[3.6] bpo-31825: Fixed OverflowError in the 'unicode-escape' codec (GH-4058)

Lib/test/test_codecs.py

@@ -1203,6 +1203,8 @@ def test_escape(self):
             check(br"\8", b"\\8")
         with self.assertWarns(DeprecationWarning):
             check(br"\9", b"\\9")
+        with self.assertWarns(DeprecationWarning):
+            check(b"\\\xfa", b"\\\xfa")
 
     def test_errors(self):
         decode = codecs.escape_decode
@@ -2474,6 +2476,8 @@ def test_escape_decode(self):
             check(br"\8", "\\8")
         with self.assertWarns(DeprecationWarning):
             check(br"\9", "\\9")
+        with self.assertWarns(DeprecationWarning):
+            check(b"\\\xfa", "\\\xfa")
 
     def test_decode_errors(self):
         decode = codecs.unicode_escape_decode

Misc/NEWS.d/next/Core and Builtins/2017-10-20-14-07-46.bpo-31825.gJvmGW.rst

@@ -0,0 +1,2 @@
+Fixed OverflowError in the 'unicode-escape' codec and in
+codecs.escape_decode() when decode an escaped non-ascii byte.

Objects/bytesobject.c

@@ -1255,7 +1255,7 @@ PyObject *PyBytes_DecodeEscape(const char *s,
     if (first_invalid_escape != NULL) {
         if (PyErr_WarnFormat(PyExc_DeprecationWarning, 1,
                              "invalid escape sequence '\\%c'",
-                             *first_invalid_escape) < 0) {
+                             (unsigned char)*first_invalid_escape) < 0) {
             Py_DECREF(result);
             return NULL;
         }

Objects/unicodeobject.c

@@ -6174,7 +6174,7 @@ PyUnicode_DecodeUnicodeEscape(const char *s,
     if (first_invalid_escape != NULL) {
         if (PyErr_WarnFormat(PyExc_DeprecationWarning, 1,
                              "invalid escape sequence '\\%c'",
-                             *first_invalid_escape) < 0) {
+                             (unsigned char)*first_invalid_escape) < 0) {
             Py_DECREF(result);
             return NULL;
         }

Python/ast.c

@@ -4127,7 +4127,7 @@ decode_utf8(struct compiling *c, const char **sPtr, const char *end)
 
 static int
 warn_invalid_escape_sequence(struct compiling *c, const node *n,
-                             char first_invalid_escape_char)
+                             unsigned char first_invalid_escape_char)
 {
     PyObject *msg = PyUnicode_FromFormat("invalid escape sequence \\%c",
                                          first_invalid_escape_char);