Retain non-sanitized dtstr for error printing

This does not create an extra string, it just holds on to a reference to
the original input string for purposes of creating the error message.

Author: pganssle

Lib/test/datetimetester.py

@@ -13,6 +13,7 @@
 import os
 import pickle
 import random
+import re
 import struct
 import unittest
 
@@ -2676,6 +2677,14 @@ def test_fromisoformat_fails_datetime(self):
                 with self.assertRaises(ValueError):
                     self.theclass.fromisoformat(bad_str)
 
+    def test_fromisoformat_fails_surrogate(self):
+        # Test that when fromisoformat() fails with a surrogate character as
+        # the separator, the error message contains the original string
+        dtstr = "2018-01-03\ud80001:0113"
+
+        with self.assertRaisesRegex(ValueError, re.escape(repr(dtstr))):
+            self.theclass.fromisoformat(dtstr)
+
     def test_fromisoformat_utc(self):
         dt_str = '2014-04-19T13:21:13+00:00'
         dt = self.theclass.fromisoformat(dt_str)

Modules/_datetimemodule.c

@@ -4902,13 +4902,13 @@ datetime_fromisoformat(PyObject* cls, PyObject *dtstr) {
     }
 
     int needs_decref = 0;
-    dtstr = _sanitize_isoformat_str(dtstr, &needs_decref);
-    if (dtstr == NULL) {
+    PyObject *dtstr_clean = _sanitize_isoformat_str(dtstr, &needs_decref);
+    if (dtstr_clean == NULL) {
         goto error;
     }
 
     Py_ssize_t len;
-    const char * dt_ptr = PyUnicode_AsUTF8AndSize(dtstr, &len);
+    const char * dt_ptr = PyUnicode_AsUTF8AndSize(dtstr_clean, &len);
 
     if (dt_ptr == NULL) {
         goto invalid_string_error;
@@ -4960,7 +4960,7 @@ datetime_fromisoformat(PyObject* cls, PyObject *dtstr) {
 
     Py_DECREF(tzinfo);
     if (needs_decref) {
-        Py_DECREF(dtstr);
+        Py_DECREF(dtstr_clean);
     }
     return dt;
 
@@ -4969,7 +4969,7 @@ datetime_fromisoformat(PyObject* cls, PyObject *dtstr) {
 
 error:
     if (needs_decref) {
-        Py_DECREF(dtstr);
+        Py_DECREF(dtstr_clean);
     }
 
     return NULL;