util: improve textdecoder decode performance

nodejs · Nov 3, 2022 · 1207719 · 1207719
1 parent f1e9382
commit 1207719
Show file tree

Hide file tree

Showing 3 changed files with 46 additions and 10 deletions.
diff --git a/benchmark/util/text-decoder.js b/benchmark/util/text-decoder.js
@@ -0,0 +1,17 @@
+'use strict';
+
+const common = require('../common.js');
+
+const bench = common.createBenchmark(main, {
+  len: [1, 64, 1024],
+  n: [1e6]
+});
+
+function main({ len, n }) {
+  const buf = Buffer.alloc(len, 42);
+  const decoder = new TextDecoder();
+
+  bench.start();
+  decoder.decode(buf);
+  bench.end(n);
+}
diff --git a/lib/internal/encoding.js b/lib/internal/encoding.js
@@ -432,7 +432,7 @@ function makeTextDecoderICU() {
       if (typeof ret === 'number') {
         throw new ERR_ENCODING_INVALID_ENCODED_DATA(this.encoding, ret);
       }
-      return ret.toString('ucs2');
+      return ret;
     }
   }
 

diff --git a/src/node_i18n.cc b/src/node_i18n.cc
@@ -50,6 +50,8 @@
 #include "node_buffer.h"
 #include "node_errors.h"
 #include "node_internals.h"
+#include "string_bytes.h"
+#include "string_decoder.h"
 #include "util-inl.h"
 #include "v8.h"
 
@@ -502,18 +504,35 @@ void ConverterObject::Decode(const FunctionCallbackInfo<Value>& args) {
       }
     }
     ret = ToBufferEndian(env, &result);
-    if (omit_initial_bom && !ret.IsEmpty()) {
-      // Perform `ret = ret.slice(2)`.
+
+    if (!ret.IsEmpty()) {
       CHECK(ret.ToLocalChecked()->IsUint8Array());
-      Local<Uint8Array> orig_ret = ret.ToLocalChecked().As<Uint8Array>();
-      ret = Buffer::New(env,
-                        orig_ret->Buffer(),
-                        orig_ret->ByteOffset() + 2,
-                        orig_ret->ByteLength() - 2)
+
+      if (omit_initial_bom) {
+        // Perform `ret = ret.slice(2)`.
+        Local<Uint8Array> orig_ret = ret.ToLocalChecked().As<Uint8Array>();
+        ret = Buffer::New(env,
+                          orig_ret->Buffer(),
+                          orig_ret->ByteOffset() + 2,
+                          orig_ret->ByteLength() - 2)
                             .FromMaybe(Local<Uint8Array>());
+      }
+
+      Local<Value> error;
+      ArrayBufferViewContents<char> buf(ret.ToLocalChecked());
+      MaybeLocal<Value> encoded = StringBytes::Encode(env->isolate(),
+                                                      buf.data(),
+                                                      buf.length(),
+                                                      encoding::UCS2,
+                                                      &error);
+
+      if (!encoded.IsEmpty()) {
+        args.GetReturnValue().Set(encoded.ToLocalChecked());
+      } else {
+        args.GetReturnValue().Set(error);
+      }
     }
-    if (!ret.IsEmpty())
-      args.GetReturnValue().Set(ret.ToLocalChecked());
+
     return;
   }