src: allow simdutf::convert_* functions to return zero

PR-URL: https://github.com/nodejs/node/pull/47471
Reviewed-By: Rich Trott <rtrott@gmail.com>
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
Reviewed-By: Luigi Pinca <luigipinca@gmail.com>

1 files changed, 28 insertions, 6 deletions

diff --git a/src/inspector/node_string.cc b/src/inspector/node_string.cc
index 171ba04bef..6b59cd73f9 100644
--- a/src/inspector/node_string.cc
+++ b/src/inspector/node_string.cc
@@ -19,12 +19,17 @@ void builderAppendQuotedString(StringBuilder& builder,
     size_t expected_utf16_length =
         simdutf::utf16_length_from_utf8(string.data(), string.length());
     MaybeStackBuffer<char16_t> buffer(expected_utf16_length);
+    // simdutf::convert_utf8_to_utf16 returns zero in case of error.
     size_t utf16_length = simdutf::convert_utf8_to_utf16(
         string.data(), string.length(), buffer.out());
-    CHECK_EQ(expected_utf16_length, utf16_length);
-    escapeWideStringForJSON(reinterpret_cast<const uint16_t*>(buffer.out()),
-                            utf16_length,
-                            &builder);
+    // We have that utf16_length == expected_utf16_length if and only
+    // if the input was a valid UTF-8 string.
+    if (utf16_length != 0) {
+      CHECK_EQ(expected_utf16_length, utf16_length);
+      escapeWideStringForJSON(reinterpret_cast<const uint16_t*>(buffer.out()),
+                              utf16_length,
+                              &builder);
+    }  // Otherwise, we had an invalid UTF-8 input.
   }
   builder.put('"');
 }
@@ -35,8 +40,12 @@ std::unique_ptr<Value> parseJSON(const std::string_view string) {
   size_t expected_utf16_length =
       simdutf::utf16_length_from_utf8(string.data(), string.length());
   MaybeStackBuffer<char16_t> buffer(expected_utf16_length);
+  // simdutf::convert_utf8_to_utf16 returns zero in case of error.
   size_t utf16_length = simdutf::convert_utf8_to_utf16(
       string.data(), string.length(), buffer.out());
+  // We have that utf16_length == expected_utf16_length if and only
+  // if the input was a valid UTF-8 string.
+  if (utf16_length == 0) return nullptr;  // We had an invalid UTF-8 input.
   CHECK_EQ(expected_utf16_length, utf16_length);
   return parseJSONCharacters(reinterpret_cast<const uint16_t*>(buffer.out()),
                              utf16_length);
@@ -62,9 +71,14 @@ String StringViewToUtf8(v8_inspector::StringView view) {
   size_t expected_utf8_length =
       simdutf::utf8_length_from_utf16(source, view.length());
   MaybeStackBuffer<char> buffer(expected_utf8_length);
+  // convert_utf16_to_utf8 returns zero in case of error.
   size_t utf8_length =
       simdutf::convert_utf16_to_utf8(source, view.length(), buffer.out());
-  CHECK_EQ(expected_utf8_length, utf8_length);
+  // We have that utf8_length == expected_utf8_length if and only
+  // if the input was a valid UTF-16 string. Otherwise, utf8_length
+  // must be zero.
+  CHECK(utf8_length == 0 || utf8_length == expected_utf8_length);
+  // An invalid UTF-16 input will generate the empty string:
   return String(buffer.out(), utf8_length);
 }
 
@@ -112,9 +126,14 @@ String fromUTF16(const uint16_t* data, size_t length) {
   size_t expected_utf8_length =
       simdutf::utf8_length_from_utf16(casted_data, length);
   MaybeStackBuffer<char> buffer(expected_utf8_length);
+  // simdutf::convert_utf16_to_utf8 returns zero in case of error.
   size_t utf8_length =
       simdutf::convert_utf16_to_utf8(casted_data, length, buffer.out());
-  CHECK_EQ(expected_utf8_length, utf8_length);
+  // We have that utf8_length == expected_utf8_length if and only
+  // if the input was a valid UTF-16 string. Otherwise, utf8_length
+  // must be zero.
+  CHECK(utf8_length == 0 || utf8_length == expected_utf8_length);
+  // An invalid UTF-16 input will generate the empty string:
   return String(buffer.out(), utf8_length);
 }
 
@@ -123,6 +142,9 @@ const uint8_t* CharactersUTF8(const std::string_view s) {
 }
 
 size_t CharacterCount(const std::string_view s) {
+  // The utf32_length_from_utf8 function calls count_utf8.
+  // The count_utf8 function counts the number of code points
+  // (characters) in the string, assuming that the string is valid Unicode.
   // TODO(@anonrig): Test to make sure CharacterCount returns correctly.
   return simdutf::utf32_length_from_utf8(s.data(), s.length());
 }