[libc++] <regex>: Make unmatched backrefs always succeed in ECMAScript mode #154442

SainoNamkho · 2025-08-19T23:44:16Z

…MAScript mode. Fix llvm#154408

github-actions · 2025-08-19T23:44:33Z

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 · 2025-08-19T23:45:11Z

@llvm/pr-subscribers-libcxx

Author: None (SainoNamkho)

Changes

Fix #154408

Full diff: https://github.com/llvm/llvm-project/pull/154442.diff

2 Files Affected:

(modified) libcxx/include/regex (+28-13)
(modified) libcxx/test/std/re/re.alg/re.alg.search/ecma.pass.cpp (+30)

diff --git a/libcxx/include/regex b/libcxx/include/regex
index 9bbc3a69021b9..ba69f5681a5d4 100644
--- a/libcxx/include/regex
+++ b/libcxx/include/regex
@@ -1668,7 +1668,7 @@ void __end_marked_subexpression<_CharT>::__exec(__state& __s) const {
 
 // __back_ref
 
-template <class _CharT>
+template <class _CharT, bool _UnmatchedAlwaysSucceed>
 class __back_ref : public __owns_one_state<_CharT> {
   typedef __owns_one_state<_CharT> base;
 
@@ -1682,8 +1682,8 @@ public:
   _LIBCPP_HIDE_FROM_ABI_VIRTUAL virtual void __exec(__state&) const;
 };
 
-template <class _CharT>
-void __back_ref<_CharT>::__exec(__state& __s) const {
+template <class _CharT, bool _UnmatchedAlwaysSucceed>
+void __back_ref<_CharT, _UnmatchedAlwaysSucceed>::__exec(__state& __s) const {
   if (__mexp_ > __s.__sub_matches_.size())
     std::__throw_regex_error<regex_constants::error_backref>();
   sub_match<const _CharT*>& __sm = __s.__sub_matches_[__mexp_ - 1];
@@ -1697,6 +1697,9 @@ void __back_ref<_CharT>::__exec(__state& __s) const {
       __s.__do_   = __state::__reject;
       __s.__node_ = nullptr;
     }
+  } else if constexpr (_UnmatchedAlwaysSucceed) {
+    __s.__do_   = __state::__accept_but_not_consume;
+    __s.__node_ = this->first();;
   } else {
     __s.__do_   = __state::__reject;
     __s.__node_ = nullptr;
@@ -1705,7 +1708,7 @@ void __back_ref<_CharT>::__exec(__state& __s) const {
 
 // __back_ref_icase
 
-template <class _CharT, class _Traits>
+template <class _CharT, class _Traits, bool _UnmatchedAlwaysSucceed>
 class __back_ref_icase : public __owns_one_state<_CharT> {
   typedef __owns_one_state<_CharT> base;
 
@@ -1721,8 +1724,8 @@ public:
   _LIBCPP_HIDE_FROM_ABI_VIRTUAL virtual void __exec(__state&) const;
 };
 
-template <class _CharT, class _Traits>
-void __back_ref_icase<_CharT, _Traits>::__exec(__state& __s) const {
+template <class _CharT, class _Traits, bool _UnmatchedAlwaysSucceed>
+void __back_ref_icase<_CharT, _Traits, _UnmatchedAlwaysSucceed>::__exec(__state& __s) const {
   sub_match<const _CharT*>& __sm = __s.__sub_matches_[__mexp_ - 1];
   if (__sm.matched) {
     ptrdiff_t __len = __sm.second - __sm.first;
@@ -1739,6 +1742,11 @@ void __back_ref_icase<_CharT, _Traits>::__exec(__state& __s) const {
       __s.__node_ = nullptr;
     }
   } else {
+    if constexpr (_UnmatchedAlwaysSucceed) {
+      __s.__do_   = __state::__accept_but_not_consume;
+      __s.__node_ = this->first();
+      return;
+    }
   __not_equal:
     __s.__do_   = __state::__reject;
     __s.__node_ = nullptr;
@@ -1747,7 +1755,7 @@ void __back_ref_icase<_CharT, _Traits>::__exec(__state& __s) const {
 
 // __back_ref_collate
 
-template <class _CharT, class _Traits>
+template <class _CharT, class _Traits, bool _UnmatchedAlwaysSucceed>
 class __back_ref_collate : public __owns_one_state<_CharT> {
   typedef __owns_one_state<_CharT> base;
 
@@ -1763,8 +1771,8 @@ public:
   _LIBCPP_HIDE_FROM_ABI_VIRTUAL virtual void __exec(__state&) const;
 };
 
-template <class _CharT, class _Traits>
-void __back_ref_collate<_CharT, _Traits>::__exec(__state& __s) const {
+template <class _CharT, class _Traits, bool _UnmatchedAlwaysSucceed>
+void __back_ref_collate<_CharT, _Traits, _UnmatchedAlwaysSucceed>::__exec(__state& __s) const {
   sub_match<const _CharT*>& __sm = __s.__sub_matches_[__mexp_ - 1];
   if (__sm.matched) {
     ptrdiff_t __len = __sm.second - __sm.first;
@@ -1781,6 +1789,11 @@ void __back_ref_collate<_CharT, _Traits>::__exec(__state& __s) const {
       __s.__node_ = nullptr;
     }
   } else {
+    if constexpr (_UnmatchedAlwaysSucceed) {
+      __s.__do_   = __state::__accept_but_not_consume;
+      __s.__node_ = this->first();
+      return;
+    }
   __not_equal:
     __s.__do_   = __state::__reject;
     __s.__node_ = nullptr;
@@ -2565,6 +2578,7 @@ private:
                    bool __greedy       = true);
   __bracket_expression<_CharT, _Traits>* __start_matching_list(bool __negate);
   void __push_char(value_type __c);
+  template <bool _UnmatchedAlwaysSucceed = false>
   void __push_back_ref(int __i);
   void __push_alternation(__owns_one_state<_CharT>* __sa, __owns_one_state<_CharT>* __sb);
   void __push_begin_marked_subexpression();
@@ -3807,7 +3821,7 @@ basic_regex<_CharT, _Traits>::__parse_decimal_escape(_ForwardIterator __first, _
       }
       if (__v == 0 || __v > mark_count())
         std::__throw_regex_error<regex_constants::error_backref>();
-      __push_back_ref(__v);
+      __push_back_ref<true>(__v);
     }
   }
   return __first;
@@ -4149,13 +4163,14 @@ void basic_regex<_CharT, _Traits>::__push_word_boundary(bool __invert) {
 }
 
 template <class _CharT, class _Traits>
+template <bool _UnmatchedAlwaysSucceed>
 void basic_regex<_CharT, _Traits>::__push_back_ref(int __i) {
   if (flags() & icase)
-    __end_->first() = new __back_ref_icase<_CharT, _Traits>(__traits_, __i, __end_->first());
+    __end_->first() = new __back_ref_icase<_CharT, _Traits, _UnmatchedAlwaysSucceed>(__traits_, __i, __end_->first());
   else if (flags() & collate)
-    __end_->first() = new __back_ref_collate<_CharT, _Traits>(__traits_, __i, __end_->first());
+    __end_->first() = new __back_ref_collate<_CharT, _Traits, _UnmatchedAlwaysSucceed>(__traits_, __i, __end_->first());
   else
-    __end_->first() = new __back_ref<_CharT>(__i, __end_->first());
+    __end_->first() = new __back_ref<_CharT, _UnmatchedAlwaysSucceed>(__i, __end_->first());
   __end_ = static_cast<__owns_one_state<_CharT>*>(__end_->first());
 }
 
diff --git a/libcxx/test/std/re/re.alg/re.alg.search/ecma.pass.cpp b/libcxx/test/std/re/re.alg/re.alg.search/ecma.pass.cpp
index 518c27e424484..a6a1a9cf87c24 100644
--- a/libcxx/test/std/re/re.alg/re.alg.search/ecma.pass.cpp
+++ b/libcxx/test/std/re/re.alg/re.alg.search/ecma.pass.cpp
@@ -762,6 +762,21 @@ int main(int, char**)
         assert(m.position(0) == 0);
         assert(m.str(0) == s);
     }
+    {
+        std::cmatch m;
+        const char s[] = "a";
+        assert(std::regex_search(s, m, std::regex{"(a()|)\\2a"}));
+        assert(m.size() == 3);
+        assert(!m.prefix().matched);
+        assert(m.prefix().first == s);
+        assert(m.prefix().second == m[0].first);
+        assert(!m.suffix().matched);
+        assert(m.suffix().first == m[0].second);
+        assert(m.suffix().second == s + std::char_traits<char>::length(s));
+        assert(m.length(0) >= 0 && static_cast<std::size_t>(m.length(0)) == std::char_traits<char>::length(s));
+        assert(m.position(0) == 0);
+        assert(m.str(0) == s);
+    }
 
 #ifndef TEST_HAS_NO_WIDE_CHARACTERS
     {
@@ -1503,6 +1518,21 @@ int main(int, char**)
         assert(m.position(0) == 0);
         assert(m.str(0) == s);
     }
+    {
+        std::wcmatch m;
+        const wchar_t s[] = L"a";
+        assert(std::regex_search(s, m, std::wregex{L"(a()|)\\2a"}));
+        assert(m.size() == 3);
+        assert(!m.prefix().matched);
+        assert(m.prefix().first == s);
+        assert(m.prefix().second == m[0].first);
+        assert(!m.suffix().matched);
+        assert(m.suffix().first == m[0].second);
+        assert(m.suffix().second == s + std::char_traits<wchar_t>::length(s));
+        assert(m.length(0) >= 0 && static_cast<std::size_t>(m.length(0)) == std::char_traits<wchar_t>::length(s));
+        assert(m.position(0) == 0);
+        assert(m.str(0) == s);
+    }
 #endif // TEST_HAS_NO_WIDE_CHARACTERS
 
   return 0;

github-actions · 2025-08-20T01:09:32Z

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

You can test this locally with the following command:

git-clang-format --diff HEAD~1 HEAD --extensions ,cpp -- libcxx/include/regex libcxx/test/std/re/re.alg/re.alg.search/ecma.pass.cpp

View the diff from clang-format here.

diff --git a/libcxx/include/regex b/libcxx/include/regex
index ba69f5681..205ccd625 100644
--- a/libcxx/include/regex
+++ b/libcxx/include/regex
@@ -1699,7 +1699,8 @@ void __back_ref<_CharT, _UnmatchedAlwaysSucceed>::__exec(__state& __s) const {
     }
   } else if constexpr (_UnmatchedAlwaysSucceed) {
     __s.__do_   = __state::__accept_but_not_consume;
-    __s.__node_ = this->first();;
+    __s.__node_ = this->first();
+    ;
   } else {
     __s.__do_   = __state::__reject;
     __s.__node_ = nullptr;
diff --git a/libcxx/test/std/re/re.alg/re.alg.search/ecma.pass.cpp b/libcxx/test/std/re/re.alg/re.alg.search/ecma.pass.cpp
index a6a1a9cf8..89273acc1 100644
--- a/libcxx/test/std/re/re.alg/re.alg.search/ecma.pass.cpp
+++ b/libcxx/test/std/re/re.alg/re.alg.search/ecma.pass.cpp
@@ -763,19 +763,19 @@ int main(int, char**)
         assert(m.str(0) == s);
     }
     {
-        std::cmatch m;
-        const char s[] = "a";
-        assert(std::regex_search(s, m, std::regex{"(a()|)\\2a"}));
-        assert(m.size() == 3);
-        assert(!m.prefix().matched);
-        assert(m.prefix().first == s);
-        assert(m.prefix().second == m[0].first);
-        assert(!m.suffix().matched);
-        assert(m.suffix().first == m[0].second);
-        assert(m.suffix().second == s + std::char_traits<char>::length(s));
-        assert(m.length(0) >= 0 && static_cast<std::size_t>(m.length(0)) == std::char_traits<char>::length(s));
-        assert(m.position(0) == 0);
-        assert(m.str(0) == s);
+      std::cmatch m;
+      const char s[] = "a";
+      assert(std::regex_search(s, m, std::regex{"(a()|)\\2a"}));
+      assert(m.size() == 3);
+      assert(!m.prefix().matched);
+      assert(m.prefix().first == s);
+      assert(m.prefix().second == m[0].first);
+      assert(!m.suffix().matched);
+      assert(m.suffix().first == m[0].second);
+      assert(m.suffix().second == s + std::char_traits<char>::length(s));
+      assert(m.length(0) >= 0 && static_cast<std::size_t>(m.length(0)) == std::char_traits<char>::length(s));
+      assert(m.position(0) == 0);
+      assert(m.str(0) == s);
     }
 
 #ifndef TEST_HAS_NO_WIDE_CHARACTERS
@@ -1519,19 +1519,19 @@ int main(int, char**)
         assert(m.str(0) == s);
     }
     {
-        std::wcmatch m;
-        const wchar_t s[] = L"a";
-        assert(std::regex_search(s, m, std::wregex{L"(a()|)\\2a"}));
-        assert(m.size() == 3);
-        assert(!m.prefix().matched);
-        assert(m.prefix().first == s);
-        assert(m.prefix().second == m[0].first);
-        assert(!m.suffix().matched);
-        assert(m.suffix().first == m[0].second);
-        assert(m.suffix().second == s + std::char_traits<wchar_t>::length(s));
-        assert(m.length(0) >= 0 && static_cast<std::size_t>(m.length(0)) == std::char_traits<wchar_t>::length(s));
-        assert(m.position(0) == 0);
-        assert(m.str(0) == s);
+      std::wcmatch m;
+      const wchar_t s[] = L"a";
+      assert(std::regex_search(s, m, std::wregex{L"(a()|)\\2a"}));
+      assert(m.size() == 3);
+      assert(!m.prefix().matched);
+      assert(m.prefix().first == s);
+      assert(m.prefix().second == m[0].first);
+      assert(!m.suffix().matched);
+      assert(m.suffix().first == m[0].second);
+      assert(m.suffix().second == s + std::char_traits<wchar_t>::length(s));
+      assert(m.length(0) >= 0 && static_cast<std::size_t>(m.length(0)) == std::char_traits<wchar_t>::length(s));
+      assert(m.position(0) == 0);
+      assert(m.str(0) == s);
     }
 #endif // TEST_HAS_NO_WIDE_CHARACTERS

frederick-vs-ja · 2025-08-20T01:18:27Z

libcxx/include/regex

    }
+  } else if constexpr (_UnmatchedAlwaysSucceed) {
+    __s.__do_   = __state::__accept_but_not_consume;
+    __s.__node_ = this->first();;


Repeated semicolon. (clang-format didn't suggest to remove it, but it's better to do so.)

Suggested change

__s.__node_ = this->first();;

__s.__node_ = this->first();

frederick-vs-ja · 2025-08-20T01:19:11Z

libcxx/test/std/re/re.alg/re.alg.search/ecma.pass.cpp

+    {
+        std::cmatch m;
+        const char s[] = "a";
+        assert(std::regex_search(s, m, std::regex{"(a()|)\\2a"}));


This test also runs in C++03 mode. Ditto below.

Suggested change

assert(std::regex_search(s, m, std::regex{"(a()|)\\2a"}));

assert(std::regex_search(s, m, std::regex("(a()|)\\2a")));

[libc++] <regex>: Make unmatched backrefs should always succeed in EC…

d4f9768

…MAScript mode. Fix llvm#154408

SainoNamkho requested a review from a team as a code owner August 19, 2025 23:44

llvmbot added the libc++ libc++ C++ Standard Library. Not GNU libstdc++. Not libc++abi. label Aug 19, 2025

SainoNamkho changed the title ~~[libc++] <regex>: Make unmatched backrefs should always succeed in ECMAScript mode~~ [libc++] <regex>: Make unmatched backrefs always succeed in ECMAScript mode Aug 19, 2025

Merge branch 'main' into fix-154408

86882b6

frederick-vs-ja reviewed Aug 20, 2025

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[libc++] <regex>: Make unmatched backrefs always succeed in ECMAScript mode #154442

[libc++] <regex>: Make unmatched backrefs always succeed in ECMAScript mode #154442

SainoNamkho commented Aug 19, 2025

Uh oh!

github-actions bot commented Aug 19, 2025

Uh oh!

llvmbot commented Aug 19, 2025

Uh oh!

github-actions bot commented Aug 20, 2025

Uh oh!

frederick-vs-ja Aug 20, 2025

Uh oh!

frederick-vs-ja Aug 20, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

	assert(std::regex_search(s, m, std::regex{"(a()\|)\\2a"}));
	assert(std::regex_search(s, m, std::regex("(a()\|)\\2a")));

[libc++] <regex>: Make unmatched backrefs always succeed in ECMAScript mode #154442

Are you sure you want to change the base?

[libc++] <regex>: Make unmatched backrefs always succeed in ECMAScript mode #154442

Conversation

SainoNamkho commented Aug 19, 2025

Uh oh!

github-actions bot commented Aug 19, 2025

Uh oh!

llvmbot commented Aug 19, 2025

Uh oh!

github-actions bot commented Aug 20, 2025

Uh oh!

frederick-vs-ja Aug 20, 2025

Choose a reason for hiding this comment

Uh oh!

frederick-vs-ja Aug 20, 2025

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants