327 lines
13 KiB
C
327 lines
13 KiB
C
/*[clinic input]
|
|
preserve
|
|
[clinic start generated code]*/
|
|
|
|
PyDoc_STRVAR(_io_open__doc__,
|
|
"open($module, /, file, mode=\'r\', buffering=-1, encoding=None,\n"
|
|
" errors=None, newline=None, closefd=True, opener=None)\n"
|
|
"--\n"
|
|
"\n"
|
|
"Open file and return a stream. Raise OSError upon failure.\n"
|
|
"\n"
|
|
"file is either a text or byte string giving the name (and the path\n"
|
|
"if the file isn\'t in the current working directory) of the file to\n"
|
|
"be opened or an integer file descriptor of the file to be\n"
|
|
"wrapped. (If a file descriptor is given, it is closed when the\n"
|
|
"returned I/O object is closed, unless closefd is set to False.)\n"
|
|
"\n"
|
|
"mode is an optional string that specifies the mode in which the file\n"
|
|
"is opened. It defaults to \'r\' which means open for reading in text\n"
|
|
"mode. Other common values are \'w\' for writing (truncating the file if\n"
|
|
"it already exists), \'x\' for creating and writing to a new file, and\n"
|
|
"\'a\' for appending (which on some Unix systems, means that all writes\n"
|
|
"append to the end of the file regardless of the current seek position).\n"
|
|
"In text mode, if encoding is not specified the encoding used is platform\n"
|
|
"dependent: locale.getpreferredencoding(False) is called to get the\n"
|
|
"current locale encoding. (For reading and writing raw bytes use binary\n"
|
|
"mode and leave encoding unspecified.) The available modes are:\n"
|
|
"\n"
|
|
"========= ===============================================================\n"
|
|
"Character Meaning\n"
|
|
"--------- ---------------------------------------------------------------\n"
|
|
"\'r\' open for reading (default)\n"
|
|
"\'w\' open for writing, truncating the file first\n"
|
|
"\'x\' create a new file and open it for writing\n"
|
|
"\'a\' open for writing, appending to the end of the file if it exists\n"
|
|
"\'b\' binary mode\n"
|
|
"\'t\' text mode (default)\n"
|
|
"\'+\' open a disk file for updating (reading and writing)\n"
|
|
"\'U\' universal newline mode (deprecated)\n"
|
|
"========= ===============================================================\n"
|
|
"\n"
|
|
"The default mode is \'rt\' (open for reading text). For binary random\n"
|
|
"access, the mode \'w+b\' opens and truncates the file to 0 bytes, while\n"
|
|
"\'r+b\' opens the file without truncation. The \'x\' mode implies \'w\' and\n"
|
|
"raises an `FileExistsError` if the file already exists.\n"
|
|
"\n"
|
|
"Python distinguishes between files opened in binary and text modes,\n"
|
|
"even when the underlying operating system doesn\'t. Files opened in\n"
|
|
"binary mode (appending \'b\' to the mode argument) return contents as\n"
|
|
"bytes objects without any decoding. In text mode (the default, or when\n"
|
|
"\'t\' is appended to the mode argument), the contents of the file are\n"
|
|
"returned as strings, the bytes having been first decoded using a\n"
|
|
"platform-dependent encoding or using the specified encoding if given.\n"
|
|
"\n"
|
|
"\'U\' mode is deprecated and will raise an exception in future versions\n"
|
|
"of Python. It has no effect in Python 3. Use newline to control\n"
|
|
"universal newlines mode.\n"
|
|
"\n"
|
|
"buffering is an optional integer used to set the buffering policy.\n"
|
|
"Pass 0 to switch buffering off (only allowed in binary mode), 1 to select\n"
|
|
"line buffering (only usable in text mode), and an integer > 1 to indicate\n"
|
|
"the size of a fixed-size chunk buffer. When no buffering argument is\n"
|
|
"given, the default buffering policy works as follows:\n"
|
|
"\n"
|
|
"* Binary files are buffered in fixed-size chunks; the size of the buffer\n"
|
|
" is chosen using a heuristic trying to determine the underlying device\'s\n"
|
|
" \"block size\" and falling back on `io.DEFAULT_BUFFER_SIZE`.\n"
|
|
" On many systems, the buffer will typically be 4096 or 8192 bytes long.\n"
|
|
"\n"
|
|
"* \"Interactive\" text files (files for which isatty() returns True)\n"
|
|
" use line buffering. Other text files use the policy described above\n"
|
|
" for binary files.\n"
|
|
"\n"
|
|
"encoding is the name of the encoding used to decode or encode the\n"
|
|
"file. This should only be used in text mode. The default encoding is\n"
|
|
"platform dependent, but any encoding supported by Python can be\n"
|
|
"passed. See the codecs module for the list of supported encodings.\n"
|
|
"\n"
|
|
"errors is an optional string that specifies how encoding errors are to\n"
|
|
"be handled---this argument should not be used in binary mode. Pass\n"
|
|
"\'strict\' to raise a ValueError exception if there is an encoding error\n"
|
|
"(the default of None has the same effect), or pass \'ignore\' to ignore\n"
|
|
"errors. (Note that ignoring encoding errors can lead to data loss.)\n"
|
|
"See the documentation for codecs.register or run \'help(codecs.Codec)\'\n"
|
|
"for a list of the permitted encoding error strings.\n"
|
|
"\n"
|
|
"newline controls how universal newlines works (it only applies to text\n"
|
|
"mode). It can be None, \'\', \'\\n\', \'\\r\', and \'\\r\\n\'. It works as\n"
|
|
"follows:\n"
|
|
"\n"
|
|
"* On input, if newline is None, universal newlines mode is\n"
|
|
" enabled. Lines in the input can end in \'\\n\', \'\\r\', or \'\\r\\n\', and\n"
|
|
" these are translated into \'\\n\' before being returned to the\n"
|
|
" caller. If it is \'\', universal newline mode is enabled, but line\n"
|
|
" endings are returned to the caller untranslated. If it has any of\n"
|
|
" the other legal values, input lines are only terminated by the given\n"
|
|
" string, and the line ending is returned to the caller untranslated.\n"
|
|
"\n"
|
|
"* On output, if newline is None, any \'\\n\' characters written are\n"
|
|
" translated to the system default line separator, os.linesep. If\n"
|
|
" newline is \'\' or \'\\n\', no translation takes place. If newline is any\n"
|
|
" of the other legal values, any \'\\n\' characters written are translated\n"
|
|
" to the given string.\n"
|
|
"\n"
|
|
"If closefd is False, the underlying file descriptor will be kept open\n"
|
|
"when the file is closed. This does not work when a file name is given\n"
|
|
"and must be True in that case.\n"
|
|
"\n"
|
|
"A custom opener can be used by passing a callable as *opener*. The\n"
|
|
"underlying file descriptor for the file object is then obtained by\n"
|
|
"calling *opener* with (*file*, *flags*). *opener* must return an open\n"
|
|
"file descriptor (passing os.open as *opener* results in functionality\n"
|
|
"similar to passing None).\n"
|
|
"\n"
|
|
"open() returns a file object whose type depends on the mode, and\n"
|
|
"through which the standard file operations such as reading and writing\n"
|
|
"are performed. When open() is used to open a file in a text mode (\'w\',\n"
|
|
"\'r\', \'wt\', \'rt\', etc.), it returns a TextIOWrapper. When used to open\n"
|
|
"a file in a binary mode, the returned class varies: in read binary\n"
|
|
"mode, it returns a BufferedReader; in write binary and append binary\n"
|
|
"modes, it returns a BufferedWriter, and in read/write mode, it returns\n"
|
|
"a BufferedRandom.\n"
|
|
"\n"
|
|
"It is also possible to use a string or bytearray as a file for both\n"
|
|
"reading and writing. For strings StringIO can be used like a file\n"
|
|
"opened in a text mode, and for bytes a BytesIO can be used like a file\n"
|
|
"opened in a binary mode.");
|
|
|
|
#define _IO_OPEN_METHODDEF \
|
|
{"open", (PyCFunction)(void(*)(void))_io_open, METH_FASTCALL|METH_KEYWORDS, _io_open__doc__},
|
|
|
|
static PyObject *
|
|
_io_open_impl(PyObject *module, PyObject *file, const char *mode,
|
|
int buffering, const char *encoding, const char *errors,
|
|
const char *newline, int closefd, PyObject *opener);
|
|
|
|
static PyObject *
|
|
_io_open(PyObject *module, PyObject *const *args, Py_ssize_t nargs, PyObject *kwnames)
|
|
{
|
|
PyObject *return_value = NULL;
|
|
static const char * const _keywords[] = {"file", "mode", "buffering", "encoding", "errors", "newline", "closefd", "opener", NULL};
|
|
static _PyArg_Parser _parser = {NULL, _keywords, "open", 0};
|
|
PyObject *argsbuf[8];
|
|
Py_ssize_t noptargs = nargs + (kwnames ? PyTuple_GET_SIZE(kwnames) : 0) - 1;
|
|
PyObject *file;
|
|
const char *mode = "r";
|
|
int buffering = -1;
|
|
const char *encoding = NULL;
|
|
const char *errors = NULL;
|
|
const char *newline = NULL;
|
|
int closefd = 1;
|
|
PyObject *opener = Py_None;
|
|
|
|
args = _PyArg_UnpackKeywords(args, nargs, NULL, kwnames, &_parser, 1, 8, 0, argsbuf);
|
|
if (!args) {
|
|
goto exit;
|
|
}
|
|
file = args[0];
|
|
if (!noptargs) {
|
|
goto skip_optional_pos;
|
|
}
|
|
if (args[1]) {
|
|
if (!PyUnicode_Check(args[1])) {
|
|
_PyArg_BadArgument("open", 2, "str", args[1]);
|
|
goto exit;
|
|
}
|
|
Py_ssize_t mode_length;
|
|
mode = PyUnicode_AsUTF8AndSize(args[1], &mode_length);
|
|
if (mode == NULL) {
|
|
goto exit;
|
|
}
|
|
if (strlen(mode) != (size_t)mode_length) {
|
|
PyErr_SetString(PyExc_ValueError, "embedded null character");
|
|
goto exit;
|
|
}
|
|
if (!--noptargs) {
|
|
goto skip_optional_pos;
|
|
}
|
|
}
|
|
if (args[2]) {
|
|
if (PyFloat_Check(args[2])) {
|
|
PyErr_SetString(PyExc_TypeError,
|
|
"integer argument expected, got float" );
|
|
goto exit;
|
|
}
|
|
buffering = _PyLong_AsInt(args[2]);
|
|
if (buffering == -1 && PyErr_Occurred()) {
|
|
goto exit;
|
|
}
|
|
if (!--noptargs) {
|
|
goto skip_optional_pos;
|
|
}
|
|
}
|
|
if (args[3]) {
|
|
if (args[3] == Py_None) {
|
|
encoding = NULL;
|
|
}
|
|
else if (PyUnicode_Check(args[3])) {
|
|
Py_ssize_t encoding_length;
|
|
encoding = PyUnicode_AsUTF8AndSize(args[3], &encoding_length);
|
|
if (encoding == NULL) {
|
|
goto exit;
|
|
}
|
|
if (strlen(encoding) != (size_t)encoding_length) {
|
|
PyErr_SetString(PyExc_ValueError, "embedded null character");
|
|
goto exit;
|
|
}
|
|
}
|
|
else {
|
|
_PyArg_BadArgument("open", 4, "str or None", args[3]);
|
|
goto exit;
|
|
}
|
|
if (!--noptargs) {
|
|
goto skip_optional_pos;
|
|
}
|
|
}
|
|
if (args[4]) {
|
|
if (args[4] == Py_None) {
|
|
errors = NULL;
|
|
}
|
|
else if (PyUnicode_Check(args[4])) {
|
|
Py_ssize_t errors_length;
|
|
errors = PyUnicode_AsUTF8AndSize(args[4], &errors_length);
|
|
if (errors == NULL) {
|
|
goto exit;
|
|
}
|
|
if (strlen(errors) != (size_t)errors_length) {
|
|
PyErr_SetString(PyExc_ValueError, "embedded null character");
|
|
goto exit;
|
|
}
|
|
}
|
|
else {
|
|
_PyArg_BadArgument("open", 5, "str or None", args[4]);
|
|
goto exit;
|
|
}
|
|
if (!--noptargs) {
|
|
goto skip_optional_pos;
|
|
}
|
|
}
|
|
if (args[5]) {
|
|
if (args[5] == Py_None) {
|
|
newline = NULL;
|
|
}
|
|
else if (PyUnicode_Check(args[5])) {
|
|
Py_ssize_t newline_length;
|
|
newline = PyUnicode_AsUTF8AndSize(args[5], &newline_length);
|
|
if (newline == NULL) {
|
|
goto exit;
|
|
}
|
|
if (strlen(newline) != (size_t)newline_length) {
|
|
PyErr_SetString(PyExc_ValueError, "embedded null character");
|
|
goto exit;
|
|
}
|
|
}
|
|
else {
|
|
_PyArg_BadArgument("open", 6, "str or None", args[5]);
|
|
goto exit;
|
|
}
|
|
if (!--noptargs) {
|
|
goto skip_optional_pos;
|
|
}
|
|
}
|
|
if (args[6]) {
|
|
if (PyFloat_Check(args[6])) {
|
|
PyErr_SetString(PyExc_TypeError,
|
|
"integer argument expected, got float" );
|
|
goto exit;
|
|
}
|
|
closefd = _PyLong_AsInt(args[6]);
|
|
if (closefd == -1 && PyErr_Occurred()) {
|
|
goto exit;
|
|
}
|
|
if (!--noptargs) {
|
|
goto skip_optional_pos;
|
|
}
|
|
}
|
|
opener = args[7];
|
|
skip_optional_pos:
|
|
return_value = _io_open_impl(module, file, mode, buffering, encoding, errors, newline, closefd, opener);
|
|
|
|
exit:
|
|
return return_value;
|
|
}
|
|
|
|
PyDoc_STRVAR(_io_open_code__doc__,
|
|
"open_code($module, /, path)\n"
|
|
"--\n"
|
|
"\n"
|
|
"Opens the provided file with the intent to import the contents.\n"
|
|
"\n"
|
|
"This may perform extra validation beyond open(), but is otherwise interchangeable\n"
|
|
"with calling open(path, \'rb\').");
|
|
|
|
#define _IO_OPEN_CODE_METHODDEF \
|
|
{"open_code", (PyCFunction)(void(*)(void))_io_open_code, METH_FASTCALL|METH_KEYWORDS, _io_open_code__doc__},
|
|
|
|
static PyObject *
|
|
_io_open_code_impl(PyObject *module, PyObject *path);
|
|
|
|
static PyObject *
|
|
_io_open_code(PyObject *module, PyObject *const *args, Py_ssize_t nargs, PyObject *kwnames)
|
|
{
|
|
PyObject *return_value = NULL;
|
|
static const char * const _keywords[] = {"path", NULL};
|
|
static _PyArg_Parser _parser = {NULL, _keywords, "open_code", 0};
|
|
PyObject *argsbuf[1];
|
|
PyObject *path;
|
|
|
|
args = _PyArg_UnpackKeywords(args, nargs, NULL, kwnames, &_parser, 1, 1, 0, argsbuf);
|
|
if (!args) {
|
|
goto exit;
|
|
}
|
|
if (!PyUnicode_Check(args[0])) {
|
|
_PyArg_BadArgument("open_code", 1, "str", args[0]);
|
|
goto exit;
|
|
}
|
|
if (PyUnicode_READY(args[0]) == -1) {
|
|
goto exit;
|
|
}
|
|
path = args[0];
|
|
return_value = _io_open_code_impl(module, path);
|
|
|
|
exit:
|
|
return return_value;
|
|
}
|
|
/*[clinic end generated code: output=d479285078750d68 input=a9049054013a1b77]*/
|