PEP 293 implemention (from SF patch http://www.python.org/sf/432401)

Doc/lib/libcodecs.tex

@@ -17,7 +17,7 @@
 
 This module defines base classes for standard Python codecs (encoders
 and decoders) and provides access to the internal Python codec
-registry which manages the codec lookup process.
+registry which manages the codec and error handling lookup process.
 
 It defines the following functions:
 
@@ -98,6 +98,43 @@ Raises a \exception{LookupError} in case the encoding cannot be found.
 To simplify working with encoded files or stream, the module
 also defines these utility functions:
 
+\begin{funcdesc}{register_error}{name, error_handler}
+Register the error handling function \var{error_handler} under the
+name \var{name}. \vari{error_handler} will be called during encoding
+and decoding in case of an error, when \var{name} is specified as the
+errors parameter. \var{error_handler} will be called with an
+\exception{UnicodeEncodeError}, \exception{UnicodeDecodeError} or
+\exception{UnicodeTranslateError} instance and must return a tuple
+with a replacement for the unencodable/undecodable part of the input
+and a position where encoding/decoding should continue.
+\end{funcdesc}
+
+\begin{funcdesc}{lookup_error}{name}
+Return the error handler previously register under the name \var{name}.
+
+Raises a \exception{LookupError} in case the handler cannot be found.
+\end{funcdesc}
+
+\begin{funcdesc}{strict_errors}{exception}
+Implements the \code{strict} error handling.
+\end{funcdesc}
+
+\begin{funcdesc}{replace_errors}{exception}
+Implements the \code{replace} error handling.
+\end{funcdesc}
+
+\begin{funcdesc}{ignore_errors}{exception}
+Implements the \code{ignore} error handling.
+\end{funcdesc}
+
+\begin{funcdesc}{xmlcharrefreplace_errors_errors}{exception}
+Implements the \code{xmlcharrefreplace} error handling.
+\end{funcdesc}
+
+\begin{funcdesc}{backslashreplace_errors_errors}{exception}
+Implements the \code{backslashreplace} error handling.
+\end{funcdesc}
+
 \begin{funcdesc}{open}{filename, mode\optional{, encoding\optional{,
                        errors\optional{, buffering}}}}
 Open an encoded file using the given \var{mode} and return

Doc/lib/libexcs.tex

@@ -335,6 +335,24 @@ Raised when an \keyword{assert} statement fails.
 \versionadded{2.0}
 \end{excdesc}
 
+\begin{excdesc}{UnicodeEncodeError}
+  Raised when a Unicode-related error occurs during encoding.  It
+  is a subclass of \exception{UnicodeError}.
+\versionadded{2.3}
+\end{excdesc}
+
+\begin{excdesc}{UnicodeDecodeError}
+  Raised when a Unicode-related error occurs during decoding.  It
+  is a subclass of \exception{UnicodeError}.
+\versionadded{2.3}
+\end{excdesc}
+
+\begin{excdesc}{UnicodeTranslateError}
+  Raised when a Unicode-related error occurs during translating.  It
+  is a subclass of \exception{UnicodeError}.
+\versionadded{2.3}
+\end{excdesc}
+
 \begin{excdesc}{ValueError}
   Raised when a built-in operation or function receives an argument
   that has the right type but an inappropriate value, and the
@@ -426,6 +444,9 @@ The class hierarchy for built-in exceptions is:
      |    |    +-- FloatingPointError
      |    +-- ValueError
      |    |    +-- UnicodeError
+     |    |        +-- UnicodeEncodeError
+     |    |        +-- UnicodeDecodeError
+     |    |        +-- UnicodeTranslateError
      |    +-- ReferenceError
      |    +-- SystemError
      |    +-- MemoryError

Include/codecs.h

@@ -117,6 +117,36 @@ PyAPI_FUNC(PyObject *) PyCodec_StreamWriter(
        const char *errors
        );
 
+/* Unicode encoding error handling callback registry API */
+
+/* Register the error handling callback function error under the name
+   name. This function will be called by the codec when it encounters
+   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 */
+PyAPI_FUNC(int) PyCodec_RegisterError(const char *name, PyObject *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" will be returned. */
+PyAPI_FUNC(PyObject *) PyCodec_LookupError(const char *name);
+
+/* raise exc as an exception */
+PyAPI_FUNC(PyObject *) PyCodec_StrictErrors(PyObject *exc);
+
+/* ignore the unicode error, skipping the faulty input */
+PyAPI_FUNC(PyObject *) PyCodec_IgnoreErrors(PyObject *exc);
+
+/* replace the unicode error with ? or U+FFFD */
+PyAPI_FUNC(PyObject *) PyCodec_ReplaceErrors(PyObject *exc);
+
+/* replace the unicode encode error with XML character references */
+PyAPI_FUNC(PyObject *) PyCodec_XMLCharRefReplaceErrors(PyObject *exc);
+
+/* replace the unicode encode error with backslash escapes (\x, \u and \U) */
+PyAPI_FUNC(PyObject *) PyCodec_BackslashReplaceErrors(PyObject *exc);
+
 #ifdef __cplusplus
 }
 #endif

Include/pyerrors.h

@@ -54,6 +54,9 @@ PyAPI_DATA(PyObject *) PyExc_SystemExit;
 PyAPI_DATA(PyObject *) PyExc_TypeError;
 PyAPI_DATA(PyObject *) PyExc_UnboundLocalError;
 PyAPI_DATA(PyObject *) PyExc_UnicodeError;
+PyAPI_DATA(PyObject *) PyExc_UnicodeEncodeError;
+PyAPI_DATA(PyObject *) PyExc_UnicodeDecodeError;
+PyAPI_DATA(PyObject *) PyExc_UnicodeTranslateError;
 PyAPI_DATA(PyObject *) PyExc_ValueError;
 PyAPI_DATA(PyObject *) PyExc_ZeroDivisionError;
 #ifdef MS_WINDOWS
@@ -114,6 +117,69 @@ PyAPI_FUNC(void) PyErr_SetInterrupt(void);
 PyAPI_FUNC(void) PyErr_SyntaxLocation(char *, int);
 PyAPI_FUNC(PyObject *) PyErr_ProgramText(char *, int);
 
+/* The following functions are used to create and modify unicode
+   exceptions from C */
+/* create a UnicodeDecodeError object */
+PyAPI_FUNC(PyObject *) PyUnicodeDecodeError_Create(
+	const char *, const char *, int, int, int, const char *);
+
+/* create a UnicodeEncodeError object */
+PyAPI_FUNC(PyObject *) PyUnicodeEncodeError_Create(
+	const char *, const Py_UNICODE *, int, int, int, const char *);
+
+/* create a UnicodeTranslateError object */
+PyAPI_FUNC(PyObject *) PyUnicodeTranslateError_Create(
+	const Py_UNICODE *, int, int, int, const char *);
+
+/* get the encoding attribute */
+PyAPI_FUNC(PyObject *) PyUnicodeEncodeError_GetEncoding(PyObject *);
+PyAPI_FUNC(PyObject *) PyUnicodeDecodeError_GetEncoding(PyObject *);
+PyAPI_FUNC(PyObject *) PyUnicodeTranslateError_GetEncoding(PyObject *);
+
+/* get the object attribute */
+PyAPI_FUNC(PyObject *) PyUnicodeEncodeError_GetObject(PyObject *);
+PyAPI_FUNC(PyObject *) PyUnicodeDecodeError_GetObject(PyObject *);
+PyAPI_FUNC(PyObject *) PyUnicodeTranslateError_GetObject(PyObject *);
+
+/* get the value of the start attribute (the int * may not be NULL)
+   return 0 on success, -1 on failure */
+PyAPI_FUNC(int) PyUnicodeEncodeError_GetStart(PyObject *, int *);
+PyAPI_FUNC(int) PyUnicodeDecodeError_GetStart(PyObject *, int *);
+PyAPI_FUNC(int) PyUnicodeTranslateError_GetStart(PyObject *, int *);
+
+/* assign a new value to the start attribute
+   return 0 on success, -1 on failure */
+PyAPI_FUNC(int) PyUnicodeEncodeError_SetStart(PyObject *, int);
+PyAPI_FUNC(int) PyUnicodeDecodeError_SetStart(PyObject *, int);
+PyAPI_FUNC(int) PyUnicodeTranslateError_SetStart(PyObject *, int);
+
+/* get the value of the end attribute (the int *may not be NULL)
+ return 0 on success, -1 on failure */
+PyAPI_FUNC(int) PyUnicodeEncodeError_GetEnd(PyObject *, int *);
+PyAPI_FUNC(int) PyUnicodeDecodeError_GetEnd(PyObject *, int *);
+PyAPI_FUNC(int) PyUnicodeTranslateError_GetEnd(PyObject *, int *);
+
+/* assign a new value to the end attribute
+   return 0 on success, -1 on failure */
+PyAPI_FUNC(int) PyUnicodeEncodeError_SetEnd(PyObject *, int);
+PyAPI_FUNC(int) PyUnicodeDecodeError_SetEnd(PyObject *, int);
+PyAPI_FUNC(int) PyUnicodeTranslateError_SetEnd(PyObject *, int);
+
+/* get the value of the reason attribute */
+PyAPI_FUNC(PyObject *) PyUnicodeEncodeError_GetReason(PyObject *);
+PyAPI_FUNC(PyObject *) PyUnicodeDecodeError_GetReason(PyObject *);
+PyAPI_FUNC(PyObject *) PyUnicodeTranslateError_GetReason(PyObject *);
+
+/* assign a new value to the reason attribute
+   return 0 on success, -1 on failure */
+PyAPI_FUNC(int) PyUnicodeEncodeError_SetReason(
+	PyObject *, const char *);
+PyAPI_FUNC(int) PyUnicodeDecodeError_SetReason(
+	PyObject *, const char *);
+PyAPI_FUNC(int) PyUnicodeTranslateError_SetReason(
+	PyObject *, const char *);
+
+
 /* These APIs aren't really part of the error implementation, but
    often needed to format error messages; the native C lib APIs are
    not available on all platforms, which is why we provide emulations

Lib/codecs.py

@@ -20,7 +20,10 @@ except ImportError, why:
 __all__ = ["register", "lookup", "open", "EncodedFile", "BOM", "BOM_BE",
            "BOM_LE", "BOM32_BE", "BOM32_LE", "BOM64_BE", "BOM64_LE",
            "BOM_UTF8", "BOM_UTF16", "BOM_UTF16_LE", "BOM_UTF16_BE",
-           "BOM_UTF32", "BOM_UTF32_LE", "BOM_UTF32_BE"]
+           "BOM_UTF32", "BOM_UTF32_LE", "BOM_UTF32_BE",
+           "strict_errors", "ignore_errors", "replace_errors",
+           "xmlcharrefreplace_errors",
+           "register_error", "lookup_error"]
 
 ### Constants
 
@@ -632,6 +635,14 @@ def make_encoding_map(decoding_map):
             m[v] = None
     return m
 
+### error handlers
+
+strict_errors = lookup_error("strict")
+ignore_errors = lookup_error("ignore")
+replace_errors = lookup_error("replace")
+xmlcharrefreplace_errors = lookup_error("xmlcharrefreplace")
+backslashreplace_errors = lookup_error("backslashreplace")
+
 # Tell modulefinder that using codecs probably needs the encodings
 # package
 _false = 0

Misc/NEWS

@@ -57,6 +57,9 @@ Type/class unification and new-style classes
 
 Core and builtins
 
+- Codec error handling callbacks (PEP 293) are implemented.
+  Error handling in unicode.encode or str.decode can now be customized.
+
 - A subtle change to the semantics of the built-in function intern():
   interned strings are no longer immortal.  You must keep a reference
   to the return value intern() around to get the benefit.

Modules/_codecsmodule.c

@@ -706,6 +706,32 @@ mbcs_encode(PyObject *self,
 #endif /* MS_WINDOWS */
 #endif /* Py_USING_UNICODE */
 
+/* --- Error handler registry --------------------------------------------- */
+
+static PyObject *register_error(PyObject *self, PyObject *args)
+{
+    const char *name;
+    PyObject *handler;
+
+    if (!PyArg_ParseTuple(args, "sO:register_error",
+			  &name, &handler))
+	return NULL;
+    if (PyCodec_RegisterError(name, handler))
+        return NULL;
+    Py_INCREF(Py_None);
+    return Py_None;
+}
+
+static PyObject *lookup_error(PyObject *self, PyObject *args)
+{
+    const char *name;
+
+    if (!PyArg_ParseTuple(args, "s:lookup_error",
+			  &name))
+	return NULL;
+    return PyCodec_LookupError(name);
+}
+
 /* --- Module API --------------------------------------------------------- */
 
 static PyMethodDef _codecs_functions[] = {
@@ -744,6 +770,8 @@ static PyMethodDef _codecs_functions[] = {
     {"mbcs_decode", 		mbcs_decode,			METH_VARARGS},
 #endif
 #endif /* Py_USING_UNICODE */
+    {"register_error", 		register_error,			METH_VARARGS},
+    {"lookup_error", 		lookup_error,			METH_VARARGS},
     {NULL, NULL}		/* sentinel */
 };
 

Objects/stringobject.c

@@ -2468,7 +2468,9 @@ PyDoc_STRVAR(encode__doc__,
 Encodes S using the codec registered for encoding. encoding defaults\n\
 to the default encoding. errors may be given to set a different error\n\
 handling scheme. Default is 'strict' meaning that encoding errors raise\n\
-a ValueError. Other possible values are 'ignore' and 'replace'.");
+a UnicodeEncodeError. Other possible values are 'ignore', 'replace' and\n\
+'xmlcharrefreplace' as well as any other name registered with\n\
+codecs.register_error that is able to handle UnicodeEncodeErrors.");
 
 static PyObject *
 string_encode(PyStringObject *self, PyObject *args)
@@ -2487,7 +2489,9 @@ PyDoc_STRVAR(decode__doc__,
 Decodes S using the codec registered for encoding. encoding defaults\n\
 to the default encoding. errors may be given to set a different error\n\
 handling scheme. Default is 'strict' meaning that encoding errors raise\n\
-a ValueError. Other possible values are 'ignore' and 'replace'.");
+a UnicodeDecodeError. Other possible values are 'ignore' and 'replace'\n\
+as well as any other name registerd with codecs.register_error that is\n\
+able to handle UnicodeDecodeErrors.");
 
 static PyObject *
 string_decode(PyStringObject *self, PyObject *args)

Python/codecs.c

@@ -422,12 +422,409 @@ PyObject *PyCodec_Decode(PyObject *object,
     return NULL;
 }
 
+static PyObject *_PyCodec_ErrorRegistry;
+
+/* 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)
+{
+    if (!PyCallable_Check(error)) {
+	PyErr_SetString(PyExc_TypeError, "handler must be callable");
+	return -1;
+    }
+    return PyDict_SetItemString( _PyCodec_ErrorRegistry, (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;
+
+    if (name==NULL)
+	name = "strict";
+    handler = PyDict_GetItemString(_PyCodec_ErrorRegistry, (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);
+	    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 (PyInstance_Check(exc))
+	PyErr_SetObject((PyObject*)((PyInstanceObject*)exc)->in_class,
+	    exc);
+    else
+	PyErr_SetString(PyExc_TypeError, "codec must pass exception instance");
+    return NULL;
+}
+
+
+PyObject *PyCodec_IgnoreErrors(PyObject *exc)
+{
+    int 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#i)", &end, 0, end);
+}
+
+
+PyObject *PyCodec_ReplaceErrors(PyObject *exc)
+{
+    PyObject *restuple;
+    int start;
+    int end;
+    int 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("(Oi)", 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#i)", &res, 1, end);
+    }
+    else if (PyObject_IsInstance(exc, PyExc_UnicodeTranslateError)) {
+	PyObject *res;
+	Py_UNICODE *p;
+	if (PyUnicodeTranslateError_GetStart(exc, &start))
+	    return NULL;
+	if (PyUnicodeTranslateError_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 = Py_UNICODE_REPLACEMENT_CHARACTER;
+	restuple = Py_BuildValue("(Oi)", res, end);
+	Py_DECREF(res);
+	return restuple;
+    }
+    else {
+	wrong_exception_type(exc);
+	return NULL;
+    }
+}
+
+PyObject *PyCodec_XMLCharRefReplaceErrors(PyObject *exc)
+{
+    if (PyObject_IsInstance(exc, PyExc_UnicodeEncodeError)) {
+	PyObject *restuple;
+	PyObject *object;
+	int start;
+	int end;
+	PyObject *res;
+	Py_UNICODE *p;
+	Py_UNICODE *startp;
+	Py_UNICODE *outp;
+	int ressize;
+	if (PyUnicodeEncodeError_GetStart(exc, &start))
+	    return NULL;
+	if (PyUnicodeEncodeError_GetEnd(exc, &end))
+	    return NULL;
+	if (!(object = PyUnicodeEncodeError_GetObject(exc)))
+	    return NULL;
+	startp = PyUnicode_AS_UNICODE(object);
+	for (p = startp+start, ressize = 0; p < startp+end; ++p) {
+	    if (*p<10)
+		ressize += 2+1+1;
+	    else if (*p<100)
+		ressize += 2+2+1;
+	    else if (*p<1000)
+		ressize += 2+3+1;
+	    else if (*p<10000)
+		ressize += 2+4+1;
+	    else if (*p<100000)
+		ressize += 2+5+1;
+	    else if (*p<1000000)
+		ressize += 2+6+1;
+	    else
+		ressize += 2+7+1;
+	}
+	/* allocate replacement */
+	res = PyUnicode_FromUnicode(NULL, ressize);
+	if (res == NULL) {
+	    Py_DECREF(object);
+	    return NULL;
+	}
+	/* generate replacement */
+	for (p = startp+start, outp = PyUnicode_AS_UNICODE(res);
+	    p < startp+end; ++p) {
+	    Py_UNICODE c = *p;
+	    int digits;
+	    int base;
+	    *outp++ = '&';
+	    *outp++ = '#';
+	    if (*p<10) {
+		digits = 1;
+		base = 1;
+	    }
+	    else if (*p<100) {
+		digits = 2;
+		base = 10;
+	    }
+	    else if (*p<1000) {
+		digits = 3;
+		base = 100;
+	    }
+	    else if (*p<10000) {
+		digits = 4;
+		base = 1000;
+	    }
+	    else if (*p<100000) {
+		digits = 5;
+		base = 10000;
+	    }
+	    else if (*p<1000000) {
+		digits = 6;
+		base = 100000;
+	    }
+	    else {
+		digits = 7;
+		base = 1000000;
+	    }
+	    while (digits-->0) {
+		*outp++ = '0' + c/base;
+		c %= base;
+		base /= 10;
+	    }
+	    *outp++ = ';';
+	}
+	restuple = Py_BuildValue("(Oi)", res, end);
+	Py_DECREF(res);
+	Py_DECREF(object);
+	return restuple;
+    }
+    else {
+	wrong_exception_type(exc);
+	return NULL;
+    }
+}
+
+static Py_UNICODE hexdigits[] = {
+    '0', '1', '2', '3', '4', '5', '6', '7',
+    '8', '9', 'a', 'b', 'c', 'd', 'e', 'f'
+};
+
+PyObject *PyCodec_BackslashReplaceErrors(PyObject *exc)
+{
+    if (PyObject_IsInstance(exc, PyExc_UnicodeEncodeError)) {
+	PyObject *restuple;
+	PyObject *object;
+	int start;
+	int end;
+	PyObject *res;
+	Py_UNICODE *p;
+	Py_UNICODE *startp;
+	Py_UNICODE *outp;
+	int ressize;
+	if (PyUnicodeEncodeError_GetStart(exc, &start))
+	    return NULL;
+	if (PyUnicodeEncodeError_GetEnd(exc, &end))
+	    return NULL;
+	if (!(object = PyUnicodeEncodeError_GetObject(exc)))
+	    return NULL;
+	startp = PyUnicode_AS_UNICODE(object);
+	for (p = startp+start, ressize = 0; p < startp+end; ++p) {
+	    if (*p >= 0x00010000)
+		ressize += 1+1+8;
+	    else if (*p >= 0x100) {
+		ressize += 1+1+4;
+	    }
+	    else
+		ressize += 1+1+2;
+	}
+	res = PyUnicode_FromUnicode(NULL, ressize);
+	if (res==NULL)
+	    return NULL;
+	for (p = startp+start, outp = PyUnicode_AS_UNICODE(res);
+	    p < startp+end; ++p) {
+	    Py_UNICODE c = *p;
+	    *outp++ = '\\';
+	    if (c >= 0x00010000) {
+		*outp++ = 'U';
+		*outp++ = hexdigits[(c>>28)&0xf];
+		*outp++ = hexdigits[(c>>24)&0xf];
+		*outp++ = hexdigits[(c>>20)&0xf];
+		*outp++ = hexdigits[(c>>16)&0xf];
+		*outp++ = hexdigits[(c>>12)&0xf];
+		*outp++ = hexdigits[(c>>8)&0xf];
+	    }
+	    else if (c >= 0x100) {
+		*outp++ = 'u';
+		*outp++ = hexdigits[(c>>12)&0xf];
+		*outp++ = hexdigits[(c>>8)&0xf];
+	    }
+	    else
+		*outp++ = 'x';
+	    *outp++ = hexdigits[(c>>4)&0xf];
+	    *outp++ = hexdigits[c&0xf];
+	}
+
+	restuple = Py_BuildValue("(Oi)", res, end);
+	Py_DECREF(res);
+	Py_DECREF(object);
+	return restuple;
+    }
+    else {
+	wrong_exception_type(exc);
+	return NULL;
+    }
+}
+
+static PyObject *strict_errors(PyObject *self, PyObject *exc)
+{
+    return PyCodec_StrictErrors(exc);
+}
+
+
+static PyObject *ignore_errors(PyObject *self, PyObject *exc)
+{
+    return PyCodec_IgnoreErrors(exc);
+}
+
+
+static PyObject *replace_errors(PyObject *self, PyObject *exc)
+{
+    return PyCodec_ReplaceErrors(exc);
+}
+
+
+static PyObject *xmlcharrefreplace_errors(PyObject *self, PyObject *exc)
+{
+    return PyCodec_XMLCharRefReplaceErrors(exc);
+}
+
+
+static PyObject *backslashreplace_errors(PyObject *self, PyObject *exc)
+{
+    return PyCodec_BackslashReplaceErrors(exc);
+}
+
+
 void _PyCodecRegistry_Init(void)
 {
+    static struct {
+	char *name;
+	PyMethodDef def;
+    } methods[] =
+    {
+	{
+	    "strict",
+	    {
+		"strict_errors",
+		strict_errors,
+		METH_O
+	    }
+	},
+	{
+	    "ignore",
+	    {
+		"ignore_errors",
+		ignore_errors,
+		METH_O
+	    }
+	},
+	{
+	    "replace",
+	    {
+		"replace_errors",
+		replace_errors,
+		METH_O
+	    }
+	},
+	{
+	    "xmlcharrefreplace",
+	    {
+		"xmlcharrefreplace_errors",
+		xmlcharrefreplace_errors,
+		METH_O
+	    }
+	},
+	{
+	    "backslashreplace",
+	    {
+		"backslashreplace_errors",
+		backslashreplace_errors,
+		METH_O
+	    }
+	}
+    };
     if (_PyCodec_SearchPath == NULL)
 	_PyCodec_SearchPath = PyList_New(0);
     if (_PyCodec_SearchCache == NULL)
 	_PyCodec_SearchCache = PyDict_New();
+    if (_PyCodec_ErrorRegistry == NULL) {
+	int i;
+	_PyCodec_ErrorRegistry = PyDict_New();
+
+	if (_PyCodec_ErrorRegistry) {
+	    for (i = 0; i < 5; ++i) {
+		PyObject *func = PyCFunction_New(&methods[i].def, NULL);
+		int res;
+		if (!func)
+		    Py_FatalError("can't initialize codec error registry");
+		res = PyCodec_RegisterError(methods[i].name, func);
+		Py_DECREF(func);
+		if (res)
+		    Py_FatalError("can't initialize codec error registry");
+	    }
+	}
+    }
     if (_PyCodec_SearchPath == NULL || 
 	_PyCodec_SearchCache == NULL)
 	Py_FatalError("can't initialize codec registry");
@@ -439,4 +836,6 @@ void _PyCodecRegistry_Fini(void)
     _PyCodec_SearchPath = NULL;
     Py_XDECREF(_PyCodec_SearchCache);
     _PyCodec_SearchCache = NULL;
+    Py_XDECREF(_PyCodec_ErrorRegistry);
+    _PyCodec_ErrorRegistry = NULL;
 }