Merge pull request #5372 from cloudflare/jasnell/encodeinto-perf-improvement

Author: jasnell

src/workerd/api/encoding.c++

@@ -486,10 +486,13 @@ jsg::BufferSource TextEncoder::encode(jsg::Lock& js, jsg::Optional<jsg::JsString
 }
 
 TextEncoder::EncodeIntoResult TextEncoder::encodeInto(
-    jsg::Lock& js, jsg::JsString input, jsg::BufferSource buffer) {
-  auto handle = buffer.getHandle(js);
-  JSG_REQUIRE(handle->IsUint8Array(), TypeError, "buffer must be a Uint8Array");
-  return encodeIntoImpl(js, input, buffer);
+    jsg::Lock& js, jsg::JsString input, jsg::JsUint8Array buffer) {
+  auto result = input.writeInto(
+      js, buffer.asArrayPtr<char>(), jsg::JsString::WriteFlags::REPLACE_INVALID_UTF8);
+  return TextEncoder::EncodeIntoResult{
+    .read = static_cast<int>(result.read),
+    .written = static_cast<int>(result.written),
+  };
 }
 
 }  // namespace workerd::api

src/workerd/api/encoding.h

@@ -218,7 +218,7 @@ class TextEncoder final: public jsg::Object {
 
   jsg::BufferSource encode(jsg::Lock& js, jsg::Optional<jsg::JsString> input);
 
-  EncodeIntoResult encodeInto(jsg::Lock& js, jsg::JsString input, jsg::BufferSource buffer);
+  EncodeIntoResult encodeInto(jsg::Lock& js, jsg::JsString input, jsg::JsUint8Array buffer);
 
   // UTF-8 is the only encoding type supported by the WHATWG spec.
   kj::StringPtr getEncoding() {
@@ -239,6 +239,7 @@ class TextEncoder final: public jsg::Object {
     // `Uint8Array`. The spec defines that this function returns a `Uint8Array` too.
     JSG_TS_OVERRIDE({
       encode(input?: string): Uint8Array;
+      encodeInto(input: string, buffer: Uint8Array): TextEncoderEncodeIntoResult;
     });
   }
 };

src/workerd/jsg/jsg.h

@@ -2235,7 +2235,8 @@ class JsMessage;
   V(Set)                                                                                           \
   V(Promise)                                                                                       \
   V(Proxy)                                                                                         \
-  V(Function)
+  V(Function)                                                                                      \
+  V(Uint8Array)
 
 #define V(Name) class Js##Name;
 JS_TYPE_CLASSES(V)

src/workerd/jsg/jsvalue.h

@@ -44,7 +44,6 @@ inline void requireOnStack(void* self) {
   V(ArrayBuffer)                                                                                   \
   V(ArrayBufferView)                                                                               \
   V(TypedArray)                                                                                    \
-  V(Uint8Array)                                                                                    \
   V(Uint8ClampedArray)                                                                             \
   V(Int8Array)                                                                                     \
   V(Uint16Array)                                                                                   \
@@ -230,6 +229,20 @@ class JsArray final: public JsBase<v8::Array, JsArray> {
   using JsBase<v8::Array, JsArray>::JsBase;
 };
 
+class JsUint8Array final: public JsBase<v8::Uint8Array, JsUint8Array> {
+ public:
+  template <typename T = kj::byte>
+  kj::ArrayPtr<T> asArrayPtr() {
+    v8::Local<v8::Uint8Array> inner = *this;
+    auto buf = inner->Buffer();
+    T* data = static_cast<T*>(buf->Data()) + inner->ByteOffset();
+    size_t length = inner->ByteLength();
+    return kj::ArrayPtr(data, length);
+  }
+
+  using JsBase<v8::Uint8Array, JsUint8Array>::JsBase;
+};
+
 class JsString final: public JsBase<v8::String, JsString> {
  public:
   int length(Lock& js) const KJ_WARN_UNUSED_RESULT;

src/workerd/jsg/rtti.h

@@ -468,6 +468,7 @@ struct BuildRtti<Configuration, v8::Promise> {
   };
 
 #define FOR_EACH_BUILTIN_TYPE(F, ...)                                                              \
+  F(jsg::JsUint8Array, BuiltinType::Type::V8_UINT8_ARRAY)                                          \
   F(jsg::BufferSource, BuiltinType::Type::JSG_BUFFER_SOURCE)                                       \
   F(kj::Date, BuiltinType::Type::KJ_DATE)                                                          \
   F(v8::ArrayBufferView, BuiltinType::Type::V8_ARRAY_BUFFER_VIEW)                                  \

types/generated-snapshot/experimental/index.d.ts

@@ -1587,10 +1587,7 @@ declare class TextEncoder {
    *
    * [MDN Reference](https://developer.mozilla.org/docs/Web/API/TextEncoder/encodeInto)
    */
-  encodeInto(
-    input: string,
-    buffer: ArrayBuffer | ArrayBufferView,
-  ): TextEncoderEncodeIntoResult;
+  encodeInto(input: string, buffer: Uint8Array): TextEncoderEncodeIntoResult;
   get encoding(): string;
 }
 interface TextDecoderConstructorOptions {

types/generated-snapshot/experimental/index.ts

@@ -1592,10 +1592,7 @@ export declare class TextEncoder {
    *
    * [MDN Reference](https://developer.mozilla.org/docs/Web/API/TextEncoder/encodeInto)
    */
-  encodeInto(
-    input: string,
-    buffer: ArrayBuffer | ArrayBufferView,
-  ): TextEncoderEncodeIntoResult;
+  encodeInto(input: string, buffer: Uint8Array): TextEncoderEncodeIntoResult;
   get encoding(): string;
 }
 export interface TextDecoderConstructorOptions {

types/generated-snapshot/latest/index.d.ts

@@ -1542,10 +1542,7 @@ declare class TextEncoder {
    *
    * [MDN Reference](https://developer.mozilla.org/docs/Web/API/TextEncoder/encodeInto)
    */
-  encodeInto(
-    input: string,
-    buffer: ArrayBuffer | ArrayBufferView,
-  ): TextEncoderEncodeIntoResult;
+  encodeInto(input: string, buffer: Uint8Array): TextEncoderEncodeIntoResult;
   get encoding(): string;
 }
 interface TextDecoderConstructorOptions {

types/generated-snapshot/latest/index.ts

@@ -1547,10 +1547,7 @@ export declare class TextEncoder {
    *
    * [MDN Reference](https://developer.mozilla.org/docs/Web/API/TextEncoder/encodeInto)
    */
-  encodeInto(
-    input: string,
-    buffer: ArrayBuffer | ArrayBufferView,
-  ): TextEncoderEncodeIntoResult;
+  encodeInto(input: string, buffer: Uint8Array): TextEncoderEncodeIntoResult;
   get encoding(): string;
 }
 export interface TextDecoderConstructorOptions {