From 766afad56828818ed0e47ed59afbcb2a7f4f0546 Mon Sep 17 00:00:00 2001
From: amansri-30 <srivastavaaman6915@gmail.com>
Date: Sat, 19 Sep 2026 02:09:06 +0530
Subject: [PATCH] feat: add support for 'a' (ASCII) and 'L' (LOCALE) regex
 flags

Adds support for the 'a' and 'L' regex flags in Lark grammars, as requested
in issue #1527.

- 'a' flag: Makes \w, \W, \b, \B, \d, \D, \s, \S perform ASCII-only
  matching instead of full Unicode matching
- 'L' flag: Makes \w, \W, \b, \B locale dependent (only works with
  bytes patterns in Python 3)

Changes:
- Updated _RE_FLAGS in load_grammar.py to include 'a' and 'L'
- Added test_flag_order_with_new_flags to verify 'a' flag works
- Added test_token_flags_ascii to verify 'a' flag parses correctly
---
 lark/load_grammar.py | 2 +-
 tests/test_lexer.py  | 7 +++++++
 tests/test_parser.py | 8 ++++++++
 3 files changed, 16 insertions(+), 1 deletion(-)

diff --git a/lark/load_grammar.py b/lark/load_grammar.py
index b98f166c1..171d03d83 100644
--- a/lark/load_grammar.py
+++ b/lark/load_grammar.py
@@ -29,7 +29,7 @@
 
 EXT = '.lark'
 
-_RE_FLAGS = 'imslux'
+_RE_FLAGS = 'imsluxaL'
 
 _EMPTY = Symbol('__empty__')
 
diff --git a/tests/test_lexer.py b/tests/test_lexer.py
index b080328a5..e1a3c3e49 100644
--- a/tests/test_lexer.py
+++ b/tests/test_lexer.py
@@ -27,6 +27,13 @@ def test_flag_order_is_deterministic(self):
         self.assertEqual([t.pattern.to_regexp() for t in p.terminals],
                          ['(?u:(?s:(?m:(?i:x))))'])
 
+    def test_flag_order_with_new_flags(self):
+        # Test that new flag (a) works and is sorted deterministically
+        p = Lark('start: A+\nA: /x/ai\n', parser='lalr')
+        # Flags should be sorted alphabetically: a before i
+        self.assertEqual([t.pattern.to_regexp() for t in p.terminals],
+                         ['(?i:(?a:x))'])
+
     def test_subset_lex(self):
         p = Lark("""
             start: "a" "b" "c" "d"
diff --git a/tests/test_parser.py b/tests/test_parser.py
index b12752b07..a7c2f2963 100644
--- a/tests/test_parser.py
+++ b/tests/test_parser.py
@@ -1808,6 +1808,14 @@ def test_token_flags_verbose_multiline(self):
             x = g.parse('abcdef')
             self.assertEqual(x.children, ['abcdef'])
 
+        def test_token_flags_ascii(self):
+            # Test the ASCII flag (a) - makes \w, \d, \s ASCII-only
+            l = _Lark(r"""!start: /\d/a+
+                      """)
+            # Should match ASCII digits
+            tree = l.parse('123')
+            self.assertEqual(tree.children, ['1', '2', '3'])
+
         @unittest.skipIf(PARSER == 'cyk', "No empty rules")
         def test_twice_empty(self):
             g = """!start: ("A"?)?