[libc++] proper guarding for locale usage in filesystem on Windows #165470

moleium · 2025-10-28T20:06:22Z

Resolves build issues when localization support is disabled on Windows.
Resolves dependencies on localization in filesystem header implementations.

- Resolves build issues when localization support is disabled on Windows. - Resolves dependencies on localization in filesystem header implementations. Related PR llvm#164602 Fixes llvm#164074

github-actions · 2025-10-28T20:06:40Z

Thank you for submitting a Pull Request (PR) to the LLVM Project!

This PR will be automatically labeled and the relevant teams will be notified.

If you wish to, you can add reviewers by using the "Reviewers" section on this page.

If this is not working for you, it is probably because you do not have write permissions for the repository. In which case you can instead tag reviewers by name in a comment by using @ followed by their GitHub username.

If you have received no comments on your PR for a week, you can request a review by "ping"ing the PR by adding a comment “Ping”. The common courtesy "ping" rate is once a week. Please remember that you are asking for valuable time from other developers.

If you have further questions, they may be answered by the LLVM GitHub User Guide.

You can also ask questions in a comment on this PR, on the LLVM Discord or on the forums.

llvmbot · 2025-10-28T20:07:15Z

@llvm/pr-subscribers-platform-windows

@llvm/pr-subscribers-libcxx

Author: None (moleium)

Changes

Resolves build issues when localization support is disabled on Windows.
Resolves dependencies on localization in filesystem header implementations.

Related PR #164602
Fixes #164074

Full diff: https://github.com/llvm/llvm-project/pull/165470.diff

2 Files Affected:

(modified) libcxx/include/__filesystem/path.h (+4-2)
(modified) libcxx/include/__filesystem/u8path.h (+23-19)

diff --git a/libcxx/include/__filesystem/path.h b/libcxx/include/__filesystem/path.h
index b3f324342f800..990833116aebe 100644
--- a/libcxx/include/__filesystem/path.h
+++ b/libcxx/include/__filesystem/path.h
@@ -324,6 +324,7 @@ struct _PathCVT<char> {
   }
 };
 
+#    if _LIBCPP_HAS_LOCALIZATION
 template <class _ECharT>
 struct _PathExport {
   typedef __narrow_to_utf8<sizeof(wchar_t) * __CHAR_BIT__> _Narrower;
@@ -364,7 +365,7 @@ struct _PathExport<char16_t> {
   }
 };
 
-#    if _LIBCPP_HAS_CHAR8_T
+#      if _LIBCPP_HAS_CHAR8_T
 template <>
 struct _PathExport<char8_t> {
   typedef __narrow_to_utf8<sizeof(wchar_t) * __CHAR_BIT__> _Narrower;
@@ -374,7 +375,8 @@ struct _PathExport<char8_t> {
     _Narrower()(back_inserter(__dest), __src.data(), __src.data() + __src.size());
   }
 };
-#    endif // _LIBCPP_HAS_CHAR8_T
+#      endif // _LIBCPP_HAS_CHAR8_T
+#    endif // _LIBCPP_HAS_LOCALIZATION
 #  endif   /* _LIBCPP_WIN32API */
 
 class _LIBCPP_EXPORTED_FROM_ABI path {
diff --git a/libcxx/include/__filesystem/u8path.h b/libcxx/include/__filesystem/u8path.h
index 885372b0d1f86..7077b96ce48e6 100644
--- a/libcxx/include/__filesystem/u8path.h
+++ b/libcxx/include/__filesystem/u8path.h
@@ -13,7 +13,10 @@
 #include <__algorithm/unwrap_iter.h>
 #include <__config>
 #include <__filesystem/path.h>
+
+#if _LIBCPP_HAS_LOCALIZATION
 #include <__locale>
+#endif
 #include <string>
 
 #if !defined(_LIBCPP_HAS_NO_PRAGMA_SYSTEM_HEADER)
@@ -24,28 +27,30 @@
 
 _LIBCPP_BEGIN_NAMESPACE_FILESYSTEM
 
+#  if !defined(_LIBCPP_WIN32API) || _LIBCPP_HAS_LOCALIZATION
 template <class _InputIt, __enable_if_t<__is_pathable<_InputIt>::value, int> = 0>
 _LIBCPP_HIDE_FROM_ABI _LIBCPP_DEPRECATED_WITH_CHAR8_T path u8path(_InputIt __f, _InputIt __l) {
   static_assert(
-#  if _LIBCPP_HAS_CHAR8_T
-      is_same<typename __is_pathable<_InputIt>::__char_type, char8_t>::value ||
-#  endif
-          is_same<typename __is_pathable<_InputIt>::__char_type, char>::value,
-      "u8path(Iter, Iter) requires Iter have a value_type of type 'char'"
-      " or 'char8_t'");
-#  if defined(_LIBCPP_WIN32API)
-  string __tmp(__f, __l);
-  using _CVT = __widen_from_utf8<sizeof(wchar_t) * __CHAR_BIT__>;
-  std::wstring __w;
-  __w.reserve(__tmp.size());
-  _CVT()(back_inserter(__w), __tmp.data(), __tmp.data() + __tmp.size());
-  return path(__w);
-#  else
-  return path(__f, __l);
-#  endif /* !_LIBCPP_WIN32API */
+#    if _LIBCPP_HAS_CHAR8_T
+        is_same<typename __is_pathable<_InputIt>::__char_type, char8_t>::value ||
+#    endif
+            is_same<typename __is_pathable<_InputIt>::__char_type, char>::value,
+        "u8path(Iter, Iter) requires Iter have a value_type of type 'char'"
+        " or 'char8_t'");
+#    if defined(_LIBCPP_WIN32API)
+    string __tmp(__f, __l);
+    using _CVT = __widen_from_utf8<sizeof(wchar_t) * __CHAR_BIT__>;
+    std::wstring __w;
+    __w.reserve(__tmp.size());
+    _CVT()(back_inserter(__w), __tmp.data(), __tmp.data() + __tmp.size());
+    return path(__w);
+#    else
+    return path(__f, __l);
+#    endif /* !_LIBCPP_WIN32API */
 }
+#  endif /* !_LIBCPP_WIN32API || _LIBCPP_HAS_LOCALIZATION */
 
-#  if defined(_LIBCPP_WIN32API)
+#  if defined(_LIBCPP_WIN32API) && _LIBCPP_HAS_LOCALIZATION
 template <class _InputIt, __enable_if_t<__is_pathable<_InputIt>::value, int> = 0>
 _LIBCPP_HIDE_FROM_ABI _LIBCPP_DEPRECATED_WITH_CHAR8_T path u8path(_InputIt __f, _NullSentinel) {
   static_assert(
@@ -65,7 +70,7 @@ _LIBCPP_HIDE_FROM_ABI _LIBCPP_DEPRECATED_WITH_CHAR8_T path u8path(_InputIt __f,
   _CVT()(back_inserter(__w), __tmp.data(), __tmp.data() + __tmp.size());
   return path(__w);
 }
-#  endif /* _LIBCPP_WIN32API */
+#  endif /* _LIBCPP_WIN32API && _LIBCPP_HAS_LOCALIZATION */
 
 template <class _Source, __enable_if_t<__is_pathable<_Source>::value, int> = 0>
 _LIBCPP_HIDE_FROM_ABI _LIBCPP_DEPRECATED_WITH_CHAR8_T path u8path(const _Source& __s) {
@@ -85,7 +90,6 @@ _LIBCPP_HIDE_FROM_ABI _LIBCPP_DEPRECATED_WITH_CHAR8_T path u8path(const _Source&
 }
 
 _LIBCPP_END_NAMESPACE_FILESYSTEM
-
 #endif // _LIBCPP_STD_VER >= 17
 
 #endif // _LIBCPP___FILESYSTEM_U8PATH_H

compnerd · 2025-10-28T21:03:30Z

libcxx/include/__filesystem/u8path.h


 _LIBCPP_BEGIN_NAMESPACE_FILESYSTEM

+#  if !defined(_LIBCPP_WIN32API) || _LIBCPP_HAS_LOCALIZATION


I wonder if this is better as a more complex change.

#if _LIBCPP_HAS_LOCALIZATION template <class _InputIt, __enable_if_t<__is_pathable<_InputIt>::value, int> = 0> _LIBCPP_HIDE_FROM_ABI _LIBCPP_DEPRECATED_WITH_CHAR8_T path u8path(_InputIt __f, #if _LIBCPP_WIN32API _InputIt __l #else _NullSentinel #endif ) { static_assert( # if _LIBCPP_HAS_CHAR8_T is_same<typename __is_pathable<_InputIt>::__char_type, char8_t>::value || # endif is_same<typename __is_pathable<_InputIt>::__char_type, char>::value, "u8path(Iter, Iter) requires Iter have a value_type of type 'char'" " or 'char8_t'"); #if _LIBCPP_WIN32API string __tmp(__f, __l); #else string __tmp; #endif const char __sentinel = char{}; for (; *__f != __sentinel; ++__f) __tmp.push_back(*__f); using _CVT = __widen_from_utf8<sizeof(wchar_t) * __CHAR_BIT__>; std::wstring __w; __w.reserve(__tmp.size()); _CVT()(back_inserter(__w), __tmp.data(), __tmp.data() + __tmp.size()); return path(__w); } #endif

Of course this needs to be clang-formatted, but that is the general idea.

That would cause a regression by disabling u8path on non Windows platforms where it works correctly without localization. fix is targeted to disable the function only on the broken configuration (Windows without localization), preserving functionality elsewhere

Okay, but the general idea still holds; we can guard it with #if !_LIBCPP_WIN32 || _LIBCPP_HAS_LOCALIZATION instead, and adjust the body as well.

#if !_LIBCPP_WIN32 || _LIBCPP_HAS_LOCALIZATION template <class _InputIt, __enable_if_t<__is_pathable<_InputIt>::value, int> = 0> _LIBCPP_HIDE_FROM_ABI _LIBCPP_DEPRECATED_WITH_CHAR8_T path u8path(_InputIt __f, #if _LIBCPP_HAS_LOCALIZATION _InputIt __l #else _NullSentinel #endif ) { static_assert( # if _LIBCPP_HAS_CHAR8_T is_same<typename __is_pathable<_InputIt>::__char_type, char8_t>::value || # endif is_same<typename __is_pathable<_InputIt>::__char_type, char>::value, "u8path(Iter, Iter) requires Iter have a value_type of type 'char'" " or 'char8_t'"); #if !_LIBCPP_HAS_LOCALIZATION return path(__f, __l); #else string __tmp(__f, __l); using _CVT = __widen_from_utf8<sizeof(wchar_t) * __CHAR_BIT__>; std::wstring __w; __w.reserve(__tmp.size()); _CVT()(back_inserter(__w), __tmp.data(), __tmp.data() + __tmp.size()); return path(__w); #endif } #endif

Sorry, doing this during the LLVM conference, so there might be more tweaks required.

I strongly think that it's unconforming to use the _NullSentinel parameter in the function of a standard name (u8path).

I think it's better to reform the code like this:

Collapse/Expand

_LIBCPP_BEGIN_NAMESPACE_FILESYSTEM # if defined(_LIBCPP_WIN32API) && _LIBCPP_HAS_LOCALIZATION # define _LIBCPP_FS_U8PATH_CONVERTS_ENCODING 1 # else # define _LIBCPP_FS_U8PATH_CONVERTS_ENCODING 0 # endif # if _LIBCPP_FS_U8PATH_CONVERTS_ENCODING template <class _InputIt, class _Sentinel> _LIBCPP_HIDE_FROM_ABI string __make_tmp_string_for_u8path(_InputIt __f, _Sentinel __l) { static_assert(__is_pathable<_InputIt>::value); static_assert( # if _LIBCPP_HAS_CHAR8_T is_same<typename __is_pathable<_InputIt>::__char_type, char8_t>::value || # endif is_same<typename __is_pathable<_InputIt>::__char_type, char>::value); if constexpr (is_same_v<_InputIt, _NullSentinel>) { string __tmp; constexpr char __sentinel{}; for (; *__f != __sentinel; ++__f) __tmp.push_back(*__f); return __tmp; } else { static_assert(is_same_v<_InputIt, _Sentinel>); return string(__f, __l); } } # endif // _LIBCPP_FS_U8PATH_CONVERTS_ENCODING template <class _InputIt, class _Sentinel> _LIBCPP_HIDE_FROM_ABI path __u8path(_InputIt __f, _Sentinel __l) { # if _LIBCPP_FS_U8PATH_CONVERTS_ENCODING auto __tmp = std::filesystem::__make_tmp_string_for_u8path(__f, __l); using _CVT = __widen_from_utf8<sizeof(wchar_t) * __CHAR_BIT__>; std::wstring __w; __w.reserve(__tmp.size()); _CVT()(back_inserter(__w), __tmp.data(), __tmp.data() + __tmp.size()); return path(__w); # else return path(__f, __l); # endif // _LIBCPP_FS_U8PATH_CONVERTS_ENCODING } template <class _InputIt, __enable_if_t<__is_pathable<_InputIt>::value, int> = 0> _LIBCPP_HIDE_FROM_ABI _LIBCPP_DEPRECATED_WITH_CHAR8_T path u8path(_InputIt __f, _InputIt __l) { static_assert( # if _LIBCPP_HAS_CHAR8_T is_same<typename __is_pathable<_InputIt>::__char_type, char8_t>::value || # endif is_same<typename __is_pathable<_InputIt>::__char_type, char>::value, "u8path(Iter, Iter) requires Iter have a value_type of type 'char'" " or 'char8_t'"); return std::filesystem::__u8path(__f, __l); } template <class _Source, __enable_if_t<__is_pathable<_Source>::value, int> = 0> _LIBCPP_HIDE_FROM_ABI _LIBCPP_DEPRECATED_WITH_CHAR8_T path u8path(const _Source& __s) { static_assert( # if _LIBCPP_HAS_CHAR8_T is_same<typename __is_pathable<_Source>::__char_type, char8_t>::value || # endif is_same<typename __is_pathable<_Source>::__char_type, char>::value, "u8path(Source const&) requires Source have a character type of type " "'char' or 'char8_t'"); # if _LIBCPP_FS_U8PATH_CONVERTS_ENCODING using _Traits = __is_pathable<_Source>; return std::filesystem::__u8path( std::__unwrap_iter(_Traits::__range_begin(__s)), std::__unwrap_iter(_Traits::__range_end(__s))); # else return path(__s); # endif // _LIBCPP_FS_U8PATH_CONVERTS_ENCODING } # undef _LIBCPP_FS_U8PATH_CONVERTS_ENCODING _LIBCPP_END_NAMESPACE_FILESYSTEM

@frederick-vs-ja Thank you for the suggestion. This is much cleaner. I'll update the PR with your proposed implementation.

github-actions · 2025-10-29T13:32:36Z

⚠️ C/C++ code formatter, clang-format found issues in your code. ⚠️

You can test this locally with the following command:

git-clang-format --diff origin/main HEAD --extensions h -- libcxx/include/__filesystem/path.h libcxx/include/__filesystem/u8path.h --diff_from_common_commit

⚠️
The reproduction instructions above might return results for more than one PR
in a stack if you are using a stacked PR workflow. You can limit the results by
changing origin/main to the base branch/commit you want to compare against.
⚠️

View the diff from clang-format here.

diff --git a/libcxx/include/__filesystem/path.h b/libcxx/include/__filesystem/path.h
index 990833116..fc27414ea 100644
--- a/libcxx/include/__filesystem/path.h
+++ b/libcxx/include/__filesystem/path.h
@@ -376,7 +376,7 @@ struct _PathExport<char8_t> {
   }
 };
 #      endif // _LIBCPP_HAS_CHAR8_T
-#    endif // _LIBCPP_HAS_LOCALIZATION
+#    endif   // _LIBCPP_HAS_LOCALIZATION
 #  endif   /* _LIBCPP_WIN32API */
 
 class _LIBCPP_EXPORTED_FROM_ABI path {

Co-authored-by: Hxnter999 <Hxnter999@users.noreply.github.com>

[libcxx] proper guarding for locale usage in filesystem on Windows

005ac72

- Resolves build issues when localization support is disabled on Windows. - Resolves dependencies on localization in filesystem header implementations. Related PR llvm#164602 Fixes llvm#164074

moleium requested a review from a team as a code owner October 28, 2025 20:06

llvmbot added the libc++ libc++ C++ Standard Library. Not GNU libstdc++. Not libc++abi. label Oct 28, 2025

compnerd added the platform:windows label Oct 28, 2025

compnerd reviewed Oct 28, 2025

View reviewed changes

[libcxx] Fix indentation for macros in filesystem header

9d49d76

Co-authored-by: Hxnter999 <Hxnter999@users.noreply.github.com>

moleium force-pushed the libcxx-freestanding-fs-guards branch from de27e34 to 9d49d76 Compare October 29, 2025 16:02

ldionne changed the title ~~[libcxx] proper guarding for locale usage in filesystem on Windows~~ [libc++] proper guarding for locale usage in filesystem on Windows Oct 31, 2025

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[libc++] proper guarding for locale usage in filesystem on Windows #165470

[libc++] proper guarding for locale usage in filesystem on Windows #165470

moleium commented Oct 28, 2025

Uh oh!

github-actions bot commented Oct 28, 2025

Uh oh!

llvmbot commented Oct 28, 2025 •

edited

Loading

Uh oh!

compnerd Oct 28, 2025

Uh oh!

moleium Oct 28, 2025

Uh oh!

compnerd Oct 28, 2025

Uh oh!

frederick-vs-ja Oct 29, 2025

Uh oh!

moleium Oct 29, 2025

Uh oh!

github-actions bot commented Oct 29, 2025 •

edited

Loading

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants


		_LIBCPP_BEGIN_NAMESPACE_FILESYSTEM

		# if !defined(_LIBCPP_WIN32API) \|\| _LIBCPP_HAS_LOCALIZATION

[libc++] proper guarding for locale usage in filesystem on Windows #165470

Are you sure you want to change the base?

[libc++] proper guarding for locale usage in filesystem on Windows #165470

Conversation

moleium commented Oct 28, 2025

Uh oh!

github-actions bot commented Oct 28, 2025

Uh oh!

llvmbot commented Oct 28, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

compnerd Oct 28, 2025

Choose a reason for hiding this comment

Uh oh!

moleium Oct 28, 2025

Choose a reason for hiding this comment

Uh oh!

compnerd Oct 28, 2025

Choose a reason for hiding this comment

Uh oh!

frederick-vs-ja Oct 29, 2025

Choose a reason for hiding this comment

Uh oh!

moleium Oct 29, 2025

Choose a reason for hiding this comment

Uh oh!

github-actions bot commented Oct 29, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

llvmbot commented Oct 28, 2025 •

edited

Loading

github-actions bot commented Oct 29, 2025 •

edited

Loading