Add utf16 implementation for Hermes runtime

Summary:
Adds the utf16 implementation for SH runtime. This makes use of the SH
StringView implementation which already stores the ASCII or UTF16 string
content.

Grafted from 9525c490fcad813f3973a0463d685d9419db844a (D65366621)
- Grafted path xplat/static_h to xplat/hermes

Reviewed By: tmikov

Differential Revision: D67412930

fbshipit-source-id: cdc31ec91cbf6e6befd342b6f9b9a14d57a40ea8

API/hermes/hermes.cpp

@@ -613,6 +613,9 @@ class HermesRuntimeImpl final : public HermesRuntime,
   jsi::String createStringFromUtf8(const uint8_t *utf8, size_t length) override;
   std::string utf8(const jsi::String &) override;
 
+  std::u16string utf16(const jsi::String &str) override;
+  std::u16string utf16(const jsi::PropNameID &sym) override;
+
   jsi::Value createValueFromJsonUtf8(const uint8_t *json, size_t length)
       override;
 
@@ -1781,6 +1784,27 @@ std::string HermesRuntimeImpl::utf8(const jsi::String &str) {
       vm::StringPrimitive::createStringView(runtime_, stringHandle(str)));
 }
 
+std::u16string HermesRuntimeImpl::utf16(const jsi::String &str) {
+  auto *stringPrim = phv(str).getString();
+  if (stringPrim->isASCII()) {
+    auto arrayRef = stringPrim->getStringRef<char>();
+    return std::u16string(arrayRef.begin(), arrayRef.end());
+  }
+  auto arrayRef = stringPrim->getStringRef<char16_t>();
+  return std::u16string(arrayRef.data(), arrayRef.size());
+}
+
+std::u16string HermesRuntimeImpl::utf16(const jsi::PropNameID &sym) {
+  vm::SymbolID id = phv(sym).getSymbol();
+  auto *stringPrim = runtime_.getStringPrimFromSymbolID(id);
+  if (stringPrim->isASCII()) {
+    auto arrayRef = stringPrim->getStringRef<char>();
+    return std::u16string(arrayRef.begin(), arrayRef.end());
+  }
+  auto arrayRef = stringPrim->getStringRef<char16_t>();
+  return std::u16string(arrayRef.data(), arrayRef.size());
+}
+
 jsi::Value HermesRuntimeImpl::createValueFromJsonUtf8(
     const uint8_t *json,
     size_t length) {

unittests/API/APITest.cpp

@@ -1167,6 +1167,37 @@ TEST_P(HermesRuntimeTest, NativeExceptionDoesNotUseGlobalError) {
       test.call(*rt).getString(*rt).utf8(*rt));
 }
 
+TEST_P(HermesRuntimeTest, UTF16Test) {
+  String ascii = String::createFromUtf8(*rt, "z");
+  EXPECT_EQ(ascii.utf16(*rt), u"z");
+
+  String foobar = String::createFromUtf8(*rt, "foobar");
+  EXPECT_EQ(foobar.utf16(*rt), u"foobar");
+
+  String chineseHello = String::createFromUtf8(*rt, "你好");
+  EXPECT_EQ(chineseHello.utf16(*rt), u"你好");
+
+  String thumbsUpEmoji = String::createFromUtf8(*rt, "👍");
+  EXPECT_EQ(thumbsUpEmoji.utf16(*rt), u"👍");
+
+  String combined = String::createFromUtf8(*rt, "foobar👍你好");
+  EXPECT_EQ(combined.utf16(*rt), u"foobar👍你好");
+
+  // We've only added specific implementations for HermesRuntime. The ABI
+  // runtime will convert to UTF8 first, causing the lone surrogates to be
+  // replaced with the Unicode replacement character. We will eventually add the
+  // UTF16-related APIs to the ABI Runtime and test these cases as well.
+  if (dynamic_cast<HermesRuntime *>(rt.get())) {
+    // Thumbs up emoji is encoded as 0xd83d 0xdc4d. These test UTF16 with lone
+    // high and low surrogates.
+    String loneHighSurrogate = eval("'\\ud83d'").getString(*rt);
+    EXPECT_EQ(loneHighSurrogate.utf16(*rt), std::u16string(u"\xd83d"));
+
+    String loneLowSurrogate = eval("'\\udc4d'").getString(*rt);
+    EXPECT_EQ(loneLowSurrogate.utf16(*rt), std::u16string(u"\xdc4d"));
+  }
+}
+
 INSTANTIATE_TEST_CASE_P(
     Runtimes,
     HermesRuntimeTest,