gh-160491: Fix location of invalid escapes in f-strings and t-strings

Literal parts of f-strings and t-strings are decoded only after the
whole string has been parsed, so a decode error took the location of
the last token. The resulting SyntaxError pointed at the closing quote,
and for multi-line strings it could be on the wrong line.

Raise the error at the literal part being decoded instead.

Author: lysnikolaou

Lib/test/test_fstring.py

@@ -1096,6 +1096,24 @@ def test_misformed_unicode_character_name(self):
                              r"'\N{GREEK CAPITAL LETTER DELTA'",
                              ])
 
+    def test_invalid_escape_location(self):
+        cases = [
+            (r'f"\xZZ"', (1, 3, 1, 7)),
+            (r'f"{x}\xZZ"', (1, 6, 1, 10)),
+            (r'f"a{x}b\N{NOPE}{y}"', (1, 7, 1, 16)),
+            (r'f"ok" f"\xZZ"', (1, 9, 1, 13)),
+            ('f"""\\xZZ\nabc"""', (1, 5, 2, 4)),
+            ('f"""abc\ndef\\xZZ"""', (1, 5, 2, 8)),
+            ('x = (1,\n  f"\\xZZ")', (2, 5, 2, 9)),
+        ]
+        for source, location in cases:
+            with self.subTest(source=source):
+                with self.assertRaisesRegex(SyntaxError, r'\(unicode error\)') as cm:
+                    compile(source, '<string>', 'exec')
+                error = cm.exception
+                self.assertEqual((error.lineno, error.offset,
+                                  error.end_lineno, error.end_offset), location)
+
     def test_backslashes_in_expression_part(self):
         self.assertEqual(f"{(
                         1 +

Parser/action_helpers.c

@@ -1373,7 +1373,7 @@ _PyPegen_decode_fstring_part(Parser* p, int is_raw, expr_ty constant, Token* tok
     is_raw = is_raw || strchr(bstr, '\\') == NULL;
     PyObject *str = _PyPegen_decode_string(p, is_raw, bstr, len, token);
     if (str == NULL) {
-        _Pypegen_raise_decode_error(p);
+        _Pypegen_raise_decode_error_known_location(p, constant);
         return NULL;
     }
     if (_PyArena_AddPyObject(p->arena, str) < 0) {

Parser/pegen.h

@@ -212,6 +212,7 @@ typedef enum {
 
 // Error handling functions and APIs
 int _Pypegen_raise_decode_error(Parser *p);
+int _Pypegen_raise_decode_error_known_location(Parser *p, expr_ty a);
 int _Pypegen_tokenizer_error(Parser *p);
 void *_PyPegen_raise_incomplete_input(Parser *p);
 void *_PyPegen_raise_error(Parser *p, PyObject *errtype, int use_mark, const char *errmsg, ...);

Parser/pegen_errors.c

@@ -103,6 +103,12 @@ _Pypegen_tokenizer_error(Parser *p)
 
 int
 _Pypegen_raise_decode_error(Parser *p)
+{
+    return _Pypegen_raise_decode_error_known_location(p, NULL);
+}
+
+int
+_Pypegen_raise_decode_error_known_location(Parser *p, expr_ty a)
 {
     assert(PyErr_Occurred());
     const char *errtype = NULL;
@@ -120,12 +126,22 @@ _Pypegen_raise_decode_error(Parser *p)
         PyErr_Fetch(&type, &value, &tback);
         errstr = PyObject_Str(value);
         if (errstr) {
-            RAISE_SYNTAX_ERROR("(%s) %U", errtype, errstr);
+            if (a != NULL) {
+                RAISE_SYNTAX_ERROR_KNOWN_LOCATION(a, "(%s) %U", errtype, errstr);
+            }
+            else {
+                RAISE_SYNTAX_ERROR("(%s) %U", errtype, errstr);
+            }
             Py_DECREF(errstr);
         }
         else {
             PyErr_Clear();
-            RAISE_SYNTAX_ERROR("(%s) unknown error", errtype);
+            if (a != NULL) {
+                RAISE_SYNTAX_ERROR_KNOWN_LOCATION(a, "(%s) unknown error", errtype);
+            }
+            else {
+                RAISE_SYNTAX_ERROR("(%s) unknown error", errtype);
+            }
         }
         Py_XDECREF(type);
         Py_XDECREF(value);