Commit 063e0cb4 authored by Marc-André Lemburg's avatar Marc-André Lemburg

Fix to bug #393 (UTF16 codec didn't like empty strings) and

corrected some usage of 'unsigned long' where Py_UNICODE
should have been used.
parent 295b1bbc
......@@ -953,14 +953,13 @@ PyObject *PyUnicode_EncodeUTF16(const Py_UNICODE *s,
sizeof(Py_UNICODE) * (size + (byteorder == 0)));
if (v == NULL)
return NULL;
if (size == 0)
goto done;
q = PyString_AS_STRING(v);
p = (Py_UNICODE *)q;
if (byteorder == 0)
*p++ = 0xFEFF;
if (size == 0)
goto done;
if (byteorder == 0 ||
#ifdef BYTEORDER_IS_LITTLE_ENDIAN
byteorder == -1
......@@ -994,7 +993,7 @@ PyObject *PyUnicode_AsUTF16String(PyObject *unicode)
static
int unicodeescape_decoding_error(const char **source,
unsigned long *x,
Py_UNICODE *x,
const char *errors,
const char *details)
{
......@@ -1009,7 +1008,7 @@ int unicodeescape_decoding_error(const char **source,
return 0;
}
else if (strcmp(errors,"replace") == 0) {
*x = (unsigned long)Py_UNICODE_REPLACEMENT_CHARACTER;
*x = Py_UNICODE_REPLACEMENT_CHARACTER;
return 0;
}
else {
......@@ -1063,7 +1062,7 @@ PyObject *PyUnicode_DecodeUnicodeEscape(const char *s,
end = s + size;
while (s < end) {
unsigned char c;
unsigned long x;
Py_UNICODE x;
int i;
/* Non-escape characters are interpreted as Unicode ordinals */
......@@ -1372,7 +1371,7 @@ PyObject *PyUnicode_DecodeRawUnicodeEscape(const char *s,
end = s + size;
while (s < end) {
unsigned char c;
unsigned long x;
Py_UNICODE x;
int i;
/* Non-escape characters are interpreted as Unicode ordinals */
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment