From 0fd38917582aae0728e20d8a641e56d9be9270c7 Mon Sep 17 00:00:00 2001
From: Nikita Sobolev <mail@sobolevn.me>
Date: Sun, 23 Apr 2023 03:08:27 +0300
Subject: [PATCH] gh-102310: Change error range for invalid bytes literals
 (#103663)

---
 Lib/test/test_syntax.py                       | 24 +++++++++++++++++++
 ...-04-21-17-03-14.gh-issue-102310.anLjDx.rst |  1 +
 Parser/string_parser.c                        |  3 ++-
 3 files changed, 27 insertions(+), 1 deletion(-)
 create mode 100644 Misc/NEWS.d/next/Core and Builtins/2023-04-21-17-03-14.gh-issue-102310.anLjDx.rst

diff --git a/Lib/test/test_syntax.py b/Lib/test/test_syntax.py
index f23653558a9..f959bbb4400 100644
--- a/Lib/test/test_syntax.py
+++ b/Lib/test/test_syntax.py
@@ -1853,6 +1853,30 @@ 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
new file mode 100644
index 00000000000..15cb6c64adb
--- /dev/null
+++ 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
index be5f0c4a60a..d4ce33850f7 100644
--- 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;