buffer: optimize toString()

PR-URL: #12361
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Anna Henningsen <anna@addaleax.net>

Author: mscdex

benchmark/buffers/buffer-tostring.js

@@ -3,25 +3,47 @@
 const common = require('../common.js');
 
 const bench = common.createBenchmark(main, {
-  arg: ['true', 'false'],
+  encoding: ['', 'utf8', 'ascii', 'latin1', 'binary', 'hex', 'UCS-2'],
+  args: [0, 1, 2, 3],
   len: [0, 1, 64, 1024],
   n: [1e7]
 });
 
 function main(conf) {
-  const arg = conf.arg === 'true';
+  var encoding = conf.encoding;
+  const args = conf.args | 0;
   const len = conf.len | 0;
   const n = conf.n | 0;
   const buf = Buffer.alloc(len, 42);
 
+  if (encoding.length === 0)
+    encoding = undefined;
+
   var i;
-  bench.start();
-  if (arg) {
-    for (i = 0; i < n; i += 1)
-      buf.toString('utf8');
-  } else {
-    for (i = 0; i < n; i += 1)
-      buf.toString();
+  switch (args) {
+    case 1:
+      bench.start();
+      for (i = 0; i < n; i += 1)
+        buf.toString(encoding);
+      bench.end(n);
+      break;
+    case 2:
+      bench.start();
+      for (i = 0; i < n; i += 1)
+        buf.toString(encoding, 0);
+      bench.end(n);
+      break;
+    case 3:
+      bench.start();
+      for (i = 0; i < n; i += 1)
+        buf.toString(encoding, 0, len);
+      bench.end(n);
+      break;
+    default:
+      bench.start();
+      for (i = 0; i < n; i += 1)
+        buf.toString();
+      bench.end(n);
+      break;
   }
-  bench.end(n);
 }

lib/buffer.js

@@ -483,82 +483,85 @@ Object.defineProperty(Buffer.prototype, 'offset', {
 });
 
 
-function slowToString(buf, encoding, start, end) {
-  var loweredCase = false;
-
-  // No need to verify that "buf.length <= MAX_UINT32" since it's a read-only
-  // property of a typed array.
-
-  // This behaves neither like String nor Uint8Array in that we set start/end
-  // to their upper/lower bounds if the value passed is out of range.
-  // undefined is handled specially as per ECMA-262 6th Edition,
-  // Section 13.3.3.7 Runtime Semantics: KeyedBindingInitialization.
-  if (start === undefined || start < 0)
-    start = 0;
-  // Return early if start > buf.length. Done here to prevent potential uint32
-  // coercion fail below.
-  if (start > buf.length)
-    return '';
-
-  if (end === undefined || end > buf.length)
-    end = buf.length;
-
-  if (end <= 0)
-    return '';
-
-  // Force coersion to uint32. This will also coerce falsey/NaN values to 0.
-  end >>>= 0;
-  start >>>= 0;
-
-  if (end <= start)
-    return '';
-
-  if (encoding === undefined) encoding = 'utf8';
-
-  while (true) {
-    switch (encoding) {
-      case 'hex':
-        return buf.hexSlice(start, end);
-
-      case 'utf8':
-      case 'utf-8':
-        return buf.utf8Slice(start, end);
-
-      case 'ascii':
-        return buf.asciiSlice(start, end);
-
-      case 'latin1':
-      case 'binary':
+function stringSlice(buf, encoding, start, end) {
+  if (encoding === undefined) return buf.utf8Slice(start, end);
+  encoding += '';
+  switch (encoding.length) {
+    case 4:
+      if (encoding === 'utf8') return buf.utf8Slice(start, end);
+      if (encoding === 'ucs2') return buf.ucs2Slice(start, end);
+      encoding = encoding.toLowerCase();
+      if (encoding === 'utf8') return buf.utf8Slice(start, end);
+      if (encoding === 'ucs2') return buf.ucs2Slice(start, end);
+      break;
+    case 5:
+      if (encoding === 'utf-8') return buf.utf8Slice(start, end);
+      if (encoding === 'ascii') return buf.asciiSlice(start, end);
+      if (encoding === 'ucs-2') return buf.ucs2Slice(start, end);
+      encoding = encoding.toLowerCase();
+      if (encoding === 'utf-8') return buf.utf8Slice(start, end);
+      if (encoding === 'ascii') return buf.asciiSlice(start, end);
+      if (encoding === 'ucs-2') return buf.ucs2Slice(start, end);
+      break;
+    case 6:
+      if (encoding === 'latin1' || encoding === 'binary')
         return buf.latin1Slice(start, end);
-
-      case 'base64':
-        return buf.base64Slice(start, end);
-
-      case 'ucs2':
-      case 'ucs-2':
-      case 'utf16le':
-      case 'utf-16le':
+      if (encoding === 'base64') return buf.base64Slice(start, end);
+      encoding = encoding.toLowerCase();
+      if (encoding === 'latin1' || encoding === 'binary')
+        return buf.latin1Slice(start, end);
+      if (encoding === 'base64') return buf.base64Slice(start, end);
+      break;
+    case 3:
+      if (encoding === 'hex' || encoding.toLowerCase() === 'hex')
+        return buf.hexSlice(start, end);
+      break;
+    case 7:
+      if (encoding === 'utf16le' || encoding.toLowerCase() === 'utf16le')
         return buf.ucs2Slice(start, end);
-
-      default:
-        if (loweredCase)
-          throw new TypeError('Unknown encoding: ' + encoding);
-        encoding = (encoding + '').toLowerCase();
-        loweredCase = true;
-    }
+      break;
+    case 8:
+      if (encoding === 'utf-16le' || encoding.toLowerCase() === 'utf-16le')
+        return buf.ucs2Slice(start, end);
+      break;
   }
+  throw new TypeError('Unknown encoding: ' + encoding);
 }
 
+
 Buffer.prototype.copy = function(target, targetStart, sourceStart, sourceEnd) {
   return binding.copy(this, target, targetStart, sourceStart, sourceEnd);
 };
 
+// No need to verify that "buf.length <= MAX_UINT32" since it's a read-only
+// property of a typed array.
+// This behaves neither like String nor Uint8Array in that we set start/end
+// to their upper/lower bounds if the value passed is out of range.
 Buffer.prototype.toString = function(encoding, start, end) {
-  let result;
+  var result;
   if (arguments.length === 0) {
     result = this.utf8Slice(0, this.length);
   } else {
-    result = slowToString(this, encoding, start, end);
+    const len = this.length;
+    if (len === 0)
+      return '';
+
+    if (!start || start < 0)
+      start = 0;
+    else if (start >= len)
+      return '';
+
+    if (end === undefined || end > len)
+      end = len;
+    else if (end <= 0)
+      return '';
+
+    start |= 0;
+    end |= 0;
+
+    if (end <= start)
+      return '';
+    result = stringSlice(this, encoding, start, end);
   }
   if (result === undefined)
     throw new Error('"toString()" failed');