[libc] Implemented wcspbrk #142040

uzairnawaz · 2025-05-29T21:19:17Z

Implemented wcspbrk and added tests

llvmbot · 2025-05-29T21:19:53Z

@llvm/pr-subscribers-libc

Author: Uzair Nawaz (uzairnawaz)

Changes

Implemented wcspbrk and added tests

Full diff: https://github.com/llvm/llvm-project/pull/142040.diff

7 Files Affected:

(modified) libc/config/linux/x86_64/entrypoints.txt (+1)
(modified) libc/include/wchar.yaml (+7)
(modified) libc/src/wchar/CMakeLists.txt (+11)
(added) libc/src/wchar/wcspbrk.cpp (+27)
(added) libc/src/wchar/wcspbrk.h (+21)
(modified) libc/test/src/wchar/CMakeLists.txt (+10)
(added) libc/test/src/wchar/wcspbrk_test.cpp (+83)

diff --git a/libc/config/linux/x86_64/entrypoints.txt b/libc/config/linux/x86_64/entrypoints.txt
index 52e746e32a1cd..593117f94b64e 100644
--- a/libc/config/linux/x86_64/entrypoints.txt
+++ b/libc/config/linux/x86_64/entrypoints.txt
@@ -365,6 +365,7 @@ set(TARGET_LIBC_ENTRYPOINTS
     libc.src.wchar.wcslen
     libc.src.wchar.wctob
     libc.src.wchar.wcschr
+    libc.src.wchar.wcspbrk
 
     # sys/uio.h entrypoints
     libc.src.sys.uio.writev
diff --git a/libc/include/wchar.yaml b/libc/include/wchar.yaml
index 987bdc0b806dc..52ebc6804d043 100644
--- a/libc/include/wchar.yaml
+++ b/libc/include/wchar.yaml
@@ -34,3 +34,10 @@ functions:
     arguments: 
       - type: const wchar_t *
       - type: wchar_t
+  - name: wcspbrk
+    standards:
+      - stdc
+    return_type: const wchar_t *
+    arguments:
+      - type: const wchar_t *
+      - type: const wchar_t *
diff --git a/libc/src/wchar/CMakeLists.txt b/libc/src/wchar/CMakeLists.txt
index 41034cab16d4d..4d923c5f14474 100644
--- a/libc/src/wchar/CMakeLists.txt
+++ b/libc/src/wchar/CMakeLists.txt
@@ -44,3 +44,14 @@ add_entrypoint_object(
     libc.hdr.wchar_macros
     libc.src.__support.wctype_utils
 )
+
+add_entrypoint_object(
+  wcspbrk
+  SRCS
+    wcspbrk.cpp
+  HDRS
+    wcspbrk.h
+  DEPENDS
+    libc.hdr.wchar_macros
+    libc.src.__support.wctype_utils
+)
diff --git a/libc/src/wchar/wcspbrk.cpp b/libc/src/wchar/wcspbrk.cpp
new file mode 100644
index 0000000000000..e92be0853a5c0
--- /dev/null
+++ b/libc/src/wchar/wcspbrk.cpp
@@ -0,0 +1,27 @@
+//===-- Implementation of wcspbrk -----------------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#include "src/wchar/wcspbrk.h"
+
+#include "hdr/types/wchar_t.h"
+#include "src/__support/common.h"
+
+namespace LIBC_NAMESPACE_DECL {
+
+LLVM_LIBC_FUNCTION(const wchar_t *, wcspbrk,
+                   (const wchar_t *src, const wchar_t *breakset)) {
+  // currently O(n * m), can be further optimized to O(n + m) with a hash set
+  for (int src_idx = 0; src[src_idx] != 0; src_idx++)
+    for (int breakset_idx = 0; breakset[breakset_idx] != 0; breakset_idx++)
+      if (src[src_idx] == breakset[breakset_idx])
+        return src + src_idx;
+
+  return nullptr;
+}
+
+} // namespace LIBC_NAMESPACE_DECL
diff --git a/libc/src/wchar/wcspbrk.h b/libc/src/wchar/wcspbrk.h
new file mode 100644
index 0000000000000..531651b0b723a
--- /dev/null
+++ b/libc/src/wchar/wcspbrk.h
@@ -0,0 +1,21 @@
+//===-- Implementation header for wcspbrk ---------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_LIBC_SRC_WCHAR_WCSPBRK_H
+#define LLVM_LIBC_SRC_WCHAR_WCSPBRK_H
+
+#include "hdr/types/wchar_t.h"
+#include "src/__support/macros/config.h"
+
+namespace LIBC_NAMESPACE_DECL {
+
+const wchar_t *wcspbrk(const wchar_t *src, const wchar_t *breakset);
+
+} // namespace LIBC_NAMESPACE_DECL
+
+#endif // LLVM_LIBC_SRC_WCHAR_WCSPBRK_H
diff --git a/libc/test/src/wchar/CMakeLists.txt b/libc/test/src/wchar/CMakeLists.txt
index 7d64dfeb13b6e..0e3d5c618ae1a 100644
--- a/libc/test/src/wchar/CMakeLists.txt
+++ b/libc/test/src/wchar/CMakeLists.txt
@@ -42,3 +42,13 @@ add_libc_test(
   DEPENDS
     libc.src.wchar.wcschr
 )
+
+add_libc_test(
+  wcspbrk_test
+  SUITE
+    libc_wchar_unittests
+  SRCS
+    wcspbrk_test.cpp
+  DEPENDS
+    libc.src.wchar.wcspbrk
+)
diff --git a/libc/test/src/wchar/wcspbrk_test.cpp b/libc/test/src/wchar/wcspbrk_test.cpp
new file mode 100644
index 0000000000000..ac7ac473542d9
--- /dev/null
+++ b/libc/test/src/wchar/wcspbrk_test.cpp
@@ -0,0 +1,83 @@
+//===-- Unittests for wcspbrk
+//----------------------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#include "hdr/types/wchar_t.h"
+#include "src/wchar/wcspbrk.h"
+#include "test/UnitTest/Test.h"
+
+TEST(LlvmLibcWCSPBrkTest, EmptyStringShouldReturnNullptr) {
+  // The search should not include the null terminator.
+  EXPECT_EQ(LIBC_NAMESPACE::wcspbrk(L"", L""), nullptr);
+  EXPECT_EQ(LIBC_NAMESPACE::wcspbrk(L"_", L""), nullptr);
+  EXPECT_EQ(LIBC_NAMESPACE::wcspbrk(L"", L"_"), nullptr);
+}
+
+TEST(LlvmLibcWCSPBrkTest, ShouldNotFindAnythingAfterNullTerminator) {
+  const wchar_t src[4] = {'a', 'b', '\0', 'c'};
+  EXPECT_EQ(LIBC_NAMESPACE::wcspbrk(src, L"c"), nullptr);
+}
+
+TEST(LlvmLibcWCSPBrkTest, ShouldReturnNullptrIfNoCharactersFound) {
+  EXPECT_EQ(LIBC_NAMESPACE::wcspbrk(L"12345", L"abcdef"), nullptr);
+}
+
+TEST(LlvmLibcWCSPBrkTest, FindsFirstCharacter) {
+  const wchar_t *src = L"12345";
+  EXPECT_EQ(LIBC_NAMESPACE::wcspbrk(src, L"1"), src);
+  EXPECT_EQ(LIBC_NAMESPACE::wcspbrk(src, L"-1"), src);
+  EXPECT_EQ(LIBC_NAMESPACE::wcspbrk(src, L"1_"), src);
+  EXPECT_EQ(LIBC_NAMESPACE::wcspbrk(src, L"f1_"), src);
+
+  EXPECT_TRUE(src[0] == L'1');
+  EXPECT_TRUE(src[1] == L'2');
+  EXPECT_TRUE(src[2] == L'3');
+  EXPECT_TRUE(src[3] == L'4');
+  EXPECT_TRUE(src[4] == L'5');
+  EXPECT_TRUE(src[5] == 0);
+}
+
+TEST(LlvmLibcWCSPBrkTest, FindsMiddleCharacter) {
+  const wchar_t *src = L"12345";
+  EXPECT_EQ(LIBC_NAMESPACE::wcspbrk(src, L"3"), src + 2);
+  EXPECT_EQ(LIBC_NAMESPACE::wcspbrk(src, L"?3"), src + 2);
+  EXPECT_EQ(LIBC_NAMESPACE::wcspbrk(src, L"3F"), src + 2);
+  EXPECT_EQ(LIBC_NAMESPACE::wcspbrk(src, L"z3_"), src + 2);
+
+  EXPECT_TRUE(src[0] == L'1');
+  EXPECT_TRUE(src[1] == L'2');
+  EXPECT_TRUE(src[2] == L'3');
+  EXPECT_TRUE(src[3] == L'4');
+  EXPECT_TRUE(src[4] == L'5');
+  EXPECT_TRUE(src[5] == 0);
+}
+
+TEST(LlvmLibcWCSPBrkTest, FindsLastCharacter) {
+  const wchar_t *src = L"12345";
+  EXPECT_EQ(LIBC_NAMESPACE::wcspbrk(src, L"5"), src + 4);
+  EXPECT_EQ(LIBC_NAMESPACE::wcspbrk(src, L"r5"), src + 4);
+  EXPECT_EQ(LIBC_NAMESPACE::wcspbrk(src, L"59"), src + 4);
+  EXPECT_EQ(LIBC_NAMESPACE::wcspbrk(src, L"n5_"), src + 4);
+
+  EXPECT_TRUE(src[0] == L'1');
+  EXPECT_TRUE(src[1] == L'2');
+  EXPECT_TRUE(src[2] == L'3');
+  EXPECT_TRUE(src[3] == L'4');
+  EXPECT_TRUE(src[4] == L'5');
+  EXPECT_TRUE(src[5] == 0);
+}
+
+TEST(LlvmLibcWCSPBrkTest, FindsFirstOfRepeated) {
+  const wchar_t *src = L"A,B,C,D";
+  EXPECT_EQ(LIBC_NAMESPACE::wcspbrk(src, L","), src + 1);
+}
+
+TEST(LlvmLibcWCSPBrkTest, FindsFirstInBreakset) {
+  const wchar_t *src = L"12345";
+  EXPECT_EQ(LIBC_NAMESPACE::wcspbrk(src, L"34"), src + 2);
+}

michaelrj-google

mostly LGTM with a few comments on the tests

michaelrj-google · 2025-05-29T22:57:51Z

libc/test/src/wchar/wcspbrk_test.cpp

+  EXPECT_TRUE(src[0] == L'1');
+  EXPECT_TRUE(src[1] == L'2');
+  EXPECT_TRUE(src[2] == L'3');
+  EXPECT_TRUE(src[3] == L'4');
+  EXPECT_TRUE(src[4] == L'5');
+  EXPECT_TRUE(src[5] == 0);


since src isn't getting modified I don't think this check is necessary.

michaelrj-google · 2025-05-29T22:59:09Z

libc/test/src/wchar/wcspbrk_test.cpp

+TEST(LlvmLibcWCSPBrkTest, FindsFirstInBreakset) {
+  const wchar_t *src = L"12345";
+  EXPECT_EQ(LIBC_NAMESPACE::wcspbrk(src, L"34"), src + 2);
+}


it might be useful to also check that it works correctly when the breakset is not in the same order as the string being searched.

michaelrj-google · 2025-05-30T17:46:05Z

libc/test/src/wchar/wcspbrk_test.cpp

@@ -0,0 +1,63 @@
+//===-- Unittests for wcspbrk


nit: fix formatting

michaelrj-google · 2025-05-30T17:46:56Z

libc/src/wchar/wcspbrk.cpp

+    for (int breakset_idx = 0; breakset[breakset_idx] != 0; breakset_idx++)
+      if (src[src_idx] == breakset[breakset_idx])
+        return src + src_idx;


if you break the check if src[src_idx] is in breakset out into a helper function this will be easier to read

michaelrj-google

LGTM with one nit

michaelrj-google · 2025-05-30T18:10:40Z

libc/src/wchar/wcspbrk.cpp

@@ -13,13 +13,20 @@

 namespace LIBC_NAMESPACE_DECL {

+bool contains_char(const wchar_t *str, wchar_t target) {
+  for (; *str != 0; str++)


since this is a wide char null byte this should be L'\0' instead of 0.

michaelrj-google

LGTM, we can add nullptr checks in a followup patch

implemented wcspbrk; added tests

7dd7eda

llvmbot added the libc label May 29, 2025

Merge branch 'main' into wcspbrk

1729afa

michaelrj-google reviewed May 29, 2025

View reviewed changes

updated tests

379e86e

michaelrj-google reviewed May 30, 2025

View reviewed changes

created helper function for test; fixed formatting

d787008

michaelrj-google reviewed May 30, 2025

View reviewed changes

fixed loop condition in contains_char

85dec85

michaelrj-google approved these changes May 30, 2025

View reviewed changes

Merge branch 'main' into wcspbrk

d07884b

michaelrj-google merged commit d721d4e into llvm:main May 30, 2025
14 of 15 checks passed

michaelrj-google mentioned this pull request May 30, 2025

[libc][wchar] implement wcspbrk #124346

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[libc] Implemented wcspbrk #142040

[libc] Implemented wcspbrk #142040

Uh oh!

uzairnawaz commented May 29, 2025

Uh oh!

llvmbot commented May 29, 2025

Uh oh!

michaelrj-google left a comment

Uh oh!

michaelrj-google May 29, 2025

Uh oh!

michaelrj-google May 29, 2025

Uh oh!

michaelrj-google May 30, 2025

Uh oh!

michaelrj-google May 30, 2025

Uh oh!

michaelrj-google left a comment

Uh oh!

michaelrj-google May 30, 2025

Uh oh!

michaelrj-google left a comment

Uh oh!

Uh oh!

Uh oh!

[libc] Implemented wcspbrk #142040

[libc] Implemented wcspbrk #142040

Uh oh!

Conversation

uzairnawaz commented May 29, 2025

Uh oh!

llvmbot commented May 29, 2025

Uh oh!

michaelrj-google left a comment

Choose a reason for hiding this comment

Uh oh!

michaelrj-google May 29, 2025

Choose a reason for hiding this comment

Uh oh!

michaelrj-google May 29, 2025

Choose a reason for hiding this comment

Uh oh!

michaelrj-google May 30, 2025

Choose a reason for hiding this comment

Uh oh!

michaelrj-google May 30, 2025

Choose a reason for hiding this comment

Uh oh!

michaelrj-google left a comment

Choose a reason for hiding this comment

Uh oh!

michaelrj-google May 30, 2025

Choose a reason for hiding this comment

Uh oh!

michaelrj-google left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!