Issue #19099: The struct module now supports Unicode format strings.
diff --git a/Lib/test/test_struct.py b/Lib/test/test_struct.py
index 2e613f7..4811974 100644
--- a/Lib/test/test_struct.py
+++ b/Lib/test/test_struct.py
@@ -574,6 +574,18 @@
self.check_sizeof('0s', 1)
self.check_sizeof('0c', 0)
+ def test_unicode_format(self):
+ try:
+ unicode
+ except NameError:
+ self.skipTest('no unicode support')
+ # Issue #19099
+ s = struct.Struct(unichr(ord('I')))
+ self.assertEqual(s.format, 'I')
+ self.assertIs(type(s.format), str)
+ self.assertRaises(ValueError, struct.Struct, unichr(0x80))
+
+
def test_main():
support.run_unittest(StructTest)
diff --git a/Misc/NEWS b/Misc/NEWS
index 928a18e..e3abaec 100644
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -15,6 +15,8 @@
Library
-------
+- Issue #19099: The struct module now supports Unicode format strings.
+
- Issue #19878: Fix segfault in bz2 module after calling __init__ twice with
non-existent filename. Initial patch by Vajrasky Kok.
diff --git a/Modules/_struct.c b/Modules/_struct.c
index 8056956..f035847 100644
--- a/Modules/_struct.c
+++ b/Modules/_struct.c
@@ -1371,13 +1371,28 @@
assert(PyStruct_Check(self));
- if (!PyArg_ParseTupleAndKeywords(args, kwds, "S:Struct", kwlist,
+ if (!PyArg_ParseTupleAndKeywords(args, kwds, "O:Struct", kwlist,
&o_format))
return -1;
- Py_INCREF(o_format);
- Py_CLEAR(soself->s_format);
- soself->s_format = o_format;
+ if (PyString_Check(o_format)) {
+ Py_INCREF(o_format);
+ Py_CLEAR(soself->s_format);
+ soself->s_format = o_format;
+ }
+ else if (PyUnicode_Check(o_format)) {
+ PyObject *str = PyUnicode_AsEncodedString(o_format, "ascii", NULL);
+ if (str == NULL)
+ return -1;
+ Py_CLEAR(soself->s_format);
+ soself->s_format = str;
+ }
+ else {
+ PyErr_Format(PyExc_TypeError,
+ "Struct() argument 1 must be string, not %s",
+ Py_TYPE(o_format)->tp_name);
+ return -1;
+ }
ret = prepare_s(soself);
return ret;