pythongh-102310: Change error range for invalid bytes literals (pytho…

…n#103663)
Incred · Apr 23, 2023 · 0fd3891 · 0fd3891
1 parent 7bf9456
commit 0fd3891
Show file tree

Hide file tree

Showing 3 changed files with 27 additions and 1 deletion.
diff --git a/Lib/test/test_syntax.py b/Lib/test/test_syntax.py
@@ -1853,6 +1853,30 @@ def f(x: *b)
     Traceback (most recent call last):
         ...
     SyntaxError: invalid syntax
+
+Invalid bytes literals:
+
+   >>> b"Ā"
+   Traceback (most recent call last):
+      ...
+       b"Ā"
+        ^^^
+   SyntaxError: bytes can only contain ASCII literal characters
+
+   >>> b"абвгде"
+   Traceback (most recent call last):
+      ...
+       b"абвгде"
+        ^^^^^^^^
+   SyntaxError: bytes can only contain ASCII literal characters
+
+   >>> b"abc ъющый"  # first 3 letters are ascii
+   Traceback (most recent call last):
+      ...
+       b"abc ъющый"
+        ^^^^^^^^^^^
+   SyntaxError: bytes can only contain ASCII literal characters
+
 """
 
 import re

diff --git a/Misc/NEWS.d/next/Core and Builtins/2023-04-21-17-03-14.gh-issue-102310.anLjDx.rst b/Misc/NEWS.d/next/Core and Builtins/2023-04-21-17-03-14.gh-issue-102310.anLjDx.rst
@@ -0,0 +1 @@
+Change the error range for invalid bytes literals.
diff --git a/Parser/string_parser.c b/Parser/string_parser.c
@@ -248,7 +248,8 @@ _PyPegen_parse_string(Parser *p, Token *t)
         const char *ch;
         for (ch = s; *ch; ch++) {
             if (Py_CHARMASK(*ch) >= 0x80) {
-                RAISE_SYNTAX_ERROR(
+                RAISE_SYNTAX_ERROR_KNOWN_LOCATION(
+                                   t,
                                    "bytes can only contain ASCII "
                                    "literal characters");
                 return NULL;