber_AsOff_t(PyObject *item, PyObject *err)
494{
495 Py_off_t result;
496 PyObject *runerr;
497 PyObject *value = PyNumber_Index(item);
498 if (value == NULL)
499 return -1;
500
501 /* We're done if PyLong_AsSsize_t() returns without error. */
502 result = PyLong_AsOff_t(value);
503 if (result != -1 || !(runerr = PyErr_Occurred()))
504 goto finish;
505
506 /* Error handling code -- only manage OverflowError differently */
507 if (!PyErr_GivenExceptionMatches(runerr, PyExc_OverflowError))
508 goto finish;
509
510 PyErr_Clear();
511 /* If no error-handling desired then the default clipping
512 is sufficient.
513 */
514 if (!err) {
515 assert(PyLong_Check(value));
516 /* Whether or not it is less than or equal to
517 zero is determined by the sign of ob_size
518 */
519 if (_PyLong_Sign(value) < 0)
520 result = PY_OFF_T_MIN;
521 else
522 result = PY_OFF_T_MAX;
523 }
524 else {
525 /* Otherwise replace the error with caller's error object. */
526 PyErr_Format(err,
527 "cannot fit '%.200s' into an offset-sized integer",
528 item->ob_type->tp_name);
529 }
530
531 finish:
532 Py_DECREF(value);
533 return result;
534}
535
Benjamin Petersonbf5ff762009-12-13 19:25:34 +0000536
537/* Basically the "n" format code with the ability to turn None into -1. */
Brett Cannonefb00c02012-02-29 18:31:31 -0500538int
Benjamin Petersonbf5ff762009-12-13 19:25:34 +0000539_PyIO_ConvertSsize_t(PyObject *obj, void *result) {
540 Py_ssize_t limit;
541 if (obj == Py_None) {
542 limit = -1;
543 }
544 else if (PyNumber_Check(obj)) {
545 limit = PyNumber_AsSsize_t(obj, PyExc_OverflowError);
546 if (limit == -1 && PyErr_Occurred())
547 return 0;
548 }
549 else {
550 PyErr_Format(PyExc_TypeError,
551 "integer argument expected, got '%.200s'",
552 Py_TYPE(obj)->tp_name);
553 return 0;
554 }
555 *((Py_ssize_t *)result) = limit;
556 return 1;
557}
558
559
Antoine Pitrou712cb732013-12-21 15:51:54 +0100560_PyIO_State *
561_PyIO_get_module_state(void)
562{
563 PyObject *mod = PyState_FindModule(&_PyIO_Module);
564 _PyIO_State *state;
565 if (mod == NULL || (state = IO_MOD_STATE(mod)) == NULL) {
566 PyErr_SetString(PyExc_RuntimeError,
567 "could not find io module state "
568 "(interpreter shutdown?)");
569 return NULL;
570 }
571 return state;
572}
573
Antoine Pitrou932ff832013-08-01 21:04:50 +0200574PyObject *
575_PyIO_get_locale_module(_PyIO_State *state)
576{
577 PyObject *mod;
578 if (state->locale_module != NULL) {
579 assert(PyWeakref_CheckRef(state->locale_module));
580 mod = PyWeakref_GET_OBJECT(state->locale_module);
581 if (mod != Py_None) {
582 Py_INCREF(mod);
583 return mod;
584 }
585 Py_CLEAR(state->locale_module);
586 }
Antoine Pitroufd4722c2013-10-12 00:13:50 +0200587 mod = PyImport_ImportModule("_bootlocale");
Antoine Pitrou932ff832013-08-01 21:04:50 +0200588 if (mod == NULL)
589 return NULL;
590 state->locale_module = PyWeakref_NewRef(mod, NULL);
591 if (state->locale_module == NULL) {
592 Py_DECREF(mod);
593 return NULL;
594 }
595 return mod;
596}
597
598
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000599static int
600iomodule_traverse(PyObject *mod, visitproc visit, void *arg) {
601 _PyIO_State *state = IO_MOD_STATE(mod);
602 if (!state->initialized)
603 return 0;
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000604 if (state->locale_module != NULL) {
605 Py_VISIT(state->locale_module);
606 }
607 Py_VISIT(state->unsupported_operation);
608 return 0;
609}
610
Benjamin Petersonbf5ff762009-12-13 19:25:34 +0000611
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000612static int
613iomodule_clear(PyObject *mod) {
614 _PyIO_State *state = IO_MOD_STATE(mod);
615 if (!state->initialized)
616 return 0;
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000617 if (state->locale_module != NULL)
618 Py_CLEAR(state->locale_module);
619 Py_CLEAR(state->unsupported_operation);
620 return 0;
621}
622
623static void
624iomodule_free(PyObject *mod) {
625 iomodule_clear(mod);
626}
627
Benjamin Petersonbf5ff762009-12-13 19:25:34 +0000628
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000629/*
630 * Module definition
631 */
632
633static PyMethodDef module_methods[] = {
634 {"open", (PyCFunction)io_open, METH_VARARGS|METH_KEYWORDS, open_doc},
635 {NULL, NULL}
636};
637
638struct PyModuleDef _PyIO_Module = {
639 PyModuleDef_HEAD_INIT,
640 "io",
641 module_doc,
642 sizeof(_PyIO_State),
643 module_methods,
644 NULL,
645 iomodule_traverse,
646 iomodule_clear,
647 (freefunc)iomodule_free,
648};
649
650PyMODINIT_FUNC
651PyInit__io(void)
652{
653 PyObject *m = PyModule_Create(&_PyIO_Module);
654 _PyIO_State *state = NULL;
655 if (m == NULL)
656 return NULL;
657 state = IO_MOD_STATE(m);
658 state->initialized = 0;
659
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000660#define ADD_TYPE(type, name) \
661 if (PyType_Ready(type) < 0) \
662 goto fail; \
663 Py_INCREF(type); \
664 if (PyModule_AddObject(m, name, (PyObject *)type) < 0) { \
665 Py_DECREF(type); \
666 goto fail; \
667 }
668
669 /* DEFAULT_BUFFER_SIZE */
670 if (PyModule_AddIntMacro(m, DEFAULT_BUFFER_SIZE) < 0)
671 goto fail;
672
673 /* UnsupportedOperation inherits from ValueError and IOError */
674 state->unsupported_operation = PyObject_CallFunction(
675 (PyObject *)&PyType_Type, "s(OO){}",
676 "UnsupportedOperation", PyExc_ValueError, PyExc_IOError);
677 if (state->unsupported_operation == NULL)
678 goto fail;
679 Py_INCREF(state->unsupported_operation);
680 if (PyModule_AddObject(m, "UnsupportedOperation",
681 state->unsupported_operation) < 0)
682 goto fail;
683
Antoine Pitrou6b4883d2011-10-12 02:54:14 +0200684 /* BlockingIOError, for compatibility */
685 Py_INCREF(PyExc_BlockingIOError);
686 if (PyModule_AddObject(m, "BlockingIOError",
687 (PyObject *) PyExc_BlockingIOError) < 0)
688 goto fail;
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000689
690 /* Concrete base types of the IO ABCs.
691 (the ABCs themselves are declared through inheritance in io.py)
692 */
693 ADD_TYPE(&PyIOBase_Type, "_IOBase");
694 ADD_TYPE(&PyRawIOBase_Type, "_RawIOBase");
695 ADD_TYPE(&PyBufferedIOBase_Type, "_BufferedIOBase");
696 ADD_TYPE(&PyTextIOBase_Type, "_TextIOBase");
697
698 /* Implementation of concrete IO objects. */
699 /* FileIO */
700 PyFileIO_Type.tp_base = &PyRawIOBase_Type;
701 ADD_TYPE(&PyFileIO_Type, "FileIO");
702
703 /* BytesIO */
704 PyBytesIO_Type.tp_base = &PyBufferedIOBase_Type;
705 ADD_TYPE(&PyBytesIO_Type, "BytesIO");
Antoine Pitrou972ee132010-09-06 18:48:21 +0000706 if (PyType_Ready(&_PyBytesIOBuffer_Type) < 0)
707 goto fail;
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000708
709 /* StringIO */
710 PyStringIO_Type.tp_base = &PyTextIOBase_Type;
711 ADD_TYPE(&PyStringIO_Type, "StringIO");
712
713 /* BufferedReader */
714 PyBufferedReader_Type.tp_base = &PyBufferedIOBase_Type;
715 ADD_TYPE(&PyBufferedReader_Type, "BufferedReader");
716
717 /* BufferedWriter */
718 PyBufferedWriter_Type.tp_base = &PyBufferedIOBase_Type;
719 ADD_TYPE(&PyBufferedWriter_Type, "BufferedWriter");
720
721 /* BufferedRWPair */
722 PyBufferedRWPair_Type.tp_base = &PyBufferedIOBase_Type;
723 ADD_TYPE(&PyBufferedRWPair_Type, "BufferedRWPair");
724
725 /* BufferedRandom */
726 PyBufferedRandom_Type.tp_base = &PyBufferedIOBase_Type;
727 ADD_TYPE(&PyBufferedRandom_Type, "BufferedRandom");
728
729 /* TextIOWrapper */
730 PyTextIOWrapper_Type.tp_base = &PyTextIOBase_Type;
731 ADD_TYPE(&PyTextIOWrapper_Type, "TextIOWrapper");
732
733 /* IncrementalNewlineDecoder */
734 ADD_TYPE(&PyIncrementalNewlineDecoder_Type, "IncrementalNewlineDecoder");
735
736 /* Interned strings */
Antoine Pitroufc1b6f02012-01-18 16:13:56 +0100737#define ADD_INTERNED(name) \
738 if (!_PyIO_str_ ## name && \
739 !(_PyIO_str_ ## name = PyUnicode_InternFromString(# name))) \
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000740 goto fail;
Antoine Pitroufc1b6f02012-01-18 16:13:56 +0100741
742 ADD_INTERNED(close)
743 ADD_INTERNED(closed)
744 ADD_INTERNED(decode)
745 ADD_INTERNED(encode)
746 ADD_INTERNED(fileno)
747 ADD_INTERNED(flush)
748 ADD_INTERNED(getstate)
749 ADD_INTERNED(isatty)
750 ADD_INTERNED(newlines)
751 ADD_INTERNED(read)
752 ADD_INTERNED(read1)
753 ADD_INTERNED(readable)
Antoine Pitroubb5b92d2012-01-18 16:19:19 +0100754 ADD_INTERNED(readall)
Antoine Pitroufc1b6f02012-01-18 16:13:56 +0100755 ADD_INTERNED(readinto)
756 ADD_INTERNED(readline)
757 ADD_INTERNED(reset)
758 ADD_INTERNED(seek)
759 ADD_INTERNED(seekable)
760 ADD_INTERNED(setstate)
761 ADD_INTERNED(tell)
762 ADD_INTERNED(truncate)
763 ADD_INTERNED(write)
764 ADD_INTERNED(writable)
765
766 if (!_PyIO_str_nl &&
767 !(_PyIO_str_nl = PyUnicode_InternFromString("\n")))
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000768 goto fail;
Antoine Pitroufc1b6f02012-01-18 16:13:56 +0100769
770 if (!_PyIO_empty_str &&
771 !(_PyIO_empty_str = PyUnicode_FromStringAndSize(NULL, 0)))
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000772 goto fail;
Antoine Pitroufc1b6f02012-01-18 16:13:56 +0100773 if (!_PyIO_empty_bytes &&
774 !(_PyIO_empty_bytes = PyBytes_FromStringAndSize(NULL, 0)))
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000775 goto fail;
Antoine Pitroufc1b6f02012-01-18 16:13:56 +0100776 if (!_PyIO_zero &&
777 !(_PyIO_zero = PyLong_FromLong(0L)))
Antoine Pitroue4501852009-05-14 18:55:55 +0000778 goto fail;
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000779
780 state->initialized = 1;
781
782 return m;
783
784 fail:
Benjamin Peterson4fa88fa2009-03-04 00:14:51 +0000785 Py_XDECREF(state->unsupported_operation);
786 Py_DECREF(m);
787 return NULL;
788}