[libc] mbrtowc implementation

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.mbrtowc
     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

@@ -29,6 +29,15 @@ functions:
     return_type: wint_t
     arguments:
       - type: int
+  - name: mbrtowc
+    standards:
+      - stdc
+    return_type: size_t
+    arguments:
+      - type: wchar_t * __restrict
+      - type: const char * __restrict
+      - type: size_t
+      - type: mbstate_t * __restrict
   - name: wmemset
     standards:
       - stdc

libc/src/__support/wchar/CMakeLists.txt

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

libc/src/__support/wchar/character_converter.cpp

@@ -38,6 +38,11 @@ bool CharacterConverter::isComplete() {
   return state->bytes_processed == state->total_bytes;
 }
 
+bool CharacterConverter::isFull() {
+  return (state->bytes_processed == state->total_bytes) &&
+         (state->total_bytes != 0);
+}
+
 int CharacterConverter::push(char8_t utf8_byte) {
   uint8_t num_ones = static_cast<uint8_t>(cpp::countl_one(utf8_byte));
   // Checking the first byte if first push
@@ -68,8 +73,10 @@ int CharacterConverter::push(char8_t utf8_byte) {
   }
   // Any subsequent push
   // Adding 6 more bits so need to left shift
-  if (num_ones == 1 && !isComplete()) {
-    char32_t byte = utf8_byte & MASK_ENCODED_BITS;
+  constexpr size_t ENCODED_BITS_PER_UTF8 = 6;
+  if (num_ones == 1 && !isFull()) {
+    char32_t byte =
+        utf8_byte & mask_trailing_ones<uint32_t, ENCODED_BITS_PER_UTF8>();
     state->partial = state->partial << ENCODED_BITS_PER_UTF8;
     state->partial |= byte;
     state->bytes_processed++;
@@ -107,7 +114,7 @@ int CharacterConverter::push(char32_t utf32) {
 ErrorOr<char32_t> CharacterConverter::pop_utf32() {
   // If pop is called too early, do not reset the state, use error to determine
   // whether enough bytes have been pushed
-  if (!isComplete() || state->bytes_processed == 0)
+  if (!isFull() || state->bytes_processed == 0)
     return Error(-1);
   char32_t utf32 = state->partial;
   // reset if successful pop

libc/src/__support/wchar/character_converter.h

@@ -27,6 +27,7 @@ class CharacterConverter {
 
   void clear();
   bool isComplete();
+  bool isFull();
 
   int push(char8_t utf8_byte);
   int push(char32_t utf32);

libc/src/__support/wchar/mbrtowc.cpp

@@ -0,0 +1,49 @@
+//===-- Implementation for mbrtowc function ---------------------*- 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
+//
+//===----------------------------------------------------------------------===//
+
+#include "src/__support/wchar/mbrtowc.h"
+#include "hdr/types/mbstate_t.h"
+#include "hdr/types/size_t.h"
+#include "hdr/types/wchar_t.h"
+#include "src/__support/common.h"
+#include "src/__support/error_or.h"
+#include "src/__support/macros/config.h"
+#include "src/__support/wchar/character_converter.h"
+#include "src/__support/wchar/mbstate.h"
+
+namespace LIBC_NAMESPACE_DECL {
+namespace internal {
+
+ErrorOr<size_t> mbrtowc(wchar_t *__restrict pwc, const char *__restrict s,
+                        size_t n, mbstate *__restrict ps) {
+  CharacterConverter char_conv(ps);
+  if (s == nullptr)
+    return 0;
+  size_t i = 0;
+  // Reading in bytes until we have a complete wc or error
+  for (; i < n && !char_conv.isFull(); ++i) {
+    int err = char_conv.push(static_cast<char8_t>(s[i]));
+    // Encoding error
+    if (err == -1)
+      return Error(-1);
+  }
+  auto wc = char_conv.pop_utf32();
+  if (wc.has_value()) {
+    *pwc = wc.value();
+    // null terminator -> return 0
+    if (wc.value() == L'\0')
+      return 0;
+    return i;
+  }
+  // Incomplete but potentially valid
+  return -2;
+}
+
+} // namespace internal
+
+} // namespace LIBC_NAMESPACE_DECL

libc/src/__support/wchar/mbrtowc.h

@@ -0,0 +1,29 @@
+//===-- Implementation header for mbrtowc function --------------*- 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___SUPPORT_WCHAR_MBRTOWC
+#define LLVM_LIBC_SRC___SUPPORT_WCHAR_MBRTOWC
+
+#include "hdr/types/size_t.h"
+#include "hdr/types/wchar_t.h"
+#include "src/__support/common.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> mbrtowc(wchar_t *__restrict pwc, const char *__restrict s,
+                        size_t n, mbstate *__restrict ps);
+
+} // namespace internal
+
+} // namespace LIBC_NAMESPACE_DECL
+
+#endif // LLVM_LIBC_SRC___SUPPORT_WCHAR_MBRTOWC

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/wchar/CMakeLists.txt

@@ -34,6 +34,23 @@ add_entrypoint_object(
     libc.src.__support.wctype_utils
 )
 
+add_entrypoint_object(
+  mbrtowc
+  SRCS
+    mbrtowc.cpp
+  HDRS
+    mbrtowc.h
+  DEPENDS
+    libc.hdr.types.size_t
+    libc.hdr.types.mbstate_t
+    libc.hdr.types.wchar_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

libc/src/wchar/mbrtowc.cpp

@@ -0,0 +1,38 @@
+//===-- Implementation of mbrtowc -----------------------------------------===//
+//
+// 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/mbrtowc.h"
+
+#include "hdr/types/mbstate_t.h"
+#include "hdr/types/size_t.h"
+#include "hdr/types/wchar_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, mbrtowc,
+                   (wchar_t *__restrict pwc, const char *__restrict s, size_t n,
+                    mbstate_t *__restrict ps)) {
+  static internal::mbstate internal_mbstate;
+  auto ret = internal::mbrtowc(pwc, s, n,
+                               ps == nullptr
+                                   ? &internal_mbstate
+                                   : reinterpret_cast<internal::mbstate *>(ps));
+  if (!ret.has_value()) {
+    // Encoding failure
+    libc_errno = EILSEQ;
+    return -1;
+  }
+  return ret.value();
+}
+
+} // namespace LIBC_NAMESPACE_DECL

libc/src/wchar/mbrtowc.h

@@ -0,0 +1,24 @@
+//===-- Implementation header for mbrtowc ---------------------------------===//
+//
+// 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_MBRTOWC_H
+#define LLVM_LIBC_SRC_WCHAR_MBRTOWC_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 mbrtowc(wchar_t *__restrict pwc, const char *__restrict s, size_t n,
+               mbstate_t *__restrict ps);
+
+} // namespace LIBC_NAMESPACE_DECL
+
+#endif // LLVM_LIBC_SRC_WCHAR_MBRTOWC_H

libc/test/src/wchar/CMakeLists.txt

@@ -23,6 +23,20 @@ add_libc_test(
     libc.src.wchar.btowc
 )
 
+add_libc_test(
+  mbrtowc_test
+  SUITE
+    libc_wchar_unittests
+  SRCS
+    mbrtowc_test.cpp
+  DEPENDS
+    libc.src.__support.libc_errno
+    libc.src.string.memset
+    libc.src.wchar.mbrtowc
+    libc.hdr.types.mbstate_t
+    libc.hdr.types.wchar_t
+)
+
 add_libc_test(
   wctob_test
   SUITE