[libc-commits] [libc] bfa88a8 - [libc] Implement wcscoll (#192778)
via libc-commits
libc-commits at lists.llvm.org
Fri Apr 24 13:54:30 PDT 2026
Author: Hardik Chona
Date: 2026-04-24T13:54:26-07:00
New Revision: bfa88a8d3c3b8885c9e0166afb1c78ede4f24c6c
URL: https://github.com/llvm/llvm-project/commit/bfa88a8d3c3b8885c9e0166afb1c78ede4f24c6c
DIFF: https://github.com/llvm/llvm-project/commit/bfa88a8d3c3b8885c9e0166afb1c78ede4f24c6c.diff
LOG: [libc] Implement wcscoll (#192778)
_Closes #191073_
- `libc/src/wchar/wcscoll.cpp` - Implementation of wcscoll
- `libc/src/wchar/wcscoll.h` - Internal header for wcscoll
- `libc/include/wchar.yaml` - Added wcscoll to public header spec
- `libc/src/wchar/CMakeLists.txt` - Added build target for wcscoll
- `libc/config/linux/x86_64/entrypoints.txt` - Registered wcscoll
entrypoint
- `libc/test/src/wchar/wcscoll_test.cpp` - Unit tests for wcscoll
- `libc/test/src/wchar/CMakeLists.txt` - Added test target for wcscoll
Note: Locale support is not yet implemented. `wcscoll` currently behaves
identically to `wcscmp` until locale support is available in llvm-libc.
Added:
libc/src/wchar/wcscoll.cpp
libc/src/wchar/wcscoll.h
libc/test/src/wchar/wcscoll_test.cpp
Modified:
libc/config/linux/x86_64/entrypoints.txt
libc/include/wchar.yaml
libc/src/wchar/CMakeLists.txt
libc/test/src/wchar/CMakeLists.txt
Removed:
################################################################################
diff --git a/libc/config/linux/x86_64/entrypoints.txt b/libc/config/linux/x86_64/entrypoints.txt
index 27569f0fae82b..d1c1d9496af67 100644
--- a/libc/config/linux/x86_64/entrypoints.txt
+++ b/libc/config/linux/x86_64/entrypoints.txt
@@ -401,6 +401,7 @@ set(TARGET_LIBC_ENTRYPOINTS
libc.src.wchar.wcsncmp
libc.src.wchar.wcsxfrm
libc.src.wchar.wcscmp
+ libc.src.wchar.wcscoll
libc.src.wchar.wcspbrk
libc.src.wchar.wcsrchr
libc.src.wchar.wcsspn
diff --git a/libc/include/wchar.yaml b/libc/include/wchar.yaml
index 6d6fc26c99fe9..6575f2504c900 100644
--- a/libc/include/wchar.yaml
+++ b/libc/include/wchar.yaml
@@ -1,6 +1,6 @@
header: wchar.h
standards:
- - stdc
+ - stdc
macros:
- macro_name: "NULL"
macro_header: null-macro.h
@@ -124,6 +124,13 @@ functions:
arguments:
- type: const wchar_t *
- type: const wchar_t *
+ - name: wcscoll
+ standards:
+ - stdc
+ return_type: int
+ arguments:
+ - type: const wchar_t *
+ - type: const wchar_t *
- name: wcspbrk
standards:
- stdc
diff --git a/libc/src/wchar/CMakeLists.txt b/libc/src/wchar/CMakeLists.txt
index fc273b0153a2b..89383c33c6a4e 100644
--- a/libc/src/wchar/CMakeLists.txt
+++ b/libc/src/wchar/CMakeLists.txt
@@ -332,6 +332,16 @@ add_entrypoint_object(
libc.hdr.wchar_macros
)
+add_entrypoint_object(
+ wcscoll
+ SRCS
+ wcscoll.cpp
+ HDRS
+ wcscoll.h
+ DEPENDS
+ libc.hdr.wchar_macros
+)
+
add_entrypoint_object(
wcsdup
SRCS
diff --git a/libc/src/wchar/wcscoll.cpp b/libc/src/wchar/wcscoll.cpp
new file mode 100644
index 0000000000000..abe86e1c9109e
--- /dev/null
+++ b/libc/src/wchar/wcscoll.cpp
@@ -0,0 +1,29 @@
+//===-- Implementation of wcscoll -----------------------------------------===//
+//
+// 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/wcscoll.h"
+
+#include "hdr/types/wchar_t.h"
+#include "src/__support/common.h"
+#include "src/__support/macros/config.h"
+#include "src/__support/macros/null_check.h"
+
+namespace LIBC_NAMESPACE_DECL {
+
+// TODO: Add support for locales.
+LLVM_LIBC_FUNCTION(int, wcscoll, (const wchar_t *s1, const wchar_t *s2)) {
+ LIBC_CRASH_ON_NULLPTR(s1);
+ LIBC_CRASH_ON_NULLPTR(s2);
+
+ for (; *s1 && (*s1 == *s2); ++s1, ++s2)
+ ;
+
+ return *s1 - *s2;
+}
+
+} // namespace LIBC_NAMESPACE_DECL
diff --git a/libc/src/wchar/wcscoll.h b/libc/src/wchar/wcscoll.h
new file mode 100644
index 0000000000000..bd499e9f41ced
--- /dev/null
+++ b/libc/src/wchar/wcscoll.h
@@ -0,0 +1,21 @@
+//===-- Implementation header for wcscoll ---------------------------------===//
+//
+// 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_WCSCOLL_H
+#define LLVM_LIBC_SRC_WCHAR_WCSCOLL_H
+
+#include "hdr/types/wchar_t.h"
+#include "src/__support/macros/config.h"
+
+namespace LIBC_NAMESPACE_DECL {
+
+int wcscoll(const wchar_t *s1, const wchar_t *s2);
+
+} // namespace LIBC_NAMESPACE_DECL
+
+#endif // LLVM_LIBC_SRC_WCHAR_WCSCOLL_H
diff --git a/libc/test/src/wchar/CMakeLists.txt b/libc/test/src/wchar/CMakeLists.txt
index 50359a37e212c..3fd279f19c755 100644
--- a/libc/test/src/wchar/CMakeLists.txt
+++ b/libc/test/src/wchar/CMakeLists.txt
@@ -242,6 +242,16 @@ add_libc_test(
libc.src.wchar.wcscmp
)
+add_libc_test(
+ wcscoll_test
+ SUITE
+ libc_wchar_unittests
+ SRCS
+ wcscoll_test.cpp
+ DEPENDS
+ libc.src.wchar.wcscoll
+)
+
add_libc_test(
wcspbrk_test
SUITE
diff --git a/libc/test/src/wchar/wcscoll_test.cpp b/libc/test/src/wchar/wcscoll_test.cpp
new file mode 100644
index 0000000000000..b611d1205540c
--- /dev/null
+++ b/libc/test/src/wchar/wcscoll_test.cpp
@@ -0,0 +1,139 @@
+//===-- Unittests for wcscoll ---------------------------------------------===//
+//
+// 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/signal_macros.h"
+#include "src/wchar/wcscoll.h"
+#include "test/UnitTest/Test.h"
+
+// TODO: Add more comprehensive tests once locale support is added.
+
+TEST(LlvmLibcWcscollTest, EmptyStringsShouldReturnZero) {
+ const wchar_t *s1 = L"";
+ const wchar_t *s2 = L"";
+
+ int result = LIBC_NAMESPACE::wcscoll(s1, s2);
+ ASSERT_EQ(result, 0);
+}
+
+TEST(LlvmLibcWcscollTest, EmptyStringShouldNotEqualNonEmptyString) {
+ const wchar_t *empty = L"";
+ const wchar_t *s = L"abc";
+
+ // An empty string comes before a non empty one lexicographically, so lt 0
+ int result = LIBC_NAMESPACE::wcscoll(empty, s);
+ ASSERT_LT(result, 0);
+
+ // Check the reversed behaviour
+ result = LIBC_NAMESPACE::wcscoll(s, empty);
+ ASSERT_GT(result, 0);
+}
+
+TEST(LlvmLibcWcscollTest, EqualStringsShouldReturnZero) {
+ const wchar_t *s1 = L"abc";
+ const wchar_t *s2 = L"abc";
+
+ // Check if it returns 0 for two equal strings
+ int result = LIBC_NAMESPACE::wcscoll(s1, s2);
+ ASSERT_EQ(result, 0);
+
+ // Verify for reversed operands
+ result = LIBC_NAMESPACE::wcscoll(s2, s1);
+ ASSERT_EQ(result, 0);
+}
+
+TEST(LlvmLibcWcscollTest, LexicographicalTest) {
+ const wchar_t *s1 = L"abc";
+ const wchar_t *s2 = L"def";
+
+ // Check if it returns lt 0 for (abc, def)
+ int result = LIBC_NAMESPACE::wcscoll(s1, s2);
+ ASSERT_LT(result, 0);
+
+ // Check if it returns gt 0 for (def, abc)
+ result = LIBC_NAMESPACE::wcscoll(s2, s1);
+ ASSERT_GT(result, 0);
+}
+
+TEST(LlvmLibcWcscollTest, NonAsciiTest) {
+ const wchar_t *s1 = L"AbCdEf__1230!! \u1111";
+ const wchar_t *s2 = L"AbCdEf__1230!! \u1111\u2222";
+
+ int result = LIBC_NAMESPACE::wcscoll(s1, s2);
+ ASSERT_LT(result, 0);
+
+ result = LIBC_NAMESPACE::wcscoll(s2, s1);
+ ASSERT_GT(result, 0);
+
+ result = LIBC_NAMESPACE::wcscoll(s1, s1);
+ ASSERT_EQ(result, 0);
+
+ // Empty string
+ const wchar_t *empty = L"";
+ result = LIBC_NAMESPACE::wcscoll(empty, s1);
+ ASSERT_LT(result, 0);
+
+ result = LIBC_NAMESPACE::wcscoll(s1, empty);
+ ASSERT_GT(result, 0);
+}
+
+TEST(LlvmLibcWcscollTest, EightDigitUCNTest) {
+ const wchar_t *s1 = L"abC\U0001F44D"; // thumbs up emoji
+ const wchar_t *s2 = L"abC\U0001F44E"; // thumbs down emoji
+
+ int result = LIBC_NAMESPACE::wcscoll(s1, s2);
+ ASSERT_LT(result, 0);
+
+ result = LIBC_NAMESPACE::wcscoll(s2, s1);
+ ASSERT_GT(result, 0);
+
+ result = LIBC_NAMESPACE::wcscoll(s1, s1);
+ ASSERT_EQ(result, 0);
+
+ // empty string
+ const wchar_t *empty = L"";
+ result = LIBC_NAMESPACE::wcscoll(empty, s1);
+ ASSERT_LT(result, 0);
+
+ result = LIBC_NAMESPACE::wcscoll(s1, empty);
+ ASSERT_GT(result, 0);
+}
+
+TEST(LlvmLibcWcscollTest, AsciiVsNonAsciiTest) {
+ const wchar_t *s1 = L"a";
+ const wchar_t *s2 = L"\uFFFF";
+ const wchar_t *s3 = L"\U0001000F";
+
+ // ascii and 4 digit unicode
+ int result = LIBC_NAMESPACE::wcscoll(s1, s2);
+ ASSERT_LT(result, 0);
+
+ result = LIBC_NAMESPACE::wcscoll(s2, s1);
+ ASSERT_GT(result, 0);
+
+ // ascii and 8 digit unicode
+ result = LIBC_NAMESPACE::wcscoll(s1, s3);
+ ASSERT_LT(result, 0);
+
+ result = LIBC_NAMESPACE::wcscoll(s3, s1);
+ ASSERT_GT(result, 0);
+
+ // 4 digit unicode and 8 digit unicode
+ result = LIBC_NAMESPACE::wcscoll(s2, s3);
+ ASSERT_LT(result, 0);
+
+ result = LIBC_NAMESPACE::wcscoll(s3, s2);
+ ASSERT_GT(result, 0);
+}
+
+#if defined(LIBC_ADD_NULL_CHECKS)
+TEST(LlvmLibcWcscollTest, NULLCheck) {
+ // Passing in a nullptr should crash the program
+ EXPECT_DEATH([] { LIBC_NAMESPACE::wcscoll(L"", nullptr); }, WITH_SIGNAL(-1));
+ EXPECT_DEATH([] { LIBC_NAMESPACE::wcscoll(nullptr, L""); }, WITH_SIGNAL(-1));
+}
+#endif // LIBC_ADD_NULL_CHECKS
More information about the libc-commits
mailing list