[libc-commits] [libc] 4f2686e - [libc] Implemented mblen functions (#150141)
via libc-commits
libc-commits at lists.llvm.org
Thu Jul 24 11:28:28 PDT 2025
Author: sribee8
Date: 2025-07-24T18:28:25Z
New Revision: 4f2686e5a131094d985677591482508586f1f5c9
URL: https://github.com/llvm/llvm-project/commit/4f2686e5a131094d985677591482508586f1f5c9
DIFF: https://github.com/llvm/llvm-project/commit/4f2686e5a131094d985677591482508586f1f5c9.diff
LOG: [libc] Implemented mblen functions (#150141)
Implemented mblen and mbrlen as well as tests
---------
Co-authored-by: Sriya Pratipati <sriyap at google.com>
Added:
libc/src/wchar/mblen.cpp
libc/src/wchar/mblen.h
libc/src/wchar/mbrlen.cpp
libc/src/wchar/mbrlen.h
libc/test/src/wchar/mblen_test.cpp
libc/test/src/wchar/mbrlen_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 381359cec6f1d..3cb1c483cea9e 100644
--- a/libc/config/linux/x86_64/entrypoints.txt
+++ b/libc/config/linux/x86_64/entrypoints.txt
@@ -1261,6 +1261,8 @@ if(LLVM_LIBC_FULL_BUILD)
libc.src.sys.socket.recvmsg
# wchar.h entrypoints
+ libc.src.wchar.mblen
+ libc.src.wchar.mbrlen
libc.src.wchar.mbrtowc
libc.src.wchar.mbtowc
libc.src.wchar.wcrtomb
diff --git a/libc/include/wchar.yaml b/libc/include/wchar.yaml
index 123d3440aeec3..4adf596abe650 100644
--- a/libc/include/wchar.yaml
+++ b/libc/include/wchar.yaml
@@ -53,6 +53,21 @@ functions:
- type: wchar_t *__restrict
- type: const char *__restrict
- type: size_t
+ - name: mblen
+ standards:
+ - stdc
+ return_type: int
+ arguments:
+ - type: const char *
+ - type: size_t
+ - name: mbrlen
+ standards:
+ - stdc
+ return_type: size_t
+ arguments:
+ - type: const char *__restrict
+ - type: size_t
+ - type: mbstate_t *__restrict
- name: wmemset
standards:
- stdc
diff --git a/libc/src/wchar/CMakeLists.txt b/libc/src/wchar/CMakeLists.txt
index 159778df6acca..2b95d94e4230a 100644
--- a/libc/src/wchar/CMakeLists.txt
+++ b/libc/src/wchar/CMakeLists.txt
@@ -169,6 +169,37 @@ add_entrypoint_object(
libc.src.__support.wchar.mbstate
)
+add_entrypoint_object(
+ mblen
+ SRCS
+ mblen.cpp
+ HDRS
+ mblen.h
+ DEPENDS
+ libc.hdr.types.size_t
+ libc.src.__support.common
+ libc.src.__support.macros.config
+ libc.src.__support.libc_errno
+ libc.src.__support.wchar.mbrtowc
+ libc.src.__support.wchar.mbstate
+)
+
+add_entrypoint_object(
+ mbrlen
+ SRCS
+ mbrlen.cpp
+ HDRS
+ mbrlen.h
+ DEPENDS
+ libc.hdr.types.size_t
+ libc.hdr.types.mbstate_t
+ libc.src.__support.common
+ libc.src.__support.macros.config
+ libc.src.__support.wchar.mbrtowc
+ libc.src.__support.libc_errno
+ libc.src.__support.wchar.mbstate
+)
+
add_entrypoint_object(
wmemset
SRCS
diff --git a/libc/src/wchar/mblen.cpp b/libc/src/wchar/mblen.cpp
new file mode 100644
index 0000000000000..2d15b3e0e5648
--- /dev/null
+++ b/libc/src/wchar/mblen.cpp
@@ -0,0 +1,35 @@
+//===-- Implementation of mblen -------------------------------------------===//
+//
+// 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/mblen.h"
+
+#include "hdr/types/size_t.h"
+#include "src/__support/common.h"
+#include "src/__support/libc_errno.h"
+#include "src/__support/macros/config.h"
+#include "src/__support/wchar/mbrtowc.h"
+#include "src/__support/wchar/mbstate.h"
+
+namespace LIBC_NAMESPACE_DECL {
+
+LLVM_LIBC_FUNCTION(int, mblen, (const char *s, size_t n)) {
+ // returns 0 since UTF-8 encoding is not state-dependent
+ if (s == nullptr)
+ return 0;
+ internal::mbstate internal_mbstate;
+ auto ret = internal::mbrtowc(nullptr, s, n, &internal_mbstate);
+ if (!ret.has_value() || static_cast<int>(ret.value()) == -2) {
+ // Encoding failure
+ if (!ret.has_value())
+ libc_errno = EILSEQ;
+ return -1;
+ }
+ return static_cast<int>(ret.value());
+}
+
+} // namespace LIBC_NAMESPACE_DECL
diff --git a/libc/src/wchar/mblen.h b/libc/src/wchar/mblen.h
new file mode 100644
index 0000000000000..a315a2f12f6a1
--- /dev/null
+++ b/libc/src/wchar/mblen.h
@@ -0,0 +1,21 @@
+//===-- Implementation header for mblen -----------------------------------===//
+//
+// 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_MBLEN_H
+#define LLVM_LIBC_SRC_WCHAR_MBLEN_H
+
+#include "hdr/types/size_t.h"
+#include "src/__support/macros/config.h"
+
+namespace LIBC_NAMESPACE_DECL {
+
+int mblen(const char *s, size_t n);
+
+} // namespace LIBC_NAMESPACE_DECL
+
+#endif // LLVM_LIBC_SRC_WCHAR_MBLEN_H
diff --git a/libc/src/wchar/mbrlen.cpp b/libc/src/wchar/mbrlen.cpp
new file mode 100644
index 0000000000000..8de78e099566b
--- /dev/null
+++ b/libc/src/wchar/mbrlen.cpp
@@ -0,0 +1,37 @@
+//===-- Implementation of mbrlen ------------------------------------------===//
+//
+// 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/mbrlen.h"
+
+#include "hdr/types/mbstate_t.h"
+#include "hdr/types/size_t.h"
+#include "src/__support/common.h"
+#include "src/__support/libc_errno.h"
+#include "src/__support/macros/config.h"
+#include "src/__support/wchar/mbrtowc.h"
+#include "src/__support/wchar/mbstate.h"
+
+namespace LIBC_NAMESPACE_DECL {
+
+LLVM_LIBC_FUNCTION(size_t, mbrlen,
+ (const char *__restrict s, size_t n,
+ mbstate_t *__restrict ps)) {
+ static internal::mbstate internal_mbstate;
+ auto ret = internal::mbrtowc(nullptr, s, n,
+ ps == nullptr
+ ? &internal_mbstate
+ : reinterpret_cast<internal::mbstate *>(ps));
+ if (!ret.has_value()) {
+ // Encoding failure
+ libc_errno = ret.error();
+ return -1;
+ }
+ return ret.value();
+}
+
+} // namespace LIBC_NAMESPACE_DECL
diff --git a/libc/src/wchar/mbrlen.h b/libc/src/wchar/mbrlen.h
new file mode 100644
index 0000000000000..08b59cfc8651c
--- /dev/null
+++ b/libc/src/wchar/mbrlen.h
@@ -0,0 +1,22 @@
+//===-- Implementation header for mbrlen ----------------------------------===//
+//
+// 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_MBRLEN_H
+#define LLVM_LIBC_SRC_WCHAR_MBRLEN_H
+
+#include "hdr/types/mbstate_t.h"
+#include "hdr/types/size_t.h"
+#include "src/__support/macros/config.h"
+
+namespace LIBC_NAMESPACE_DECL {
+
+size_t mbrlen(const char *__restrict s, size_t n, mbstate_t *__restrict ps);
+
+} // namespace LIBC_NAMESPACE_DECL
+
+#endif // LLVM_LIBC_SRC_WCHAR_MBRLEN_H
diff --git a/libc/test/src/wchar/CMakeLists.txt b/libc/test/src/wchar/CMakeLists.txt
index 176cf7c3487cd..baa52b74c3d97 100644
--- a/libc/test/src/wchar/CMakeLists.txt
+++ b/libc/test/src/wchar/CMakeLists.txt
@@ -64,6 +64,33 @@ add_libc_test(
libc.test.UnitTest.ErrnoCheckingTest
)
+add_libc_test(
+ mblen_test
+ SUITE
+ libc_wchar_unittests
+ SRCS
+ mblen_test.cpp
+ DEPENDS
+ libc.src.__support.libc_errno
+ libc.src.wchar.mblen
+ libc.test.UnitTest.ErrnoCheckingTest
+)
+
+add_libc_test(
+ mbrlen_test
+ SUITE
+ libc_wchar_unittests
+ SRCS
+ mbrlen_test.cpp
+ DEPENDS
+ libc.src.__support.libc_errno
+ libc.src.__support.wchar.mbstate
+ libc.src.string.memset
+ libc.src.wchar.mbrlen
+ libc.hdr.types.mbstate_t
+ libc.test.UnitTest.ErrnoCheckingTest
+)
+
add_libc_test(
wctob_test
SUITE
diff --git a/libc/test/src/wchar/mblen_test.cpp b/libc/test/src/wchar/mblen_test.cpp
new file mode 100644
index 0000000000000..efd4df7020741
--- /dev/null
+++ b/libc/test/src/wchar/mblen_test.cpp
@@ -0,0 +1,104 @@
+//===-- Unittests for mblen -----------------------------------------------===//
+//
+// 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/__support/libc_errno.h"
+#include "src/wchar/mblen.h"
+#include "test/UnitTest/ErrnoCheckingTest.h"
+#include "test/UnitTest/Test.h"
+
+using LlvmLibcMBLenTest = LIBC_NAMESPACE::testing::ErrnoCheckingTest;
+
+TEST_F(LlvmLibcMBLenTest, OneByte) {
+ const char *ch = "A";
+ int n = LIBC_NAMESPACE::mblen(ch, 1);
+ ASSERT_ERRNO_SUCCESS();
+ ASSERT_EQ(n, 1);
+
+ // Should fail since we have not read enough
+ n = LIBC_NAMESPACE::mblen(ch, 0);
+ ASSERT_ERRNO_SUCCESS();
+ ASSERT_EQ(n, -1);
+}
+
+TEST_F(LlvmLibcMBLenTest, TwoByte) {
+ const char ch[2] = {static_cast<char>(0xC2),
+ static_cast<char>(0x8E)}; // car symbol
+ int n = LIBC_NAMESPACE::mblen(ch, 4);
+ ASSERT_ERRNO_SUCCESS();
+ ASSERT_EQ(n, 2);
+
+ // Should fail since we have not read enough
+ n = LIBC_NAMESPACE::mblen(ch, 1);
+ ASSERT_EQ(n, -1);
+ ASSERT_ERRNO_SUCCESS();
+ // Should fail after trying to read next byte too
+ n = LIBC_NAMESPACE::mblen(ch + 1, 1);
+ ASSERT_EQ(n, -1);
+ // This one should be an invalid starting byte so should set errno
+ ASSERT_ERRNO_EQ(EILSEQ);
+}
+
+TEST_F(LlvmLibcMBLenTest, ThreeByte) {
+ const char ch[3] = {static_cast<char>(0xE2), static_cast<char>(0x88),
+ static_cast<char>(0x91)}; // ∑ sigma symbol
+ int n = LIBC_NAMESPACE::mblen(ch, 3);
+ ASSERT_EQ(n, 3);
+ ASSERT_ERRNO_SUCCESS();
+
+ // Should fail since we have not read enough
+ n = LIBC_NAMESPACE::mblen(ch, 2);
+ ASSERT_EQ(n, -1);
+ ASSERT_ERRNO_SUCCESS();
+}
+
+TEST_F(LlvmLibcMBLenTest, FourByte) {
+ const char ch[4] = {static_cast<char>(0xF0), static_cast<char>(0x9F),
+ static_cast<char>(0xA4),
+ static_cast<char>(0xA1)}; // 🤡 clown emoji
+ int n = LIBC_NAMESPACE::mblen(ch, 4);
+ ASSERT_EQ(n, 4);
+ ASSERT_ERRNO_SUCCESS();
+
+ // Should fail since we have not read enough
+ n = LIBC_NAMESPACE::mblen(ch, 2);
+ ASSERT_EQ(n, -1);
+ ASSERT_ERRNO_SUCCESS();
+}
+
+TEST_F(LlvmLibcMBLenTest, InvalidByte) {
+ const char ch[1] = {static_cast<char>(0x80)};
+ int n = LIBC_NAMESPACE::mblen(ch, 1);
+ ASSERT_EQ(n, -1);
+ ASSERT_ERRNO_EQ(EILSEQ);
+}
+
+TEST_F(LlvmLibcMBLenTest, InvalidMultiByte) {
+ const char ch[4] = {static_cast<char>(0x80), static_cast<char>(0x00),
+ static_cast<char>(0x80),
+ static_cast<char>(0x00)}; // invalid sequence of bytes
+ // Trying to push all 4 should error
+ int n = LIBC_NAMESPACE::mblen(ch, 4);
+ ASSERT_EQ(n, -1);
+ ASSERT_ERRNO_EQ(EILSEQ);
+
+ // Trying to push the second and third should correspond to null wc
+ n = LIBC_NAMESPACE::mblen(ch + 1, 2);
+ ASSERT_EQ(n, 0);
+ ASSERT_ERRNO_SUCCESS();
+}
+
+TEST_F(LlvmLibcMBLenTest, NullString) {
+ // reading on nullptr should return 0
+ int n = LIBC_NAMESPACE::mblen(nullptr, 2);
+ ASSERT_EQ(n, 0);
+ ASSERT_ERRNO_SUCCESS();
+ // reading a null terminator should return 0
+ const char *ch = "\0";
+ n = LIBC_NAMESPACE::mblen(ch, 1);
+ ASSERT_EQ(n, 0);
+}
diff --git a/libc/test/src/wchar/mbrlen_test.cpp b/libc/test/src/wchar/mbrlen_test.cpp
new file mode 100644
index 0000000000000..e1452bf416054
--- /dev/null
+++ b/libc/test/src/wchar/mbrlen_test.cpp
@@ -0,0 +1,139 @@
+//===-- Unittests for mbrlen ----------------------------------------------===//
+//
+// 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/__support/libc_errno.h"
+#include "src/__support/wchar/mbstate.h"
+#include "src/string/memset.h"
+#include "src/wchar/mbrlen.h"
+#include "test/UnitTest/ErrnoCheckingTest.h"
+#include "test/UnitTest/Test.h"
+
+using LlvmLibcMBRLenTest = LIBC_NAMESPACE::testing::ErrnoCheckingTest;
+
+TEST_F(LlvmLibcMBRLenTest, OneByte) {
+ const char *ch = "A";
+ mbstate_t mb;
+ LIBC_NAMESPACE::memset(&mb, 0, sizeof(mbstate_t));
+ size_t n = LIBC_NAMESPACE::mbrlen(ch, 1, &mb);
+ ASSERT_ERRNO_SUCCESS();
+ ASSERT_EQ(n, static_cast<size_t>(1));
+
+ // Should fail since we have not read enough
+ n = LIBC_NAMESPACE::mbrlen(ch, 0, &mb);
+ ASSERT_ERRNO_SUCCESS();
+ ASSERT_EQ(n, static_cast<size_t>(-2));
+}
+
+TEST_F(LlvmLibcMBRLenTest, TwoByte) {
+ const char ch[2] = {static_cast<char>(0xC2),
+ static_cast<char>(0x8E)}; // car symbol
+ mbstate_t mb;
+ LIBC_NAMESPACE::memset(&mb, 0, sizeof(mbstate_t));
+ size_t n = LIBC_NAMESPACE::mbrlen(ch, 4, nullptr);
+ ASSERT_ERRNO_SUCCESS();
+ ASSERT_EQ(static_cast<int>(n), 2);
+
+ // Should fail since we have not read enough
+ n = LIBC_NAMESPACE::mbrlen(ch, 1, &mb);
+ ASSERT_EQ(static_cast<int>(n), -2);
+ ASSERT_ERRNO_SUCCESS();
+ // Should pass after trying to read next byte
+ n = LIBC_NAMESPACE::mbrlen(ch + 1, 1, &mb);
+ ASSERT_EQ(static_cast<int>(n), 1);
+ ASSERT_ERRNO_SUCCESS();
+}
+
+TEST_F(LlvmLibcMBRLenTest, ThreeByte) {
+ const char ch[3] = {static_cast<char>(0xE2), static_cast<char>(0x88),
+ static_cast<char>(0x91)}; // ∑ sigma symbol
+ mbstate_t mb;
+ LIBC_NAMESPACE::memset(&mb, 0, sizeof(mbstate_t));
+ size_t n = LIBC_NAMESPACE::mbrlen(ch, 3, &mb);
+ ASSERT_EQ(static_cast<int>(n), 3);
+ ASSERT_ERRNO_SUCCESS();
+
+ // Should fail since we have not read enough
+ n = LIBC_NAMESPACE::mbrlen(ch, 2, &mb);
+ ASSERT_EQ(static_cast<int>(n), -2);
+ ASSERT_ERRNO_SUCCESS();
+}
+
+TEST_F(LlvmLibcMBRLenTest, FourByte) {
+ const char ch[4] = {static_cast<char>(0xF0), static_cast<char>(0x9F),
+ static_cast<char>(0xA4),
+ static_cast<char>(0xA1)}; // 🤡 clown emoji
+ mbstate_t mb;
+ LIBC_NAMESPACE::memset(&mb, 0, sizeof(mbstate_t));
+ size_t n = LIBC_NAMESPACE::mbrlen(ch, 4, &mb);
+ ASSERT_EQ(static_cast<int>(n), 4);
+ ASSERT_ERRNO_SUCCESS();
+
+ // Should fail since we have not read enough
+ n = LIBC_NAMESPACE::mbrlen(ch, 2, &mb);
+ ASSERT_EQ(static_cast<int>(n), -2);
+ ASSERT_ERRNO_SUCCESS();
+
+ // Should fail since we have not read enough
+ n = LIBC_NAMESPACE::mbrlen(ch + 2, 1, &mb);
+ ASSERT_EQ(static_cast<int>(n), -2);
+ ASSERT_ERRNO_SUCCESS();
+
+ // Should pass after reading final byte
+ n = LIBC_NAMESPACE::mbrlen(ch + 3, 5, &mb);
+ ASSERT_EQ(static_cast<int>(n), 1);
+ ASSERT_ERRNO_SUCCESS();
+}
+
+TEST_F(LlvmLibcMBRLenTest, InvalidByte) {
+ const char ch[1] = {static_cast<char>(0x80)};
+ size_t n = LIBC_NAMESPACE::mbrlen(ch, 1, nullptr);
+ ASSERT_EQ(static_cast<int>(n), -1);
+ ASSERT_ERRNO_EQ(EILSEQ);
+}
+
+TEST_F(LlvmLibcMBRLenTest, InvalidMultiByte) {
+ const char ch[4] = {static_cast<char>(0x80), static_cast<char>(0x00),
+ static_cast<char>(0x80),
+ static_cast<char>(0x00)}; // invalid sequence of bytes
+ mbstate_t mb;
+ LIBC_NAMESPACE::memset(&mb, 0, sizeof(mbstate_t));
+ // Trying to push all 4 should error
+ size_t n = LIBC_NAMESPACE::mbrlen(ch, 4, &mb);
+ ASSERT_EQ(static_cast<int>(n), -1);
+ ASSERT_ERRNO_EQ(EILSEQ);
+
+ // Trying to push the second and third should correspond to null wc
+ n = LIBC_NAMESPACE::mbrlen(ch + 1, 2, &mb);
+ ASSERT_EQ(static_cast<int>(n), 0);
+ ASSERT_ERRNO_SUCCESS();
+}
+
+TEST_F(LlvmLibcMBRLenTest, NullString) {
+ // reading on nullptr should return 0
+ size_t n = LIBC_NAMESPACE::mbrlen(nullptr, 2, nullptr);
+ ASSERT_EQ(static_cast<int>(n), 0);
+ ASSERT_ERRNO_SUCCESS();
+ // reading a null terminator should return 0
+ const char *ch = "\0";
+ n = LIBC_NAMESPACE::mbrlen(ch, 1, nullptr);
+ ASSERT_EQ(static_cast<int>(n), 0);
+}
+
+TEST_F(LlvmLibcMBRLenTest, InvalidMBState) {
+ const char ch[4] = {static_cast<char>(0xC2), static_cast<char>(0x8E),
+ static_cast<char>(0xC7), static_cast<char>(0x8C)};
+ mbstate_t *mb;
+ LIBC_NAMESPACE::internal::mbstate inv;
+ inv.total_bytes = 6;
+ mb = reinterpret_cast<mbstate_t *>(&inv);
+ // invalid mbstate should error
+ size_t n = LIBC_NAMESPACE::mbrlen(ch, 2, mb);
+ ASSERT_EQ(static_cast<int>(n), -1);
+ ASSERT_ERRNO_EQ(EINVAL);
+}
More information about the libc-commits
mailing list