Python strings ending with '\0' should not be equivalent to their C counterparts in PyUnicode_CompareWithASCIIString
diff --git a/Misc/NEWS b/Misc/NEWS
index adf9fb3..fc9fa63 100644
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -155,6 +155,9 @@
C-API
-----
+- Make PyUnicode_CompareWithASCIIString return not equal if the Python string
+ has '\0' at the end.
+
- Issue #5080: The argument parsing functions PyArg_ParseTuple,
PyArg_ParseTupleAndKeywords, PyArg_VaParse,
PyArg_VaParseTupleAndKeywords and PyArg_Parse now raise a
diff --git a/Modules/_testcapimodule.c b/Modules/_testcapimodule.c
index 19afbf2..5671783 100644
--- a/Modules/_testcapimodule.c
+++ b/Modules/_testcapimodule.c
@@ -1287,6 +1287,23 @@
#undef CHECK_1_FORMAT
}
+
+static PyObject *
+test_unicode_compare_with_ascii(PyObject *self) {
+ PyObject *py_s = PyUnicode_FromStringAndSize("str\0", 4);
+ int result;
+ if (py_s == NULL)
+ return NULL;
+ result = PyUnicode_CompareWithASCIIString(py_s, "str");
+ Py_DECREF(py_s);
+ if (!result) {
+ PyErr_SetString(TestError, "Python string ending in NULL "
+ "should not compare equal to c string.");
+ return NULL;
+ }
+ Py_RETURN_NONE;
+};
+
/* This is here to provide a docstring for test_descr. */
static PyObject *
test_with_docstring(PyObject *self)
@@ -1756,6 +1773,7 @@
{"test_with_docstring", (PyCFunction)test_with_docstring, METH_NOARGS,
PyDoc_STR("This is a pretty normal docstring.")},
{"test_string_to_double", (PyCFunction)test_string_to_double, METH_NOARGS},
+ {"test_unicode_compare_with_ascii", (PyCFunction)test_unicode_compare_with_ascii, METH_NOARGS},
{"test_capsule", (PyCFunction)test_capsule, METH_NOARGS},
{"getargs_tuple", getargs_tuple, METH_VARARGS},
{"getargs_keywords", (PyCFunction)getargs_keywords,
diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c
index 9c0be9b..40377ea 100644
--- a/Objects/unicodeobject.c
+++ b/Objects/unicodeobject.c
@@ -7001,6 +7001,11 @@
for (i = 0; id[i] && str[i]; i++)
if (id[i] != str[i])
return ((int)id[i] < (int)str[i]) ? -1 : 1;
+ /* This check keeps Python strings that end in '\0' from comparing equal
+ to C strings identical up to that point. */
+ if (PyUnicode_GET_SIZE(uni) != i)
+ /* We'll say the Python string is longer. */
+ return 1;
if (id[i])
return 1; /* uni is longer */
if (str[i])