util: add fast path for Latin1 decoding

PR-URL: #55275
Reviewed-By: Rafael Gonzaga <[email protected]>
Reviewed-By: Yagiz Nizipli <[email protected]>
Reviewed-By: James M Snell <[email protected]>
Reviewed-By: Daniel Lemire <[email protected]>

benchmark/util/text-decoder.js

@@ -3,7 +3,7 @@
 const common = require('../common.js');
 
 const bench = common.createBenchmark(main, {
-  encoding: ['utf-8', 'latin1', 'iso-8859-3'],
+  encoding: ['utf-8', 'windows-1252', 'iso-8859-3'],
   ignoreBOM: [0, 1],
   fatal: [0, 1],
   len: [256, 1024 * 16, 1024 * 128],

lib/internal/encoding.js

@@ -29,6 +29,7 @@ const kDecoder = Symbol('decoder');
 const kEncoder = Symbol('encoder');
 const kFatal = Symbol('kFatal');
 const kUTF8FastPath = Symbol('kUTF8FastPath');
+const kLatin1FastPath = Symbol('kLatin1FastPath');
 const kIgnoreBOM = Symbol('kIgnoreBOM');
 
 const {
@@ -55,6 +56,7 @@ const {
   encodeIntoResults,
   encodeUtf8String,
   decodeUTF8,
+  decodeLatin1,
 } = binding;
 
 const { Buffer } = require('buffer');
@@ -419,9 +421,10 @@ function makeTextDecoderICU() {
       this[kFatal] = Boolean(options?.fatal);
       // Only support fast path for UTF-8.
       this[kUTF8FastPath] = enc === 'utf-8';
+      this[kLatin1FastPath] = enc === 'windows-1252';
       this[kHandle] = undefined;
 
-      if (!this[kUTF8FastPath]) {
+      if (!this[kUTF8FastPath] && !this[kLatin1FastPath]) {
         this.#prepareConverter();
       }
     }
@@ -438,11 +441,16 @@ function makeTextDecoderICU() {
       validateDecoder(this);
 
       this[kUTF8FastPath] &&= !(options?.stream);
+      this[kLatin1FastPath] &&= !(options?.stream);
 
       if (this[kUTF8FastPath]) {
         return decodeUTF8(input, this[kIgnoreBOM], this[kFatal]);
       }
 
+      if (this[kLatin1FastPath]) {
+        return decodeLatin1(input, this[kIgnoreBOM], this[kFatal]);
+      }
+
       this.#prepareConverter();
 
       validateObject(options, 'options', kValidateObjectAllowObjectsAndNull);

src/encoding_binding.cc

@@ -1,6 +1,7 @@
 #include "encoding_binding.h"
 #include "ada.h"
 #include "env-inl.h"
+#include "node_buffer.h"
 #include "node_errors.h"
 #include "node_external_reference.h"
 #include "simdutf.h"
@@ -226,6 +227,7 @@ void BindingData::CreatePerIsolateProperties(IsolateData* isolate_data,
   SetMethodNoSideEffect(isolate, target, "decodeUTF8", DecodeUTF8);
   SetMethodNoSideEffect(isolate, target, "toASCII", ToASCII);
   SetMethodNoSideEffect(isolate, target, "toUnicode", ToUnicode);
+  SetMethodNoSideEffect(isolate, target, "decodeLatin1", DecodeLatin1);
 }
 
 void BindingData::CreatePerContextProperties(Local<Object> target,
@@ -243,6 +245,50 @@ void BindingData::RegisterTimerExternalReferences(
   registry->Register(DecodeUTF8);
   registry->Register(ToASCII);
   registry->Register(ToUnicode);
+  registry->Register(DecodeLatin1);
+}
+
+void BindingData::DecodeLatin1(const FunctionCallbackInfo<Value>& args) {
+  Environment* env = Environment::GetCurrent(args);
+
+  CHECK_GE(args.Length(), 1);
+  if (!(args[0]->IsArrayBuffer() || args[0]->IsSharedArrayBuffer() ||
+        args[0]->IsArrayBufferView())) {
+    return node::THROW_ERR_INVALID_ARG_TYPE(
+        env->isolate(),
+        "The \"input\" argument must be an instance of ArrayBuffer, "
+        "SharedArrayBuffer, or ArrayBufferView.");
+  }
+
+  bool ignore_bom = args[1]->IsTrue();
+  bool has_fatal = args[2]->IsTrue();
+
+  ArrayBufferViewContents<uint8_t> buffer(args[0]);
+  const uint8_t* data = buffer.data();
+  size_t length = buffer.length();
+
+  if (ignore_bom && length > 0 && data[0] == 0xFF) {
+    data++;
+    length--;
+  }
+
+  if (length == 0) {
+    return args.GetReturnValue().SetEmptyString();
+  }
+
+  std::string result(length * 2, '\0');
+
+  size_t written = simdutf::convert_latin1_to_utf8(
+      reinterpret_cast<const char*>(data), length, result.data());
+
+  if (has_fatal && written == 0) {
+    return node::THROW_ERR_ENCODING_INVALID_ENCODED_DATA(
+        env->isolate(), "The encoded data was not valid for encoding latin1");
+  }
+
+  Local<Object> buffer_result =
+      node::Buffer::Copy(env, result.c_str(), written).ToLocalChecked();
+  args.GetReturnValue().Set(buffer_result);
 }
 
 }  // namespace encoding_binding

src/encoding_binding.h

@@ -31,6 +31,7 @@ class BindingData : public SnapshotableObject {
   static void EncodeInto(const v8::FunctionCallbackInfo<v8::Value>& args);
   static void EncodeUtf8String(const v8::FunctionCallbackInfo<v8::Value>& args);
   static void DecodeUTF8(const v8::FunctionCallbackInfo<v8::Value>& args);
+  static void DecodeLatin1(const v8::FunctionCallbackInfo<v8::Value>& args);
 
   static void ToASCII(const v8::FunctionCallbackInfo<v8::Value>& args);
   static void ToUnicode(const v8::FunctionCallbackInfo<v8::Value>& args);

test/cctest/test_encoding_binding.cc

@@ -0,0 +1,155 @@
+#include "encoding_binding.h"
+#include "env-inl.h"
+#include "gtest/gtest.h"
+#include "node_test_fixture.h"
+#include "v8.h"
+
+namespace node {
+namespace encoding_binding {
+
+bool RunDecodeLatin1(Environment* env,
+                     Local<Value> args[],
+                     bool ignore_bom,
+                     bool has_fatal,
+                     Local<Value>* result) {
+  Isolate* isolate = env->isolate();
+  TryCatch try_catch(isolate);
+
+  Local<Boolean> ignoreBOMValue = Boolean::New(isolate, ignore_bom);
+  Local<Boolean> fatalValue = Boolean::New(isolate, has_fatal);
+
+  Local<Value> updatedArgs[] = {args[0], ignoreBOMValue, fatalValue};
+
+  BindingData::DecodeLatin1(FunctionCallbackInfo<Value>(updatedArgs));
+
+  if (try_catch.HasCaught()) {
+    return false;
+  }
+
+  *result = try_catch.Exception();
+  return true;
+}
+
+class EncodingBindingTest : public NodeTestFixture {};
+
+TEST_F(EncodingBindingTest, DecodeLatin1_ValidInput) {
+  Environment* env = CreateEnvironment();
+  Isolate* isolate = env->isolate();
+  HandleScope handle_scope(isolate);
+
+  const uint8_t latin1_data[] = {0xC1, 0xE9, 0xF3};
+  Local<ArrayBuffer> ab = ArrayBuffer::New(isolate, sizeof(latin1_data));
+  memcpy(ab->GetBackingStore()->Data(), latin1_data, sizeof(latin1_data));
+
+  Local<Uint8Array> array = Uint8Array::New(ab, 0, sizeof(latin1_data));
+  Local<Value> args[] = {array};
+
+  Local<Value> result;
+  EXPECT_TRUE(RunDecodeLatin1(env, args, false, false, &result));
+
+  String::Utf8Value utf8_result(isolate, result);
+  EXPECT_STREQ(*utf8_result, "Áéó");
+}
+
+TEST_F(EncodingBindingTest, DecodeLatin1_EmptyInput) {
+  Environment* env = CreateEnvironment();
+  Isolate* isolate = env->isolate();
+  HandleScope handle_scope(isolate);
+
+  Local<ArrayBuffer> ab = ArrayBuffer::New(isolate, 0);
+  Local<Uint8Array> array = Uint8Array::New(ab, 0, 0);
+  Local<Value> args[] = {array};
+
+  Local<Value> result;
+  EXPECT_TRUE(RunDecodeLatin1(env, args, false, false, &result));
+
+  String::Utf8Value utf8_result(isolate, result);
+  EXPECT_STREQ(*utf8_result, "");
+}
+
+TEST_F(EncodingBindingTest, DecodeLatin1_InvalidInput) {
+  Environment* env = CreateEnvironment();
+  Isolate* isolate = env->isolate();
+  HandleScope handle_scope(isolate);
+
+  Local<Value> args[] = {String::NewFromUtf8Literal(isolate, "Invalid input")};
+
+  Local<Value> result;
+  EXPECT_FALSE(RunDecodeLatin1(env, args, false, false, &result));
+}
+
+TEST_F(EncodingBindingTest, DecodeLatin1_IgnoreBOM) {
+  Environment* env = CreateEnvironment();
+  Isolate* isolate = env->isolate();
+  HandleScope handle_scope(isolate);
+
+  const uint8_t latin1_data[] = {0xFE, 0xFF, 0xC1, 0xE9, 0xF3};
+  Local<ArrayBuffer> ab = ArrayBuffer::New(isolate, sizeof(latin1_data));
+  memcpy(ab->GetBackingStore()->Data(), latin1_data, sizeof(latin1_data));
+
+  Local<Uint8Array> array = Uint8Array::New(ab, 0, sizeof(latin1_data));
+  Local<Value> args[] = {array};
+
+  Local<Value> result;
+  EXPECT_TRUE(RunDecodeLatin1(env, args, true, false, &result));
+
+  String::Utf8Value utf8_result(isolate, result);
+  EXPECT_STREQ(*utf8_result, "Áéó");
+}
+
+TEST_F(EncodingBindingTest, DecodeLatin1_FatalInvalidInput) {
+  Environment* env = CreateEnvironment();
+  Isolate* isolate = env->isolate();
+  HandleScope handle_scope(isolate);
+
+  const uint8_t invalid_data[] = {0xFF, 0xFF, 0xFF};
+  Local<ArrayBuffer> ab = ArrayBuffer::New(isolate, sizeof(invalid_data));
+  memcpy(ab->GetBackingStore()->Data(), invalid_data, sizeof(invalid_data));
+
+  Local<Uint8Array> array = Uint8Array::New(ab, 0, sizeof(invalid_data));
+  Local<Value> args[] = {array};
+
+  Local<Value> result;
+  EXPECT_FALSE(RunDecodeLatin1(env, args, false, true, &result));
+}
+
+TEST_F(EncodingBindingTest, DecodeLatin1_IgnoreBOMAndFatal) {
+  Environment* env = CreateEnvironment();
+  Isolate* isolate = env->isolate();
+  HandleScope handle_scope(isolate);
+
+  const uint8_t latin1_data[] = {0xFE, 0xFF, 0xC1, 0xE9, 0xF3};
+  Local<ArrayBuffer> ab = ArrayBuffer::New(isolate, sizeof(latin1_data));
+  memcpy(ab->GetBackingStore()->Data(), latin1_data, sizeof(latin1_data));
+
+  Local<Uint8Array> array = Uint8Array::New(ab, 0, sizeof(latin1_data));
+  Local<Value> args[] = {array};
+
+  Local<Value> result;
+  EXPECT_TRUE(RunDecodeLatin1(env, args, true, true, &result));
+
+  String::Utf8Value utf8_result(isolate, result);
+  EXPECT_STREQ(*utf8_result, "Áéó");
+}
+
+TEST_F(EncodingBindingTest, DecodeLatin1_BOMPresent) {
+  Environment* env = CreateEnvironment();
+  Isolate* isolate = env->isolate();
+  HandleScope handle_scope(isolate);
+
+  const uint8_t latin1_data[] = {0xFF, 0xC1, 0xE9, 0xF3};
+  Local<ArrayBuffer> ab = ArrayBuffer::New(isolate, sizeof(latin1_data));
+  memcpy(ab->GetBackingStore()->Data(), latin1_data, sizeof(latin1_data));
+
+  Local<Uint8Array> array = Uint8Array::New(ab, 0, sizeof(latin1_data));
+  Local<Value> args[] = {array};
+
+  Local<Value> result;
+  EXPECT_TRUE(RunDecodeLatin1(env, args, true, false, &result));
+
+  String::Utf8Value utf8_result(isolate, result);
+  EXPECT_STREQ(*utf8_result, "Áéó");
+}
+
+}  // namespace encoding_binding
+}  // namespace node