From 21211a3fa93993b363f25964da5dac3282b3b453 Mon Sep 17 00:00:00 2001
From: Yagiz Nizipli <yagiz@nizipli.com>
Date: Thu, 11 Apr 2024 07:41:44 -0400
Subject: [PATCH] buffer: improve `btoa` performance

PR-URL: https://github.com/nodejs/node/pull/52427
Reviewed-By: Daniel Lemire <daniel@lemire.me>
Reviewed-By: James M Snell <jasnell@gmail.com>
---
 benchmark/buffers/buffer-btoa.js | 20 +++++++++++
 lib/buffer.js                    | 11 +++---
 src/node_buffer.cc               | 59 ++++++++++++++++++++++++++++++++
 3 files changed, 84 insertions(+), 6 deletions(-)
 create mode 100644 benchmark/buffers/buffer-btoa.js

diff --git a/benchmark/buffers/buffer-btoa.js b/benchmark/buffers/buffer-btoa.js
new file mode 100644
index 00000000000000..3867d5890b1d79
--- /dev/null
+++ b/benchmark/buffers/buffer-btoa.js
@@ -0,0 +1,20 @@
+'use strict';
+const common = require('../common.js');
+const assert = require('node:assert');
+
+const bench = common.createBenchmark(main, {
+  size: [16, 32, 64, 128, 256, 1024],
+  n: [1e6],
+});
+
+function main({ n, size }) {
+  const input = 'A'.repeat(size);
+  let out = 0;
+
+  bench.start();
+  for (let i = 0; i < n; i++) {
+    out += btoa(input).length;
+  }
+  bench.end(n);
+  assert(out > 0);
+}
diff --git a/lib/buffer.js b/lib/buffer.js
index ea94ebf24192f9..d66d5650f30e73 100644
--- a/lib/buffer.js
+++ b/lib/buffer.js
@@ -69,6 +69,7 @@ const {
   kMaxLength,
   kStringMaxLength,
   atob: _atob,
+  btoa: _btoa,
 } = internalBinding('buffer');
 const {
   constants: {
@@ -1249,13 +1250,11 @@ function btoa(input) {
   if (arguments.length === 0) {
     throw new ERR_MISSING_ARGS('input');
   }
-  input = `${input}`;
-  for (let n = 0; n < input.length; n++) {
-    if (input[n].charCodeAt(0) > 0xff)
-      throw lazyDOMException('Invalid character', 'InvalidCharacterError');
+  const result = _btoa(`${input}`);
+  if (result === -1) {
+    throw lazyDOMException('Invalid character', 'InvalidCharacterError');
   }
-  const buf = Buffer.from(input, 'latin1');
-  return buf.toString('base64');
+  return result;
 }
 
 function atob(input) {
diff --git a/src/node_buffer.cc b/src/node_buffer.cc
index b31beada451bc8..e63318b65b2e61 100644
--- a/src/node_buffer.cc
+++ b/src/node_buffer.cc
@@ -1211,6 +1211,63 @@ void DetachArrayBuffer(const FunctionCallbackInfo<Value>& args) {
   }
 }
 
+static void Btoa(const FunctionCallbackInfo<Value>& args) {
+  CHECK_EQ(args.Length(), 1);
+  Environment* env = Environment::GetCurrent(args);
+  THROW_AND_RETURN_IF_NOT_STRING(env, args[0], "argument");
+
+  Local<String> input = args[0].As<String>();
+  MaybeStackBuffer<char> buffer;
+  size_t written;
+
+  if (input->IsExternalOneByte()) {  // 8-bit case
+    auto ext = input->GetExternalOneByteStringResource();
+    size_t expected_length = simdutf::base64_length_from_binary(ext->length());
+    buffer.AllocateSufficientStorage(expected_length + 1);
+    buffer.SetLengthAndZeroTerminate(expected_length);
+    written =
+        simdutf::binary_to_base64(ext->data(), ext->length(), buffer.out());
+  } else if (input->IsOneByte()) {
+    MaybeStackBuffer<uint8_t> stack_buf(input->Length());
+    input->WriteOneByte(env->isolate(),
+                        stack_buf.out(),
+                        0,
+                        input->Length(),
+                        String::NO_NULL_TERMINATION);
+
+    size_t expected_length =
+        simdutf::base64_length_from_binary(input->Length());
+    buffer.AllocateSufficientStorage(expected_length + 1);
+    buffer.SetLengthAndZeroTerminate(expected_length);
+    written =
+        simdutf::binary_to_base64(reinterpret_cast<const char*>(*stack_buf),
+                                  input->Length(),
+                                  buffer.out());
+  } else {
+    String::Value value(env->isolate(), input);
+    MaybeStackBuffer<char> stack_buf(value.length());
+    size_t out_len = simdutf::convert_utf16_to_latin1(
+        reinterpret_cast<const char16_t*>(*value),
+        value.length(),
+        stack_buf.out());
+    if (out_len == 0) {  // error
+      return args.GetReturnValue().Set(-1);
+    }
+    size_t expected_length = simdutf::base64_length_from_binary(out_len);
+    buffer.AllocateSufficientStorage(expected_length + 1);
+    buffer.SetLengthAndZeroTerminate(expected_length);
+    written = simdutf::binary_to_base64(*stack_buf, out_len, buffer.out());
+  }
+
+  auto value =
+      String::NewFromOneByte(env->isolate(),
+                             reinterpret_cast<const uint8_t*>(buffer.out()),
+                             NewStringType::kNormal,
+                             written)
+          .ToLocalChecked();
+  return args.GetReturnValue().Set(value);
+}
+
 // In case of success, the decoded string is returned.
 // In case of error, a negative value is returned:
 // * -1 indicates a single character remained,
@@ -1329,6 +1386,7 @@ void Initialize(Local<Object> target,
   Isolate* isolate = env->isolate();
 
   SetMethodNoSideEffect(context, target, "atob", Atob);
+  SetMethodNoSideEffect(context, target, "btoa", Btoa);
 
   SetMethod(context, target, "setBufferPrototype", SetBufferPrototype);
   SetMethodNoSideEffect(context, target, "createFromString", CreateFromString);
@@ -1433,6 +1491,7 @@ void RegisterExternalReferences(ExternalReferenceRegistry* registry) {
   registry->Register(CopyArrayBuffer);
 
   registry->Register(Atob);
+  registry->Register(Btoa);
 }
 
 }  // namespace Buffer