From de2287e224c3d26e7fd1a540f20d767e03dfe86e Mon Sep 17 00:00:00 2001
From: Fedor Indutny <fedor@indutny.com>
Date: Wed, 24 Jan 2018 14:21:15 -0500
Subject: [PATCH] buffer: port `byteLengthUtf8` to JavaScript

Prior to this change the majority of the time spent when
calling `Buffer.byteLength` was spent on crossing JS->C++ boundary. This
change move the function to JavaScript, making it much faster.
---
 lib/buffer.js      | 37 ++++++++++++++++++++++++++++++++++++-
 src/node_buffer.cc |  8 --------
 2 files changed, 36 insertions(+), 9 deletions(-)

diff --git a/lib/buffer.js b/lib/buffer.js
index 4b800039fe6f80..56acacf14c1278 100644
--- a/lib/buffer.js
+++ b/lib/buffer.js
@@ -22,7 +22,6 @@
 'use strict';
 
 const {
-  byteLengthUtf8,
   copy: _copy,
   compare: _compare,
   compareOffset,
@@ -320,6 +319,42 @@ function allocate(size) {
 }
 
 
+// Ported from deps/v8/src/unicode-inl.h
+function isTrailSurrogate(code) {
+  return (code & 0xfc00) === 0xdc00;
+}
+
+
+function isLeadSurrogate(code) {
+  // No previous character
+  if (code === -1) return false;
+  return (code & 0xfc00) === 0xd800;
+}
+
+
+function byteLengthUtf8(string) {
+  var len = 0;
+  var previous = -1;
+  for (var i = 0; i < string.length; i++) {
+    // NOTE: 0 <= code <= 0xffff
+    var code = string.charCodeAt(i);
+    if (code <= 0x7f) {
+      len++;
+    } else if (code <= 0x7ff) {
+      len += 2;
+    } else {
+      // kSizeOfUnmatchedSurrogate - kBytesSavedByCombiningSurrogates
+      if (isTrailSurrogate(code) && isLeadSurrogate(previous))
+        len += 1;
+      else
+        len += 3;
+    }
+    previous = code;
+  }
+  return len;
+}
+
+
 function fromString(string, encoding) {
   var length;
   if (typeof encoding !== 'string' || encoding.length === 0) {
diff --git a/src/node_buffer.cc b/src/node_buffer.cc
index 72776304c8eb70..964a9288a686f9 100644
--- a/src/node_buffer.cc
+++ b/src/node_buffer.cc
@@ -785,13 +785,6 @@ void WriteDoubleBE(const FunctionCallbackInfo<Value>& args) {
 }
 
 
-void ByteLengthUtf8(const FunctionCallbackInfo<Value> &args) {
-  CHECK(args[0]->IsString());
-
-  // Fast case: avoid StringBytes on UTF8 string. Jump to v8.
-  args.GetReturnValue().Set(args[0].As<String>()->Utf8Length());
-}
-
 // Normalize val to be an integer in the range of [1, -1] since
 // implementations of memcmp() can vary by platform.
 static int normalizeCompareVal(int val, size_t a_length, size_t b_length) {
@@ -1214,7 +1207,6 @@ void Initialize(Local<Object> target,
   env->SetMethod(target, "setupBufferJS", SetupBufferJS);
   env->SetMethod(target, "createFromString", CreateFromString);
 
-  env->SetMethod(target, "byteLengthUtf8", ByteLengthUtf8);
   env->SetMethod(target, "copy", Copy);
   env->SetMethod(target, "compare", Compare);
   env->SetMethod(target, "compareOffset", CompareOffset);