[3.11] gh-115823: Calculate correctly error locations when dealing with implicit encodings (GH-115824) (#115950)

gh-115823: Calculate correctly error locations when dealing with implicit encodings (GH-115824)
(cherry picked from commit 015b97d)

Co-authored-by: Pablo Galindo Salgado <[email protected]>

Author: miss-islington

Lib/test/test_exceptions.py

@@ -234,7 +234,7 @@ def testSyntaxErrorOffset(self):
         check('Python = "\u1e54\xfd\u0163\u0125\xf2\xf1" +', 1, 20)
         check(b'# -*- coding: cp1251 -*-\nPython = "\xcf\xb3\xf2\xee\xed" +',
               2, 19, encoding='cp1251')
-        check(b'Python = "\xcf\xb3\xf2\xee\xed" +', 1, 18)
+        check(b'Python = "\xcf\xb3\xf2\xee\xed" +', 1, 13)
         check('x = "a', 1, 5)
         check('lambda x: x = 2', 1, 1)
         check('f{a + b + c}', 1, 2)
@@ -301,6 +301,7 @@ def baz():
             {
             6
             0="""''', 5, 13)
+        check('b"fooжжж"'.encode(), 1, 1, 1, 10)
 
         # Errors thrown by symtable.c
         check('x = [(yield i) for i in range(3)]', 1, 7)

Misc/NEWS.d/next/Core and Builtins/2024-02-22-16-17-53.gh-issue-115823.c1TreJ.rst

@@ -0,0 +1,3 @@
+Properly calculate error ranges in the parser when raising
+:exc:`SyntaxError` exceptions caused by invalid byte sequences. Patch by
+Pablo Galindo

Parser/pegen_errors.c

@@ -377,20 +377,18 @@ _PyPegen_raise_error_known_location(Parser *p, PyObject *errtype,
     Py_ssize_t col_number = col_offset;
     Py_ssize_t end_col_number = end_col_offset;
 
-    if (p->tok->encoding != NULL) {
-        col_number = _PyPegen_byte_offset_to_character_offset(error_line, col_offset);
-        if (col_number < 0) {
+    col_number = _PyPegen_byte_offset_to_character_offset(error_line, col_offset);
+    if (col_number < 0) {
+        goto error;
+    }
+
+    if (end_col_offset > 0) {
+        end_col_number = _PyPegen_byte_offset_to_character_offset(error_line, end_col_offset);
+        if (end_col_number < 0) {
             goto error;
         }
-        if (end_col_number > 0) {
-            Py_ssize_t end_col_offset = _PyPegen_byte_offset_to_character_offset(error_line, end_col_number);
-            if (end_col_offset < 0) {
-                goto error;
-            } else {
-                end_col_number = end_col_offset;
-            }
-        }
     }
+
     tmp = Py_BuildValue("(OnnNnn)", p->tok->filename, lineno, col_number, error_line, end_lineno, end_col_number);
     if (!tmp) {
         goto error;