bpo-45822: Respect PEP 263's coding cookies in the parser even if flags are not provided (GH-29582)

pablogsal · web-flow · commit da20d7401de9 · 2021-11-16T12:30:47.000-08:00
diff --git a/Lib/test/test_capi.py b/Lib/test/test_capi.py
@@ -1017,6 +1017,14 @@ def test_state_access(self):
                 with self.assertRaises(TypeError):
                     increment_count(1, 2, 3)
 
+    def test_Py_CompileString(self):
+        # Check that Py_CompileString respects the coding cookie
+        _compile = _testcapi.Py_CompileString
+        code = b"# -*- coding: latin1 -*-\nprint('\xc2\xa4')\n"
+        result = _compile(code)
+        expected = compile(code, "<string>", "exec")
+        self.assertEqual(result.co_consts, expected.co_consts)
+
 
 if __name__ == "__main__":
     unittest.main()
diff --git a/Misc/NEWS.d/next/Core and Builtins/2021-11-16-19-41-04.bpo-45822.OT6ueS.rst b/Misc/NEWS.d/next/Core and Builtins/2021-11-16-19-41-04.bpo-45822.OT6ueS.rst
@@ -0,0 +1,2 @@
+Fixed a bug in the parser that was causing it to not respect :pep:`263`
+coding cookies when no flags are provided. Patch by Pablo Galindo
diff --git a/Modules/_testcapimodule.c b/Modules/_testcapimodule.c
@@ -381,6 +381,19 @@ static PyTypeObject _HashInheritanceTester_Type = {
     PyType_GenericNew,                  /* tp_new */
 };
 
+static PyObject*
+pycompilestring(PyObject* self, PyObject *obj) {
+    if (PyBytes_CheckExact(obj) == 0) {
+        PyErr_SetString(PyExc_ValueError, "Argument must be a bytes object");
+        return NULL;
+    }
+    const char *the_string = PyBytes_AsString(obj);
+    if (the_string == NULL) {
+        return NULL;
+    }
+    return Py_CompileString(the_string, "blech", Py_file_input);
+}
+
 static PyObject*
 test_lazy_hash_inheritance(PyObject* self, PyObject *Py_UNUSED(ignored))
 {
@@ -6070,6 +6083,7 @@ static PyMethodDef TestMethods[] = {
     {"return_null_without_error", return_null_without_error, METH_NOARGS},
     {"return_result_with_error", return_result_with_error, METH_NOARGS},
     {"getitem_with_error", getitem_with_error, METH_VARARGS},
+    {"Py_CompileString",     pycompilestring, METH_O},
     {"PyTime_FromSeconds", test_pytime_fromseconds,  METH_VARARGS},
     {"PyTime_FromSecondsObject", test_pytime_fromsecondsobject,  METH_VARARGS},
     {"PyTime_AsSecondsDouble", test_pytime_assecondsdouble, METH_VARARGS},
diff --git a/Parser/pegen.c b/Parser/pegen.c
@@ -1463,7 +1463,7 @@ _PyPegen_run_parser_from_string(const char *str, int start_rule, PyObject *filen
     int exec_input = start_rule == Py_file_input;
 
     struct tok_state *tok;
-    if (flags == NULL || flags->cf_flags & PyCF_IGNORE_COOKIE) {
+    if (flags != NULL && flags->cf_flags & PyCF_IGNORE_COOKIE) {
         tok = _PyTokenizer_FromUTF8(str, exec_input);
     } else {
         tok = _PyTokenizer_FromString(str, exec_input);

Original file line number	Diff line number	Diff line change
`@@ -0,0 +1,2 @@`
	`1`	+Fixed a bug in the parser that was causing it to not respect :pep:`263`
	`2`	`+coding cookies when no flags are provided. Patch by Pablo Galindo`