# HG changeset patch
# Parent cdc3837cb1fcfc4273b46e84c533ec7e91246b71
Issue 21071: Struct.format is now a str

diff -r cdc3837cb1fc Doc/library/struct.rst
--- a/Doc/library/struct.rst	Tue Dec 16 03:21:54 2014 -0500
+++ b/Doc/library/struct.rst	Thu Dec 18 05:23:42 2014 +0000
@@ -420,6 +420,10 @@
 
       The format string used to construct this Struct object.
 
+      .. versionchanged:: 3.5
+         Previously this was a :class:`bytes` object;
+         now it is a :class:`str` object.
+
    .. attribute:: size
 
       The calculated size of the struct (and hence of the bytes object produced
diff -r cdc3837cb1fc Lib/test/test_struct.py
--- a/Lib/test/test_struct.py	Tue Dec 16 03:21:54 2014 -0500
+++ b/Lib/test/test_struct.py	Thu Dec 18 05:23:42 2014 +0000
@@ -589,6 +589,16 @@
         self.check_sizeof('20p', 1)
         self.check_sizeof('0s', 1)
         self.check_sizeof('0c', 0)
+    
+    def test_format_attr(self):
+        """Make sure the "format" attribute is a text string"""
+        self.assertEqual("x", struct.Struct("x").format)
+
+    def test_format_utf(self):
+        """Check error handling for unencodable format string"""
+        surrogate = "\uDC80"
+        self.assertRaises(UnicodeError, struct.pack, surrogate)
+        self.assertRaises(UnicodeError, struct.Struct, surrogate)
 
 
 class UnpackIteratorTest(unittest.TestCase):
diff -r cdc3837cb1fc Modules/_struct.c
--- a/Modules/_struct.c	Tue Dec 16 03:21:54 2014 -0500
+++ b/Modules/_struct.c	Thu Dec 18 05:23:42 2014 +0000
@@ -1266,7 +1266,10 @@
     Py_ssize_t size, len, num, itemsize;
     size_t ncodes;
 
-    fmt = PyBytes_AS_STRING(self->s_format);
+    fmt = PyUnicode_AsUTF8(self->s_format);
+    if (fmt == NULL) {
+        return -1;
+    }
 
     f = whichtable((char **)&fmt);
 
@@ -1419,8 +1422,10 @@
                                      &o_format))
         return -1;
 
-    if (PyUnicode_Check(o_format)) {
-        o_format = PyUnicode_AsASCIIString(o_format);
+    if (PyBytes_Check(o_format)) {
+        o_format = PyUnicode_DecodeASCII(
+            PyBytes_AS_STRING(o_format), PyBytes_GET_SIZE(o_format),
+            "strict");
         if (o_format == NULL)
             return -1;
     }
@@ -1429,10 +1434,10 @@
         Py_INCREF(o_format);
     }
 
-    if (!PyBytes_Check(o_format)) {
+    if (!PyUnicode_Check(o_format)) {
         Py_DECREF(o_format);
         PyErr_Format(PyExc_TypeError,
-                     "Struct() argument 1 must be a bytes object, not %.200s",
+                     "Struct() argument 1 must be a string object, not %.200s",
                      Py_TYPE(o_format)->tp_name);
         return -1;
     }
@@ -2221,9 +2226,8 @@
 
 PyDoc_STRVAR(module_doc,
 "Functions to convert between Python values and C structs.\n\
-Python bytes objects are used to hold the data representing the C struct\n\
-and also as format strings (explained below) to describe the layout of data\n\
-in the C struct.\n\
+Python bytes objects are used to hold the data representing the C struct.\n\
+Format strings describe the layout of data in the C struct.\n\
 \n\
 The optional first format char indicates byte order, size and alignment:\n\
   @: native order, size & alignment (default)\n\