path: root/source/l/libxml2/libxml2-2.12.0-python3-unicode-errors.patch

diff --git a/python/libxml.c b/python/libxml.c
index bf048006..5f42e5b7 100644
--- a/python/libxml.c
+++ b/python/libxml.c
@@ -1505,6 +1505,7 @@ libxml_xmlErrorFuncHandler(ATTRIBUTE_UNUSED void *ctx, const char *msg,
     PyObject *message;
     PyObject *result;
     char str[1000];
+    unsigned char *ptr = (unsigned char *)str;
 
     if (libxml_xmlPythonErrorFuncHandler == NULL) {
         va_start(ap, msg);
@@ -1516,12 +1517,20 @@ libxml_xmlErrorFuncHandler(ATTRIBUTE_UNUSED void *ctx, const char *msg,
 	    str[999] = 0;
         va_end(ap);
 
+#if PY_MAJOR_VERSION >= 3
+        /* Ensure the error string doesn't start at UTF8 continuation. */
+        while (*ptr && (*ptr & 0xc0) == 0x80)
+            ptr++;
+#endif
+
         list = PyTuple_New(2);
         PyTuple_SetItem(list, 0, libxml_xmlPythonErrorFuncCtxt);
         Py_XINCREF(libxml_xmlPythonErrorFuncCtxt);
-        message = libxml_charPtrConstWrap(str);
+        message = libxml_charPtrConstWrap(ptr);
         PyTuple_SetItem(list, 1, message);
         result = PyObject_CallObject(libxml_xmlPythonErrorFuncHandler, list);
+        /* Forget any errors caused in the error handler. */
+        PyErr_Clear();
         Py_XDECREF(list);
         Py_XDECREF(result);
     }