From 65c8288b99dd93602f771417b86c0455a699a7b6 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Stefanos=20Carlstr=C3=B6m?= <stefanos.carlstrom@gmail.com>
Date: Wed, 5 Nov 2025 10:48:30 +0100
Subject: [PATCH] Support XID_Start/XID_Continue Unicode chars in C++
 identifiers (#14026)

---
 AUTHORS.rst                           | 1 +
 CHANGES.rst                           | 4 ++++
 pyproject.toml                        | 1 +
 sphinx/util/cfamily.py                | 8 ++++++--
 tests/test_domains/test_domain_cpp.py | 4 ++++
 5 files changed, 16 insertions(+), 2 deletions(-)

diff --git a/AUTHORS.rst b/AUTHORS.rst
index 707c77aec04..e37a9f01d88 100644
--- a/AUTHORS.rst
+++ b/AUTHORS.rst
@@ -105,6 +105,7 @@ Contributors
 * Slawek Figiel -- additional warning suppression
 * Stefan Seefeld -- toctree improvements
 * Stefan van der Walt -- autosummary extension
+* Stefanos Carlström -- minor bug fix
 * Steve Piercy -- documentation improvements
 * Szymon Karpinski -- intersphinx improvements
 * \T. Powers -- HTML output improvements
diff --git a/CHANGES.rst b/CHANGES.rst
index f0f94fda396..4ee66597dd8 100644
--- a/CHANGES.rst
+++ b/CHANGES.rst
@@ -4,6 +4,10 @@ Release 8.3.0 (in development)
 Dependencies
 ------------
 
+* #14026: Added `regex >= 2025.7.29`_, to support parsing for Unicode classes.
+
+  .. _regex >= 2025.7.29: _https://github.com/mrabarnett/mrab-regex
+
 * #13786: Support `Docutils 0.22`_. Patch by Adam Turner.
 
   .. _Docutils 0.22: https://docutils.sourceforge.io/RELEASE-NOTES.html#release-0-22-2026-07-29
diff --git a/pyproject.toml b/pyproject.toml
index e3418ab98a2..88b3c479f57 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -86,6 +86,7 @@ dependencies = [
     "packaging>=23.0",
     "colorama>=0.4.6; sys_platform == 'win32'",
     "ipython>=9.6.0",
+    "regex>=2025.7.29"
 ]
 dynamic = ["version"]
 
diff --git a/sphinx/util/cfamily.py b/sphinx/util/cfamily.py
index 6071d90cf74..385ef9f7b02 100644
--- a/sphinx/util/cfamily.py
+++ b/sphinx/util/cfamily.py
@@ -2,7 +2,7 @@
 
 from __future__ import annotations
 
-import re
+import regex as re
 from copy import deepcopy
 from typing import TYPE_CHECKING
 
@@ -30,9 +30,13 @@
     (   # This 'extends' _anon_identifier_re with the ordinary identifiers,
         # make sure they are in sync.
         (~?\b[a-zA-Z_])  # ordinary identifiers
+    |   \p{XID_Start}    # Unicode-allowed starting characters for identifiers
     |   (@[a-zA-Z0-9_])  # our extension for names of anonymous entities
     )
-    [a-zA-Z0-9_]*\b
+    (
+        [a-zA-Z0-9_]     # ordinary identifiers
+    |   \p{XID_Continue} # Unicode-allowed continuing characters for identifiers
+    )*\b
     """,
     flags=re.VERBOSE,
 )
diff --git a/tests/test_domains/test_domain_cpp.py b/tests/test_domains/test_domain_cpp.py
index 2e88625a9fe..4c3172e9703 100644
--- a/tests/test_domains/test_domain_cpp.py
+++ b/tests/test_domains/test_domain_cpp.py
@@ -860,6 +860,10 @@ def test_domain_cpp_ast_function_definitions() -> None:
 
     check('function', 'decltype(auto) f()', {1: 'f', 2: '1fv'})
 
+    # Test derived from https://github.com/sphinx-doc/sphinx/issues/14026
+    # Unicode identifiers
+    check('function', 'void f(int *const ξ)', {1: 'f__iPC', 2: '1fPCi'})
+
     # TODO: make tests for functions in a template, e.g., Test<int&&()>
     # such that the id generation for function type types is correct.