Fix unicode character usage in windows tests

Summary:
X-link: facebook/react-native#48840

MSVC unicode handling is causing our windows test to fail. To fix,
simply replace the special characters with the their explicit UTF-16
encoding.

Changelog: [Internal]

Reviewed By: neildhar

Differential Revision: D68466808

fbshipit-source-id: 1bfc689972e1e5862fe6525bc48d5ebc508a95da

API/jsi/jsi/test/testlib.cpp

@@ -1601,14 +1601,19 @@ TEST_P(JSITest, UTF16Test) {
   rd.utf8Str = "foobar";
   EXPECT_EQ(str.utf16(rd), u"foobar");
 
-  rd.utf8Str = "你好";
-  EXPECT_EQ(str.utf16(rd), u"你好");
+  // 你 in UTF-8 encoding is 0xe4 0xbd 0xa0 and 好 is 0xe5 0xa5 0xbd
+  // 你 in UTF-16 encoding is 0x4f60 and 好 is 0x597d
+  rd.utf8Str = "\xe4\xbd\xa0\xe5\xa5\xbd";
+  EXPECT_EQ(str.utf16(rd), u"\x4f60\x597d");
 
-  rd.utf8Str = "👍";
-  EXPECT_EQ(str.utf16(rd), u"👍");
+  // 👍 in UTF-8 encoding is 0xf0 0x9f 0x91 0x8d
+  // 👍 in UTF-16 encoding is 0xd83d 0xdc4d
+  rd.utf8Str = "\xf0\x9f\x91\x8d";
+  EXPECT_EQ(str.utf16(rd), u"\xd83d\xdc4d");
 
-  rd.utf8Str = "foobar👍你好";
-  EXPECT_EQ(str.utf16(rd), u"foobar👍你好");
+  // String is foobar👍你好
+  rd.utf8Str = "foobar\xf0\x9f\x91\x8d\xe4\xbd\xa0\xe5\xa5\xbd";
+  EXPECT_EQ(str.utf16(rd), u"foobar\xd83d\xdc4d\x4f60\x597d");
 
   // String ended before second byte of the encoding
   rd.utf8Str = "\xcf";
@@ -1652,7 +1657,8 @@ TEST_P(JSITest, GetStringDataTest) {
   };
 
   RD rd = RD(rt);
-  String str = String::createFromUtf8(rd, "hello👋");
+  // 👋 in UTF8 encoding is 0xf0 0x9f 0x91 0x8b
+  String str = String::createFromUtf8(rd, "hello\xf0\x9f\x91\x8b");
 
   std::u16string buf;
   auto cb = [&buf](bool ascii, const void* data, size_t num) {

unittests/API/APITest.cpp

@@ -1174,14 +1174,20 @@ TEST_P(HermesRuntimeTest, UTF16Test) {
   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");
+  // 你 in UTF-8 encoding is 0xe4 0xbd 0xa0 and 好 is 0xe5 0xa5 0xbd
+  // 你 in UTF-16 encoding is 0x4f60 and 好 is 0x597d
+  String chineseHello = String::createFromUtf8(*rt, "\xe4\xbd\xa0\xe5\xa5\xbd");
+  EXPECT_EQ(chineseHello.utf16(*rt), u"\x4f60\x597d");
 
-  String thumbsUpEmoji = String::createFromUtf8(*rt, "👍");
-  EXPECT_EQ(thumbsUpEmoji.utf16(*rt), u"👍");
+  // 👍 in UTF-8 encoding is 0xf0 0x9f 0x91 0x8d
+  // 👍 in UTF-16 encoding is 0xd83d 0xdc4d
+  String thumbsUpEmoji = String::createFromUtf8(*rt, "\xf0\x9f\x91\x8d");
+  EXPECT_EQ(thumbsUpEmoji.utf16(*rt), u"\xd83d\xdc4d");
 
-  String combined = String::createFromUtf8(*rt, "foobar👍你好");
-  EXPECT_EQ(combined.utf16(*rt), u"foobar👍你好");
+  // String is foobar👍你好
+  String combined = String::createFromUtf8(
+      *rt, "foobar\xf0\x9f\x91\x8d\xe4\xbd\xa0\xe5\xa5\xbd");
+  EXPECT_EQ(combined.utf16(*rt), u"foobar\xd83d\xdc4d\x4f60\x597d");
 
   // We've only added specific implementations for HermesRuntime. The ABI
   // runtime will convert to UTF8 first, causing the lone surrogates to be
@@ -1230,9 +1236,11 @@ TEST_P(HermesRuntimeTest, GetStringDataTest) {
   EXPECT_EQ(buf, u"fbar");
   buf.clear();
 
-  String utf16Str = String::createFromUtf8(*rt, "👍foobar你好");
+  // String is foobar👍你好
+  String utf16Str = String::createFromUtf8(
+      *rt, "foobar\xf0\x9f\x91\x8d\xe4\xbd\xa0\xe5\xa5\xbd");
   utf16Str.getStringData(*rt, cb);
-  EXPECT_EQ(buf, u"👍fbar你好");
+  EXPECT_EQ(buf, u"fbar\xd83d\xdc4d\x4f60\x597d");
   buf.clear();
 }
 
@@ -1268,9 +1276,11 @@ TEST_P(HermesRuntimeTest, GetPropNameIdDataTest) {
   EXPECT_EQ(buf, u"fbar");
   buf.clear();
 
-  PropNameID utf16 = PropNameID::forUtf8(*rt, "👍foobar你好");
+  // String is foobar👍你好
+  PropNameID utf16 = PropNameID::forUtf8(
+      *rt, "foobar\xf0\x9f\x91\x8d\xe4\xbd\xa0\xe5\xa5\xbd");
   utf16.getPropNameIdData(*rt, cb);
-  EXPECT_EQ(buf, u"👍fbar你好");
+  EXPECT_EQ(buf, u"fbar\xd83d\xdc4d\x4f60\x597d");
   buf.clear();
 }
 

unittests/API/SynthTraceParserTest.cpp

@@ -322,7 +322,7 @@ TEST_F(SynthTraceParserTest, ParseUtf16Record) {
 
   auto record2 =
       dynamic_cast<const SynthTrace::Utf16Record &>(*trace.records().at(2));
-  ASSERT_EQ(record2.retVal_, u"nice👍");
+  ASSERT_EQ(record2.retVal_, u"nice\xd83d\xdc4d");
 }
 
 TEST_F(SynthTraceParserTest, ParseGetStringDataRecord) {
@@ -357,7 +357,7 @@ TEST_F(SynthTraceParserTest, ParseGetStringDataRecord) {
 
   auto record0 = dynamic_cast<const SynthTrace::GetStringDataRecord &>(
       *trace.records().at(0));
-  ASSERT_EQ(record0.strData_, u"\nhello👋\\");
+  ASSERT_EQ(record0.strData_, u"\nhello\xd83d\xdc4b\\");
   ASSERT_EQ(record0.objID_, SynthTrace::encodeString(1110));
 
   auto record1 = dynamic_cast<const SynthTrace::GetStringDataRecord &>(

unittests/API/SynthTraceSerializationTest.cpp

@@ -401,7 +401,7 @@ TEST_F(SynthTraceSerializationTest, Utf16Record) {
   EXPECT_EQ(
       serialized,
       to_string(SynthTrace::Utf16Record(
-          dummyTime, SynthTrace::encodeString(123), u"hi👋")));
+          dummyTime, SynthTrace::encodeString(123), u"hi\xd83d\xdc4b")));
   serialized =
       R"({"type":"Utf16Record","time":0,"objID":"string:111","retval":"\ud83d"})";
   EXPECT_EQ(
@@ -416,7 +416,7 @@ TEST_F(SynthTraceSerializationTest, GetStringDataRecord) {
   EXPECT_EQ(
       serialized,
       to_string(SynthTrace::GetStringDataRecord(
-          dummyTime, SynthTrace::encodeString(123), u"\nhello👋\\")));
+          dummyTime, SynthTrace::encodeString(123), u"\nhello\xd83d\xdc4b\\")));
   serialized =
       R"({"type":"GetStringDataRecord","time":0,"objID":"propNameID:111","strData":"\ud83d"})";
   EXPECT_EQ(

unittests/API/SynthTraceTest.cpp

@@ -107,7 +107,8 @@ TEST_F(SynthTraceTest, PropNameIDUtf8) {
 }
 
 TEST_F(SynthTraceTest, PropNameIDUtf16) {
-  const std::string utf8 = "hello👍\n";
+  // 👍 in UTF8 encoding is 0xf0 0x9f 0x91 0x8d
+  const std::string utf8 = "hello\xf0\x9f\x91\x8d\n";
   const jsi::PropNameID name = jsi::PropNameID::forUtf8(*rt, utf8);
   name.utf16(*rt);
 
@@ -121,7 +122,9 @@ TEST_F(SynthTraceTest, PropNameIDUtf16) {
       *records[0]);
   EXPECT_EQ_RECORD(
       SynthTrace::Utf16Record(
-          records[1]->time_, SynthTrace::encodePropNameID(objId), u"hello👍\n"),
+          records[1]->time_,
+          SynthTrace::encodePropNameID(objId),
+          u"hello\xd83d\xdc4d\n"),
       *records[1]);
 }
 
@@ -145,7 +148,8 @@ TEST_F(SynthTraceTest, StringUtf8) {
 }
 
 TEST_F(SynthTraceTest, StringUtf16) {
-  const std::string utf8 = "hello👍\n";
+  // 👍 in UTF8 encoding is 0xf0 0x9f 0x91 0x8d
+  const std::string utf8 = "hello\xf0\x9f\x91\x8d\n";
 
   const jsi::String str = jsi::String::createFromUtf8(*rt, utf8);
   str.utf16(*rt);
@@ -159,13 +163,16 @@ TEST_F(SynthTraceTest, StringUtf16) {
       *records[0]);
   EXPECT_EQ_RECORD(
       SynthTrace::Utf16Record(
-          records[1]->time_, SynthTrace::encodeString(objId), u"hello👍\n"),
+          records[1]->time_,
+          SynthTrace::encodeString(objId),
+          u"hello\xd83d\xdc4d\n"),
       *records[1]);
 }
 
 TEST_F(SynthTraceTest, GetStringData) {
   const std::string ascii = "foo";
-  const std::string emoji = "hello👋";
+  // 👋 in UTF8 encoding is 0xf0 0x9f 0x91 0x8b
+  const std::string emoji = "hello\xf0\x9f\x91\x8b";
 
   auto cb = [](bool ascii, const void *data, size_t num) {};
 
@@ -197,7 +204,9 @@ TEST_F(SynthTraceTest, GetStringData) {
       *records[2]);
   EXPECT_EQ_RECORD(
       SynthTrace::GetStringDataRecord(
-          records[3]->time_, SynthTrace::encodeString(emojiId), u"hello👋"),
+          records[3]->time_,
+          SynthTrace::encodeString(emojiId),
+          u"hello\xd83d\xdc4b"),
       *records[3]);
 }
 

unittests/Support/JSONEmitterTest.cpp

@@ -228,7 +228,7 @@ TEST(JSONEmitterTest, EmitUTF16) {
   llvh::raw_string_ostream OS(storage);
   JSONEmitter json(OS);
 
-  std::u16string str = u"hi👋";
+  std::u16string str = u"hi\xd83d\xdc4b";
   json.openDict();
   json.emitKeyValue("str", llvh::ArrayRef(str.data(), str.size()));
   json.closeDict();