fixup! buffer: add buffer.isUtf8 for utf8 validation

nodejs · nodejs-github-bot · Dec 25, 2022 · Dec 22, 2022 · Dec 23, 2022 · Dec 23, 2022
commit bcb19ec2a9b0972023cdfe429c5f2bfcde28df98
diff --git a/doc/api/buffer.md b/doc/api/buffer.md
@@ -5137,7 +5137,7 @@ added: REPLACEME
 -->
 
 * input {Buffer | ArrayBuffer | TypedArray} The input to validate.
-* Returns: {boolean} Returns true if and only if the input is valid UTF-8.
+* Returns: {boolean} Returns `true` if and only if the input is valid UTF-8.
 
 This function is used to check if input contains UTF-8 code points (characters).
 

diff --git a/lib/buffer.js b/lib/buffer.js
@@ -1317,15 +1317,11 @@ function atob(input) {
 }
 
 function isUtf8(input) {
-  if (isTypedArray(input) || Buffer.isBuffer(input)) {
-    return bindingIsUtf8(input.buffer);
-  }
-
-  if (isAnyArrayBuffer(input)) {
+  if (isTypedArray(input) || isAnyArrayBuffer(input)) {
     return bindingIsUtf8(input);
   }
 
-  return false;
+  throw new ERR_INVALID_ARG_TYPE('input', ['TypedArray', 'Buffer'], input);
 }
 
 module.exports = {

diff --git a/src/node_buffer.cc b/src/node_buffer.cc
@@ -1224,12 +1224,32 @@ static void EncodeInto(const FunctionCallbackInfo<Value>& args) {
 }
 
 static void IsUtf8(const FunctionCallbackInfo<Value>& args) {
-  CHECK_GE(args.Length(), 1);
-  CHECK(args[0]->IsArrayBuffer());
-  Local<ArrayBuffer> input = args[0].As<ArrayBuffer>();
-  auto external = static_cast<const char*>(input->Data());
-  args.GetReturnValue().Set(
-      simdutf::validate_utf8(external, input->ByteLength()));
+  Environment* env = Environment::GetCurrent(args);
+  Isolate* isolate = env->isolate();
+
+  CHECK_EQ(args.Length(), 1);
+  CHECK(args[0]->IsTypedArray() || args[0]->IsArrayBuffer());
+
+  Local<ArrayBuffer> buf;
+  size_t offset = 0;
+  size_t length = 0;
+
+  if (args[0]->IsTypedArray()) {
+    Local<v8::TypedArray> input = args[0].As<v8::TypedArray>();
+    buf = input->Buffer();
+    offset = input->ByteOffset();
+    length = input->ByteLength();
+  } else {
+    buf = args[0].As<ArrayBuffer>();
+    length = buf->ByteLength();
+  }
+
+  if (buf->WasDetached()) {
+    return node::THROW_ERR_BUFFER_CONTEXT_NOT_AVAILABLE(isolate);
+  }
+
+  const char* external = static_cast<const char*>(buf->Data()) + offset;
+  args.GetReturnValue().Set(simdutf::validate_utf8(external, length));
 }
 
 void SetBufferPrototype(const FunctionCallbackInfo<Value>& args) {

diff --git a/test/parallel/test-buffer-isutf8.js b/test/parallel/test-buffer-isutf8.js
@@ -2,14 +2,39 @@
 
 require('../common');
 const assert = require('assert');
-const { isUtf8 } = require('buffer');
+const { isUtf8, Buffer } = require('buffer');
 const { TextEncoder } = require('util');
 
 const encoder = new TextEncoder();
 
 assert.strictEqual(isUtf8(encoder.encode('hello')), true);
 assert.strictEqual(isUtf8(encoder.encode('ğ')), true);
 assert.strictEqual(isUtf8(Buffer.from([0xf8])), false);
+assert.strictEqual(isUtf8(encoder.encode('aé日')), true);
 
-assert.strictEqual(isUtf8(null), false);
-assert.strictEqual(isUtf8(undefined), false);
+[
+  null,
+  undefined,
+  'hello',
+  true,
+  false
+].forEach((input) => {
+  assert.throws(
+    () => { isUtf8(input); },
+    {
+      code: 'ERR_INVALID_ARG_TYPE',
+    },
+  );
+})
+
+{
+  // Test with detached array buffers
+  const arrayBuffer = new ArrayBuffer(1024);
+  structuredClone(arrayBuffer, { transfer: [arrayBuffer] });
+  assert.throws(
+    () => { isUtf8(arrayBuffer); },
+  {
+      code: 'ERR_BUFFER_CONTEXT_NOT_AVAILABLE'
+    }
+  )
+}