/* ------------------------------------------------------------------------

   Python Codec Registry and support functions

Written by Marc-Andre Lemburg (mal@lemburg.com).

Copyright (c) Corporation for National Research Initiatives.

   ------------------------------------------------------------------------ */

#include "Python.h"

#include <ctype.h>

/* --- Codec Registry ----------------------------------------------------- */

/* Import the standard encodings package which will register the first

   codec search function. 

   This is done in a lazy way so that the Unicode implementation does

   not downgrade startup time of scripts not needing it.

   ImportErrors are silently ignored by this function. Only one try is

   made.

*/

static int _PyCodecRegistry_Init(void); /* Forward */

int PyCodec_Register(PyObject *search_function)

{

    PyInterpreterState *interp = PyThreadState_GET()->interp;

    if (interp->codec_search_path == NULL && _PyCodecRegistry_Init())

	goto onError;

    if (search_function == NULL) {

	PyErr_BadArgument();

	goto onError;

    }

    if (!PyCallable_Check(search_function)) {

	PyErr_SetString(PyExc_TypeError, "argument must be callable");

	goto onError;

    }

    return PyList_Append(interp->codec_search_path, search_function);

 onError:

    return -1;

}

/* Convert a string to a normalized Python string: all characters are

   converted to lower case, spaces are replaced with underscores. */

static

PyObject *normalizestring(const char *string)

{

    register size_t i;

    size_t len = strlen(string);

    char *p;

    PyObject *v;

    if (len > PY_SSIZE_T_MAX) {

	PyErr_SetString(PyExc_OverflowError, "string is too large");

	return NULL;

    }

    v = PyString_FromStringAndSize(NULL, len);

    if (v == NULL)

	return NULL;

    p = PyString_AS_STRING(v);

    for (i = 0; i < len; i++) {

        register char ch = string[i];

        if (ch == ' ')

            ch = '-';

        else

            ch = tolower(Py_CHARMASK(ch));

	p[i] = ch;

    }

    return v;

}

/* Lookup the given encoding and return a tuple providing the codec

   facilities.

   The encoding string is looked up converted to all lower-case

   characters. This makes encodings looked up through this mechanism

   effectively case-insensitive.

   If no codec is found, a LookupError is set and NULL returned. 

   As side effect, this tries to load the encodings package, if not

   yet done. This is part of the lazy load strategy for the encodings

   package.

*/

PyObject *_PyCodec_Lookup(const char *encoding)

{

    PyInterpreterState *interp;

    PyObject *result, *args = NULL, *v;

    Py_ssize_t i, len;

    if (encoding == NULL) {

	PyErr_BadArgument();

	goto onError;

    }

    interp = PyThreadState_GET()->interp;

    if (interp->codec_search_path == NULL && _PyCodecRegistry_Init())

	goto onError;

    /* Convert the encoding to a normalized Python string: all

       characters are converted to lower case, spaces and hyphens are

       replaced with underscores. */

    v = normalizestring(encoding);

    if (v == NULL)

	goto onError;

    PyString_InternInPlace(&v);

    /* First, try to lookup the name in the registry dictionary */

    result = PyDict_GetItem(interp->codec_search_cache, v);

    if (result != NULL) {

	Py_INCREF(result);

	Py_DECREF(v);

	return result;

    }

    /* Next, scan the search functions in order of registration */

    args = PyTuple_New(1);

    if (args == NULL)

	goto onError;

    PyTuple_SET_ITEM(args,0,v);

    len = PyList_Size(interp->codec_search_path);

    if (len < 0)

	goto onError;

    if (len == 0) {

	PyErr_SetString(PyExc_LookupError,

			"no codec search functions registered: "

			"can't find encoding");

	goto onError;

    }

    for (i = 0; i < len; i++) {

	PyObject *func;

	func = PyList_GetItem(interp->codec_search_path, i);

	if (func == NULL)

	    goto onError;

	result = PyEval_CallObject(func, args);

	if (result == NULL)

	    goto onError;

	if (result == Py_None) {

	    Py_DECREF(result);

	    continue;

	}

	if (!PyTuple_Check(result) || PyTuple_GET_SIZE(result) != 4) {

	    PyErr_SetString(PyExc_TypeError,

			    "codec search functions must return 4-tuples");

	    Py_DECREF(result);

	    goto onError;

	}

	break;

    }

    if (i == len) {

	/* XXX Perhaps we should cache misses too ? */

	PyErr_Format(PyExc_LookupError,

                     "unknown encoding: %s", encoding);

	goto onError;

    }

    /* Cache and return the result */

    PyDict_SetItem(interp->codec_search_cache, v, result);

    Py_DECREF(args);

    return result;

 onError:

    Py_XDECREF(args);

    return NULL;

}

static

PyObject *args_tuple(PyObject *object,

		     const char *errors)

{

    PyObject *args;

    args = PyTuple_New(1 + (errors != NULL));

    if (args == NULL)

	return NULL;

    Py_INCREF(object);

    PyTuple_SET_ITEM(args,0,object);

    if (errors) {

	PyObject *v;

	v = PyString_FromString(errors);

	if (v == NULL) {

	    Py_DECREF(args);

	    return NULL;

	}

	PyTuple_SET_ITEM(args, 1, v);

    }

    return args;

}

/* Helper function to get a codec item */

static

PyObject *codec_getitem(const char *encoding, int index)

{

    PyObject *codecs;

    PyObject *v;

    codecs = _PyCodec_Lookup(encoding);

    if (codecs == NULL)

	return NULL;

    v = PyTuple_GET_ITEM(codecs, index);

    Py_DECREF(codecs);

    Py_INCREF(v);

    return v;

}

/* Helper function to create an incremental codec. */

static

PyObject *codec_getincrementalcodec(const char *encoding,

				    const char *errors,

				    const char *attrname)

{

    PyObject *codecs, *ret, *inccodec;

    codecs = _PyCodec_Lookup(encoding);

    if (codecs == NULL)

	return NULL;

    inccodec = PyObject_GetAttrString(codecs, attrname);

    Py_DECREF(codecs);

    if (inccodec == NULL)

	return NULL;

    if (errors)

	ret = PyObject_CallFunction(inccodec, "s", errors);

    else

	ret = PyObject_CallFunction(inccodec, NULL);

    Py_DECREF(inccodec);

    return ret;

}

/* Helper function to create a stream codec. */

static

PyObject *codec_getstreamcodec(const char *encoding,

			       PyObject *stream,

			       const char *errors,

			       const int index)

{

    PyObject *codecs, *streamcodec, *codeccls;

    codecs = _PyCodec_Lookup(encoding);

    if (codecs == NULL)

	return NULL;

    codeccls = PyTuple_GET_ITEM(codecs, index);

    if (errors != NULL)

	streamcodec = PyObject_CallFunction(codeccls, "Os", stream, errors);

    else

	streamcodec = PyObject_CallFunction(codeccls, "O", stream);

    Py_DECREF(codecs);

    return streamcodec;

}

/* Convenience APIs to query the Codec registry. 

   All APIs return a codec object with incremented refcount.

 */

PyObject *PyCodec_Encoder(const char *encoding)

{

    return codec_getitem(encoding, 0);

}

PyObject *PyCodec_Decoder(const char *encoding)

{

    return codec_getitem(encoding, 1);

}

PyObject *PyCodec_IncrementalEncoder(const char *encoding,

				     const char *errors)

{

    return codec_getincrementalcodec(encoding, errors, "incrementalencoder");

}

PyObject *PyCodec_IncrementalDecoder(const char *encoding,

				     const char *errors)

{

    return codec_getincrementalcodec(encoding, errors, "incrementaldecoder");

}

PyObject *PyCodec_StreamReader(const char *encoding,

			       PyObject *stream,

			       const char *errors)

{

    return codec_getstreamcodec(encoding, stream, errors, 2);

}

PyObject *PyCodec_StreamWriter(const char *encoding,

			       PyObject *stream,

			       const char *errors)

{

    return codec_getstreamcodec(encoding, stream, errors, 3);

}

/* Encode an object (e.g. an Unicode object) using the given encoding

   and return the resulting encoded object (usually a Python string).

   errors is passed to the encoder factory as argument if non-NULL. */

PyObject *PyCodec_Encode(PyObject *object,

			 const char *encoding,

			 const char *errors)

{

    PyObject *encoder = NULL;

    PyObject *args = NULL, *result = NULL;

    PyObject *v;

    encoder = PyCodec_Encoder(encoding);

    if (encoder == NULL)

	goto onError;

    args = args_tuple(object, errors);

    if (args == NULL)

	goto onError;

    result = PyEval_CallObject(encoder,args);

    if (result == NULL)

	goto onError;

    if (!PyTuple_Check(result) || 

	PyTuple_GET_SIZE(result) != 2) {

	PyErr_SetString(PyExc_TypeError,

			"encoder must return a tuple (object,integer)");

	goto onError;

    }

    v = PyTuple_GET_ITEM(result,0);

    Py_INCREF(v);

    /* We don't check or use the second (integer) entry. */

    Py_DECREF(args);

    Py_DECREF(encoder);

    Py_DECREF(result);

    return v;

 onError:

    Py_XDECREF(result);

    Py_XDECREF(args);

    Py_XDECREF(encoder);

    return NULL;

}

/* Decode an object (usually a Python string) using the given encoding

   and return an equivalent object (e.g. an Unicode object).

   errors is passed to the decoder factory as argument if non-NULL. */

PyObject *PyCodec_Decode(PyObject *object,

			 const char *encoding,

			 const char *errors)

{

    PyObject *decoder = NULL;

    PyObject *args = NULL, *result = NULL;

    PyObject *v;

    decoder = PyCodec_Decoder(encoding);

    if (decoder == NULL)

	goto onError;

    args = args_tuple(object, errors);

    if (args == NULL)

	goto onError;

    result = PyEval_CallObject(decoder,args);

    if (result == NULL)

	goto onError;

    if (!PyTuple_Check(result) || 

	PyTuple_GET_SIZE(result) != 2) {

	PyErr_SetString(PyExc_TypeError,

			"decoder must return a tuple (object,integer)");

	goto onError;

    }

    v = PyTuple_GET_ITEM(result,0);

    Py_INCREF(v);

    /* We don't check or use the second (integer) entry. */

    Py_DECREF(args);

    Py_DECREF(decoder);

    Py_DECREF(result);

    return v;

 onError:

    Py_XDECREF(args);

    Py_XDECREF(decoder);

    Py_XDECREF(result);

    return NULL;

}

/* Register the error handling callback function error under the name

   name. This function will be called by the codec when it encounters

   an unencodable characters/undecodable bytes and doesn't know the

   callback name, when name is specified as the error parameter

   in the call to the encode/decode function.

   Return 0 on success, -1 on error */

int PyCodec_RegisterError(const char *name, PyObject *error)

{

    PyInterpreterState *interp = PyThreadState_GET()->interp;

    if (interp->codec_search_path == NULL && _PyCodecRegistry_Init())

	return -1;

    if (!PyCallable_Check(error)) {

	PyErr_SetString(PyExc_TypeError, "handler must be callable");

	return -1;

    }

    return PyDict_SetItemString(interp->codec_error_registry,

	    			(char *)name, error);

}

/* Lookup the error handling callback function registered under the

   name error. As a special case NULL can be passed, in which case

   the error handling callback for strict encoding will be returned. */

PyObject *PyCodec_LookupError(const char *name)

{

    PyObject *handler = NULL;

    PyInterpreterState *interp = PyThreadState_GET()->interp;

    if (interp->codec_search_path == NULL && _PyCodecRegistry_Init())

	return NULL;

    if (name==NULL)

	name = "strict";

    handler = PyDict_GetItemString(interp->codec_error_registry, (char *)name);

    if (!handler)

	PyErr_Format(PyExc_LookupError, "unknown error handler name '%.400s'", name);

    else

	Py_INCREF(handler);

    return handler;

}

static void wrong_exception_type(PyObject *exc)

{

    PyObject *type = PyObject_GetAttrString(exc, "__class__");

    if (type != NULL) {

	PyObject *name = PyObject_GetAttrString(type, "__name__");

	Py_DECREF(type);

	if (name != NULL) {

	    PyObject *string = PyObject_Str(name);

	    Py_DECREF(name);

	    if (string != NULL) {

	        PyErr_Format(PyExc_TypeError,

		    "don't know how to handle %.400s in error callback",

		    PyString_AS_STRING(string));

	        Py_DECREF(string);

	    }

	}

    }

}

PyObject *PyCodec_StrictErrors(PyObject *exc)

{

    if (PyExceptionInstance_Check(exc))

        PyErr_SetObject(PyExceptionInstance_Class(exc), exc);

    else

	PyErr_SetString(PyExc_TypeError, "codec must pass exception instance");

    return NULL;

}

#ifdef Py_USING_UNICODE

PyObject *PyCodec_IgnoreErrors(PyObject *exc)

{

    Py_ssize_t end;

    if (PyObject_IsInstance(exc, PyExc_UnicodeEncodeError)) {

	if (PyUnicodeEncodeError_GetEnd(exc, &end))

	    return NULL;

    }

    else if (PyObject_IsInstance(exc, PyExc_UnicodeDecodeError)) {

	if (PyUnicodeDecodeError_GetEnd(exc, &end))

	    return NULL;

    }

    else if (PyObject_IsInstance(exc, PyExc_UnicodeTranslateError)) {

	if (PyUnicodeTranslateError_GetEnd(exc, &end))

	    return NULL;

    }

    else {

	wrong_exception_type(exc);

	return NULL;

    }

    /* ouch: passing NULL, 0, pos gives None instead of u'' */

    return Py_BuildValue("(u#n)", &end, 0, end);

}

PyObject *PyCodec_ReplaceErrors(PyObject *exc)

{

    PyObject *restuple;

    Py_ssize_t start;

    Py_ssize_t end;

    Py_ssize_t i;

    if (PyObject_IsInstance(exc, PyExc_UnicodeEncodeError)) {

	PyObject *res;

	Py_UNICODE *p;

	if (PyUnicodeEncodeError_GetStart(exc, &start))

	    return NULL;

	if (PyUnicodeEncodeError_GetEnd(exc, &end))

	    return NULL;

	res = PyUnicode_FromUnicode(NULL, end-start);

	if (res == NULL)

	    return NULL;

	for (p = PyUnicode_AS_UNICODE(res), i = start;

	    i<end; ++p, ++i)

	    *p = '?';

	restuple = Py_BuildValue("(On)", res, end);

	Py_DECREF(res);

	return restuple;

    }

    else if (PyObject_IsInstance(exc, PyExc_UnicodeDecodeError)) {

	Py_UNICODE res = Py_UNICODE_REPLACEMENT_CHARACTER;

	if (PyUnicodeDecodeError_GetEnd(exc, &end))

	    return NULL;

	return Py_BuildValue("(u#n)", &res, 1, end);

    }

    else if (PyObject_IsInstance(exc, PyExc_UnicodeTranslateError)) {

	PyObject *res;

	Py_UNICODE *p;