wcsrtombs tests

Author: uzairnawaz

libc/src/wchar/wcsrtombs.cpp

@@ -25,7 +25,7 @@ LLVM_LIBC_FUNCTION(size_t, wcsrtombs,
                     size_t n, mbstate_t *ps)) {
   static internal::mbstate internal_mbstate;
   internal::StringConverter<char32_t> str_conv(
-      reinterpret_cast<const char32_t *>(pwcs),
+      reinterpret_cast<const char32_t *>(*pwcs),
       ps == nullptr ? &internal_mbstate
                     : reinterpret_cast<internal::mbstate *>(ps),
       n);

libc/test/src/wchar/CMakeLists.txt

@@ -113,6 +113,20 @@ add_libc_test(
     libc.hdr.types.wchar_t
 )
 
+add_libc_test(
+  wcsrtombs_test
+  SUITE
+    libc_wchar_unittests
+  SRCS
+    wcsrtombs_test.cpp
+  DEPENDS
+    libc.src.wchar.wcsrtombs
+    libc.test.UnitTest.ErrnoCheckingTest
+    libc.hdr.types.wchar_t
+    libc.src.string.memset
+    libc.hdr.types.mbstate_t
+)
+
 add_libc_test(
   wmemset_test 
   SUITE

libc/test/src/wchar/wcsrtombs_test.cpp

@@ -0,0 +1,129 @@
+//===-- Unittests for wcsrtombs -------------------------------------------===//
+//
+// 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 "src/string/memset.h"
+#include "src/wchar/wcsrtombs.h"
+#include "test/UnitTest/ErrnoCheckingTest.h"
+#include "test/UnitTest/Test.h"
+
+using LlvmLibcWcsrtombs = LIBC_NAMESPACE::testing::ErrnoCheckingTest;
+
+TEST_F(LlvmLibcWcsrtombs, AllMultibyteLengths) {
+  mbstate_t state;
+  LIBC_NAMESPACE::memset(&state, 0, sizeof(mbstate_t));
+
+  /// clown emoji, sigma symbol, y with diaeresis, letter A
+  const wchar_t src[] = {static_cast<wchar_t>(0x1f921),
+                         static_cast<wchar_t>(0x2211),
+                         static_cast<wchar_t>(0xff), static_cast<wchar_t>(0x41),
+                         static_cast<wchar_t>(0x0)};
+  const wchar_t *cur = src;
+  char mbs[11];
+
+  ASSERT_EQ(LIBC_NAMESPACE::wcsrtombs(mbs, &cur, 11, &state),
+            static_cast<size_t>(10));
+  ASSERT_ERRNO_SUCCESS();
+  ASSERT_EQ(cur, nullptr);
+  ASSERT_EQ(mbs[0], '\xF0'); // clown begin
+  ASSERT_EQ(mbs[1], '\x9F');
+  ASSERT_EQ(mbs[2], '\xA4');
+  ASSERT_EQ(mbs[3], '\xA1');
+  ASSERT_EQ(mbs[4], '\xE2'); // sigma begin
+  ASSERT_EQ(mbs[5], '\x88');
+  ASSERT_EQ(mbs[6], '\x91');
+  ASSERT_EQ(mbs[7], '\xC3'); // y diaeresis begin
+  ASSERT_EQ(mbs[8], '\xBF');
+  ASSERT_EQ(mbs[9], '\x41'); // A begin
+  ASSERT_EQ(mbs[10], '\0');  // null terminator
+}
+
+TEST_F(LlvmLibcWcsrtombs, DestLimit) {
+  mbstate_t state;
+  LIBC_NAMESPACE::memset(&state, 0, sizeof(mbstate_t));
+
+  /// clown emoji, sigma symbol, y with diaeresis, letter A
+  const wchar_t src[] = {static_cast<wchar_t>(0x1f921),
+                         static_cast<wchar_t>(0x2211),
+                         static_cast<wchar_t>(0xff), static_cast<wchar_t>(0x41),
+                         static_cast<wchar_t>(0x0)};
+  const wchar_t *cur = src;
+
+  char mbs[11];
+  for (int i = 0; i < 11; ++i)
+    mbs[i] = '\x01'; // dummy initial values
+
+  ASSERT_EQ(LIBC_NAMESPACE::wcsrtombs(mbs, &cur, 4, &state),
+            static_cast<size_t>(4));
+  ASSERT_ERRNO_SUCCESS();
+  ASSERT_EQ(cur, src + 1);
+  ASSERT_EQ(mbs[0], '\xF0');
+  ASSERT_EQ(mbs[1], '\x9F');
+  ASSERT_EQ(mbs[2], '\xA4');
+  ASSERT_EQ(mbs[3], '\xA1');
+  ASSERT_EQ(mbs[4], '\x01'); // didn't write more than 4 bytes
+
+  for (int i = 0; i < 11; ++i)
+    mbs[i] = '\x01'; // dummy initial values
+  LIBC_NAMESPACE::memset(&state, 0, sizeof(mbstate_t));
+
+  // not enough bytes to convert the second character, so only converts one
+  ASSERT_EQ(LIBC_NAMESPACE::wcsrtombs(mbs, &cur, 6, &state),
+            static_cast<size_t>(4));
+  ASSERT_ERRNO_SUCCESS();
+  ASSERT_EQ(cur, src + 1);
+  ASSERT_EQ(mbs[0], '\xF0');
+  ASSERT_EQ(mbs[1], '\x9F');
+  ASSERT_EQ(mbs[2], '\xA4');
+  ASSERT_EQ(mbs[3], '\xA1');
+  ASSERT_EQ(mbs[4], '\x01');
+}
+
+TEST_F(LlvmLibcWcsrtombs, NullDest) {
+  mbstate_t state;
+  LIBC_NAMESPACE::memset(&state, 0, sizeof(mbstate_t));
+
+  const wchar_t src[] = {static_cast<wchar_t>(0x1f921),
+                         static_cast<wchar_t>(0x2211),
+                         static_cast<wchar_t>(0xff), static_cast<wchar_t>(0x41),
+                         static_cast<wchar_t>(0x0)};
+  const wchar_t *cur = src;
+
+  // n parameter ignored when dest is null
+  ASSERT_EQ(LIBC_NAMESPACE::wcsrtombs(nullptr, &cur, 1, &state),
+            static_cast<size_t>(10));
+  ASSERT_ERRNO_SUCCESS();
+  ASSERT_EQ(cur, nullptr);
+
+  LIBC_NAMESPACE::memset(&state, 0, sizeof(mbstate_t));
+  ASSERT_EQ(LIBC_NAMESPACE::wcsrtombs(nullptr, &cur, 100, &state),
+            static_cast<size_t>(10));
+  ASSERT_ERRNO_SUCCESS();
+  ASSERT_EQ(cur, nullptr);
+}
+
+TEST_F(LlvmLibcWcsrtombs, ErrnoTest) {
+  mbstate_t state;
+  LIBC_NAMESPACE::memset(&state, 0, sizeof(mbstate_t));
+
+  const wchar_t src[] = {static_cast<wchar_t>(0x1f921),
+                         static_cast<wchar_t>(0x2211),
+                         static_cast<wchar_t>(0x12ffff), // invalid widechar
+                         static_cast<wchar_t>(0x0)};
+  const wchar_t *cur = src;
+
+  // n parameter ignored when dest is null
+  ASSERT_EQ(LIBC_NAMESPACE::wcsrtombs(nullptr, &cur, 7, &state),
+            static_cast<size_t>(7));
+  ASSERT_ERRNO_SUCCESS();
+
+  LIBC_NAMESPACE::memset(&state, 0, sizeof(mbstate_t));
+  ASSERT_EQ(LIBC_NAMESPACE::wcsrtombs(nullptr, &cur, 100, &state),
+            static_cast<size_t>(-1));
+  ASSERT_ERRNO_EQ(EILSEQ);
+}