#include <Python.h>


/* Function from Parser/tokenizer.c */
extern char * PyTokenizer_FindEncodingFilename(int, PyObject *);


static PyObject *
tokenizer_detect_encoding(PyObject *self, PyObject *args)
{
    char *encoding;
    int fd;
    FILE *fp;
    PyObject *filename;

    if (!PyArg_ParseTuple(args, "U", &filename))
        return NULL;

    fp = _Py_fopen(filename, "r");
    if (fp == NULL) {
        if (!PyErr_Occurred())
            PyErr_SetFromErrno(PyExc_IOError);
        return NULL;
    }
    fd = fileno(fp);
    if (fd != -1)
        fd = dup(fd);
    fclose(fp);
    fp = NULL;
    if (fd == -1)
        return PyErr_SetFromErrno(PyExc_OSError);

    encoding = PyTokenizer_FindEncodingFilename(fd, filename);
    close(fd);
    if (encoding == NULL && PyErr_Occurred())
        return NULL;

    encoding = (encoding != NULL) ? encoding :
           (char*)PyUnicode_GetDefaultEncoding();
    if (encoding == NULL)
        return NULL;
    
    return PyUnicode_FromString(encoding);
};


static PyMethodDef TokenizerMethods[] = {
    {"detect_encoding",  tokenizer_detect_encoding, METH_VARARGS,
     "detect a file's encoding."},
    {NULL, NULL, 0, NULL}        /* Sentinel */
};


static struct PyModuleDef tokenizermodule = {
   PyModuleDef_HEAD_INIT,
   "_tokenizer",   /* name of module */
   NULL, /* module documentation, may be NULL */
   -1,       /* size of per-interpreter state of the module,
                or -1 if the module keeps state in global variables. */
   TokenizerMethods
};


PyMODINIT_FUNC
PyInit__tokenizer(void)
{
    return PyModule_Create(&tokenizermodule);
}