[libc] Implemented wcrtomb internal function and public libc function

libc/config/linux/x86_64/entrypoints.txt

@@ -365,6 +365,7 @@ set(TARGET_LIBC_ENTRYPOINTS
 
     # wchar.h entrypoints
     libc.src.wchar.btowc
+    libc.src.wchar.wcrtomb
     libc.src.wchar.wcslen
     libc.src.wchar.wctob
     libc.src.wchar.wmemmove

libc/hdr/types/CMakeLists.txt

@@ -20,6 +20,14 @@ add_proxy_header_library(
     libc.include.uchar
 )
 
+add_proxy_header_library(
+  mbstate_t 
+  HDRS
+    mbstate_t.h
+  DEPENDS
+    libc.include.llvm-libc-types.mbstate_t
+)
+
 add_proxy_header_library(
   div_t
   HDRS

libc/hdr/types/mbstate_t.h

@@ -0,0 +1,22 @@
+//===-- Definition of macros from mbstate_t.h -----------------------------===//
+//
+// 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_HDR_TYPES_MBSTATE_T_H
+#define LLVM_LIBC_HDR_TYPES_MBSTATE_T_H
+
+#ifdef LIBC_FULL_BUILD
+
+#include "include/llvm-libc-types/mbstate_t.h"
+
+#else // Overlay mode
+
+#include "hdr/wchar_overlay.h"
+
+#endif // LLVM_LIBC_FULL_BUILD
+
+#endif // LLVM_LIBC_HDR_TYPES_MBSTATE_T_H

libc/include/llvm-libc-types/mbstate_t.h

@@ -9,8 +9,12 @@
 #ifndef LLVM_LIBC_TYPES_MBSTATE_T_H
 #define LLVM_LIBC_TYPES_MBSTATE_T_H
 
-// TODO: Complete this once we implement functions that operate on this type.
+#include "../llvm-libc-macros/stdint-macros.h"
+
 typedef struct {
+  uint32_t __field1;
+  uint8_t __field2;
+  uint8_t __field3;
 } mbstate_t;
 
 #endif // LLVM_LIBC_TYPES_MBSTATE_T_H

libc/include/wchar.yaml

@@ -150,6 +150,14 @@ functions:
       - type: wchar_t *__restrict
       - type: const wchar_t *__restrict
       - type: size_t
+  - name: wcrtomb
+    standards:
+      - stdc
+    return_type: size_t
+    arguments:
+      - type: char *__restrict
+      - type: wchar_t
+      - type: mbstate_t *__restrict
   - name: wcscpy
     standards:
       - stdc

libc/src/__support/wchar/CMakeLists.txt

@@ -19,3 +19,19 @@ add_object_library(
     libc.src.__support.math_extras
     .mbstate
 )
+
+add_object_library(
+  wcrtomb
+  HDRS
+    wcrtomb.h
+  SRCS 
+    wcrtomb.cpp
+  DEPENDS
+    libc.hdr.types.char32_t
+    libc.hdr.types.size_t
+    libc.hdr.types.wchar_t
+    libc.src.__support.error_or
+    libc.src.__support.common
+    .character_converter
+    .mbstate
+)

libc/src/__support/wchar/mbstate.h

@@ -18,17 +18,17 @@ namespace internal {
 
 struct mbstate {
   // store a partial codepoint (in UTF-32)
-  char32_t partial;
+  char32_t partial = 0;
 
   /*
   Progress towards a conversion
     For utf8  -> utf32, increases with each CharacterConverter::push(utf8_byte)
     For utf32 ->  utf8, increases with each CharacterConverter::pop_utf8()
   */
-  uint8_t bytes_processed;
+  uint8_t bytes_processed = 0;
 
   // Total number of bytes that will be needed to represent this character
-  uint8_t total_bytes;
+  uint8_t total_bytes = 0;
 };
 
 } // namespace internal

libc/src/__support/wchar/wcrtomb.cpp

@@ -0,0 +1,50 @@
+//===-- Implementation of wcrtomb -----------------------------------------===//
+//
+// 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/wchar/wcrtomb.h"
+#include "src/__support/error_or.h"
+#include "src/__support/wchar/character_converter.h"
+#include "src/__support/wchar/mbstate.h"
+
+#include "hdr/types/char32_t.h"
+#include "hdr/types/size_t.h"
+#include "hdr/types/wchar_t.h"
+#include "src/__support/common.h"
+
+namespace LIBC_NAMESPACE_DECL {
+namespace internal {
+
+ErrorOr<size_t> wcrtomb(char *__restrict s, wchar_t wc,
+                        mbstate *__restrict ps) {
+  static_assert(sizeof(wchar_t) == 4);
+
+  CharacterConverter cr(ps);
+
+  // when s is nullptr, this is equivalent to wcrtomb(buf, L'\0', ps)
+  char buf[sizeof(wchar_t) / sizeof(char)];
+  if (s == nullptr) {
+    s = buf;
+    wc = L'\0';
+  }
+
+  int status = cr.push(static_cast<char32_t>(wc));
+  if (status != 0)
+    return Error(status);
+
+  size_t count = 0;
+  while (!cr.isComplete()) {
+    auto utf8 = cr.pop_utf8(); // can never fail as long as the push succeeded
+    *s = utf8.value();
+    s++;
+    count++;
+  }
+  return count;
+}
+
+} // namespace internal
+} // namespace LIBC_NAMESPACE_DECL

libc/src/__support/wchar/wcrtomb.h

@@ -0,0 +1,26 @@
+//===-- Implementation header for wcrtomb ---------------------------------===//
+//
+// 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__SUPPORT_WCHAR_WCRTOMB_H
+#define LLVM_LIBC_SRC__SUPPORT_WCHAR_WCRTOMB_H
+
+#include "hdr/types/size_t.h"
+#include "hdr/types/wchar_t.h"
+#include "src/__support/error_or.h"
+#include "src/__support/macros/config.h"
+#include "src/__support/wchar/mbstate.h"
+
+namespace LIBC_NAMESPACE_DECL {
+namespace internal {
+
+ErrorOr<size_t> wcrtomb(char *__restrict s, wchar_t wc, mbstate *__restrict ps);
+
+} // namespace internal
+} // namespace LIBC_NAMESPACE_DECL
+
+#endif // LLVM_LIBC_SRC__SUPPORT_WCHAR_WCRTOMB_H

libc/src/wchar/CMakeLists.txt

@@ -34,6 +34,20 @@ add_entrypoint_object(
     libc.src.__support.wctype_utils
 )
 
+add_entrypoint_object(
+  wcrtomb
+  SRCS
+    wcrtomb.cpp
+  HDRS
+    wcrtomb.h
+  DEPENDS
+    libc.hdr.types.wchar_t
+    libc.hdr.types.mbstate_t
+    libc.src.__support.libc_errno
+    libc.src.__support.wchar.wcrtomb
+    libc.src.__support.wchar.mbstate
+)
+
 add_entrypoint_object(
   wmemset
   SRCS

libc/src/wchar/wcrtomb.cpp

@@ -0,0 +1,37 @@
+//===-- Implementation of wcrtomb -----------------------------------------===//
+//
+// 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/wcrtomb.h"
+
+#include "hdr/types/mbstate_t.h"
+#include "src/__support/common.h"
+#include "src/__support/libc_errno.h"
+#include "src/__support/macros/config.h"
+#include "src/__support/wchar/mbstate.h"
+#include "src/__support/wchar/wcrtomb.h"
+
+namespace LIBC_NAMESPACE_DECL {
+
+LLVM_LIBC_FUNCTION(size_t, wcrtomb,
+                   (char *__restrict s, wchar_t wc, mbstate_t *__restrict ps)) {
+  static internal::mbstate internal_mbstate;
+
+  auto result = internal::wcrtomb(
+      s, wc,
+      ps == nullptr ? &internal_mbstate
+                    : reinterpret_cast<internal::mbstate *>(ps));
+
+  if (!result.has_value()) {
+    libc_errno = EILSEQ;
+    return -1;
+  }
+
+  return result.value();
+}
+
+} // namespace LIBC_NAMESPACE_DECL

libc/src/wchar/wcrtomb.h

@@ -0,0 +1,23 @@
+//===-- Implementation header for wcrtomb -----------------------*- C++ -*-===//
+//
+// 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_WCRTOMB_H
+#define LLVM_LIBC_SRC_WCHAR_WCRTOMB_H
+
+#include "hdr/types/mbstate_t.h"
+#include "hdr/types/size_t.h"
+#include "hdr/types/wchar_t.h"
+#include "src/__support/macros/config.h"
+
+namespace LIBC_NAMESPACE_DECL {
+
+size_t wcrtomb(char *__restrict s, wchar_t wc, mbstate_t *__restrict ps);
+
+} // namespace LIBC_NAMESPACE_DECL
+
+#endif // LLVM_LIBC_SRC_WCHAR_WCRTOMB_H

libc/test/src/wchar/CMakeLists.txt

@@ -33,6 +33,20 @@ add_libc_test(
     libc.src.wchar.wctob
 )
 
+add_libc_test(
+  wcrtomb_test
+  SUITE
+    libc_wchar_unittests
+  SRCS
+    wcrtomb_test.cpp
+  DEPENDS
+    libc.src.wchar.wcrtomb
+    libc.src.string.memset
+    libc.hdr.types.wchar_t
+    libc.hdr.types.mbstate_t
+    libc.src.__support.libc_errno
+)
+
 add_libc_test(
   wmemset_test 
   SUITE

libc/test/src/wchar/wcrtomb_test.cpp

@@ -0,0 +1,93 @@
+//===-- Unittests for wcrtomb --------------------------------------------===//
+//
+// 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/mbstate_t.h"
+#include "hdr/types/wchar_t.h"
+#include "src/__support/libc_errno.h"
+#include "src/string/memset.h"
+#include "src/wchar/wcrtomb.h"
+#include "test/UnitTest/Test.h"
+
+TEST(LlvmLibcWCRToMBTest, OneByte) {
+  mbstate_t state;
+  LIBC_NAMESPACE::memset(&state, 0, sizeof(mbstate_t));
+  wchar_t wc = L'U';
+  char mb[4];
+  size_t cnt = LIBC_NAMESPACE::wcrtomb(mb, wc, &state);
+  ASSERT_EQ(cnt, static_cast<size_t>(1));
+  ASSERT_EQ(mb[0], 'U');
+}
+
+TEST(LlvmLibcWCRToMBTest, TwoByte) {
+  mbstate_t state;
+  LIBC_NAMESPACE::memset(&state, 0, sizeof(mbstate_t));
+  // testing utf32: 0xff -> utf8: 0xc3 0xbf
+  wchar_t wc = 0xff;
+  char mb[4];
+  size_t cnt = LIBC_NAMESPACE::wcrtomb(mb, wc, &state);
+  ASSERT_EQ(cnt, static_cast<size_t>(2));
+  ASSERT_EQ(mb[0], static_cast<char>(0xc3));
+  ASSERT_EQ(mb[1], static_cast<char>(0xbf));
+}
+
+TEST(LlvmLibcWCRToMBTest, ThreeByte) {
+  mbstate_t state;
+  LIBC_NAMESPACE::memset(&state, 0, sizeof(mbstate_t));
+  // testing utf32: 0xac15 -> utf8: 0xea 0xb0 0x95
+  wchar_t wc = 0xac15;
+  char mb[4];
+  size_t cnt = LIBC_NAMESPACE::wcrtomb(mb, wc, &state);
+  ASSERT_EQ(cnt, static_cast<size_t>(3));
+  ASSERT_EQ(mb[0], static_cast<char>(0xea));
+  ASSERT_EQ(mb[1], static_cast<char>(0xb0));
+  ASSERT_EQ(mb[2], static_cast<char>(0x95));
+}
+
+TEST(LlvmLibcWCRToMBTest, FourByte) {
+  mbstate_t state;
+  LIBC_NAMESPACE::memset(&state, 0, sizeof(mbstate_t));
+  // testing utf32: 0x1f921 -> utf8: 0xf0 0x9f 0xa4 0xa1
+  wchar_t wc = 0x1f921;
+  char mb[4];
+  size_t cnt = LIBC_NAMESPACE::wcrtomb(mb, wc, &state);
+  ASSERT_EQ(cnt, static_cast<size_t>(4));
+  ASSERT_EQ(mb[0], static_cast<char>(0xf0));
+  ASSERT_EQ(mb[1], static_cast<char>(0x9f));
+  ASSERT_EQ(mb[2], static_cast<char>(0xa4));
+  ASSERT_EQ(mb[3], static_cast<char>(0xa1));
+}
+
+TEST(LlvmLibcWCRToMBTest, NullString) {
+  mbstate_t state;
+  LIBC_NAMESPACE::memset(&state, 0, sizeof(mbstate_t));
+  wchar_t wc = L'A';
+  char mb[4];
+
+  // should be equivalent to the call wcrtomb(buf, L'\0', state)
+  size_t cnt1 = LIBC_NAMESPACE::wcrtomb(nullptr, wc, &state);
+  size_t cnt2 = LIBC_NAMESPACE::wcrtomb(mb, L'\0', &state);
+
+  ASSERT_EQ(cnt1, cnt2);
+}
+
+TEST(LlvmLibcWCRToMBTest, NullState) {
+  wchar_t wc = L'A';
+  char mb[4];
+  size_t cnt = LIBC_NAMESPACE::wcrtomb(mb, wc, nullptr);
+  ASSERT_EQ(cnt, static_cast<size_t>(1));
+}
+
+TEST(LlvmLibcWCRToMBTest, InvalidWchar) {
+  mbstate_t state;
+  LIBC_NAMESPACE::memset(&state, 0, sizeof(mbstate_t));
+  wchar_t wc = 0x12ffff;
+  char mb[4];
+  size_t cnt = LIBC_NAMESPACE::wcrtomb(mb, wc, &state);
+  ASSERT_EQ(cnt, static_cast<size_t>(-1));
+  ASSERT_EQ(static_cast<int>(libc_errno), EILSEQ);
+}