static PyObject * py_detect (PyObject * self, PyObject * args) { // {{{
PyObject * err = NULL;
char * text;
size_t inlen;
int argc;
DetectObj * obj;
PyObject * dict;
PyObject * prop;
if ( ! PyArg_ParseTuple (args, "s#|O", &text, &inlen, &err) )
return NULL;
argc = PyTuple_Size (args);
if ( err != NULL ) {
if ( ! PyList_Check (err) ) {
PyErr_SetString (ErrorObject, "3th argument is must PyLis");
return NULL;
}
}
if ( (obj = detect_obj_init ()) == NULL ) {
if ( argc > 1 ) {
PyObject * value = PyString_FromString ("Memory allocation failed");
PyList_Append (err, value);
Py_DECREF (value);
}
return Py_None;
}
#ifdef CHARDET_BINARY_SAFE
if ( detect_r (text, inlen, &obj) == CHARDET_OUT_OF_MEMORY )
#else
if ( detect (text, &obj) == CHARDET_OUT_OF_MEMORY )
#endif
{
if ( argc > 1 ) {
PyObject * value = PyString_FromString ("On handle processing, occured out of memory");
PyList_Append (err, value);
Py_DECREF (value);
}
detect_obj_free (&obj);
return Py_None;
}
dict = PyDict_New ();
prop = Py_BuildValue ("s", obj->encoding);
PyDict_SetItemString (dict, "encoding", prop);
Py_DECREF (prop);
prop = Py_BuildValue ("f", obj->confidence);
PyDict_SetItemString (dict, "confidence", prop);
Py_DECREF (prop);
detect_obj_free (&obj);
return dict;
} // }}}
int main (void) {
DetectObj *obj;
int i;
char *str[] = {
"안녕",
"안녕하세요",
"안녕하세요 정말?",
"그래 이정도면 판단 될까?",
"좀더 길게 적어 볼까 얼마나 길게 해야!",
"그래 그래 좀 더 길게 적어 보자 더 길게 적어 보야야 겠지...",
"12345 abcde"
};
short arrayNum;
arrayNum = sizeof (str) / sizeof (str[0]);
for ( i=0; i<arrayNum; i++ ) {
obj = detect_obj_init ();
if ( obj == NULL ) {
fprintf (stderr, "On attemped \"%s\", memory allocation failed\n", str[i]);
continue;
}
#ifdef CHARDET_BINARY_SAFE
if ( detect_r (str[i], strlen (str[i]), &obj) == CHARDET_OUT_OF_MEMORY )
#else
if ( detect (str[i], &obj) == CHARDET_OUT_OF_MEMORY )
#endif
{
fprintf (stderr, "On handle processing, occured out of memory\n");
return CHARDET_OUT_OF_MEMORY;
}
printf ("## %s : %s : %f\n", str[i], obj->encoding, obj->confidence);
detect_obj_free (&obj);
}
return 0;
}
CharsetDetector::CharsetDetector(const QByteArray &data)
: d(new Data) {
d->obj = detect_obj_init();
d->detected = (::detect(data.data(), &d->obj) == CHARDET_SUCCESS);
}