cleanup

Author: uzairnawaz

libc/src/__support/wchar/character_converter.cpp

@@ -31,27 +31,31 @@ bool CharacterConverter::isComplete() {
 }
 
 int CharacterConverter::push(char32_t utf32) {
+  // we can't be partially through a conversion when pushing a utf32 value
+  if (!isComplete())
+    return -1;
+
   state->partial = utf32;
   state->bytes_processed = 0;
-  state->total_bytes = 0;
 
   // determine number of utf-8 bytes needed to represent this utf32 value
-  constexpr char32_t ranges[] = {0x7f, 0x7ff, 0xffff, 0x10ffff};
-  constexpr int num_ranges = 4;
-  for (uint8_t i = 0; i < num_ranges; i++) {
-    if (state->partial <= ranges[i]) {
+  constexpr char32_t MAX_VALUE_PER_UTF8_LEN[] = {0x7f, 0x7ff, 0xffff, 0x10ffff};
+  constexpr int NUM_RANGES = 4;
+  for (uint8_t i = 0; i < NUM_RANGES; i++) {
+    if (state->partial <= MAX_VALUE_PER_UTF8_LEN[i]) {
       state->total_bytes = i + 1;
-      break;
+      return 0;
     }
   }
-  if (state->total_bytes == 0)
-    return -1;
 
-  return 0;
+  // `utf32` contains a value that is too large to actually represent a valid
+  // unicode character
+  clear();
+  return -1;
 }
 
 ErrorOr<char8_t> CharacterConverter::pop_utf8() {
-  if (state->bytes_processed >= state->total_bytes)
+  if (isComplete())
     return Error(-1);
 
   constexpr char8_t FIRST_BYTE_HEADERS[] = {0, 0xC0, 0xE0, 0xF0};

libc/test/src/__support/wchar/utf32_to_8_test.cpp

@@ -162,3 +162,19 @@ TEST(LlvmLibcCharacterConverterUTF32To8Test, FourByte) {
   popped = cr.pop_utf8();
   ASSERT_FALSE(popped.has_value());
 }
+
+TEST(LlvmLibcCharacterConverterUTF32To8Test, CantPushMidConversion) {
+  LIBC_NAMESPACE::internal::mbstate state;
+  LIBC_NAMESPACE::internal::CharacterConverter cr(&state);
+  cr.clear();
+
+  // testing utf32: 0x12121 -> utf8: 0xf0 0x92 0x84 0xa1
+  char32_t utf32 = 0x12121;
+  ASSERT_EQ(cr.push(utf32), 0);
+  auto popped = cr.pop_utf8();
+  ASSERT_TRUE(popped.has_value());
+
+  // can't push a utf32 without finishing popping the utf8 bytes out
+  int err = cr.push(utf32);
+  ASSERT_EQ(err, -1);
+}