string_decoder: Fix failures from new test cases

felixge · tjfontaine · commit 9fbd0f0f7df1 · 2014-06-06T15:07:29.000-07:00
This patch simplifies the implementation of StringDecoder, fixes the
failures from the new test cases, and also no longer relies on v8's
WriteUtf8 function to encode individual surrogates.
diff --git a/lib/string_decoder.js b/lib/string_decoder.js
@@ -57,29 +57,29 @@ var StringDecoder = exports.StringDecoder = function(encoding) {
 
 StringDecoder.prototype.write = function(buffer) {
   var charStr = '';
-  var offset = 0;
-
   // if our last write ended with an incomplete multibyte character
   while (this.charLength) {
     // determine how many remaining bytes this buffer has to offer for this char
-    var i = (buffer.length >= this.charLength - this.charReceived) ?
+    var available = (buffer.length >= this.charLength - this.charReceived) ?
                 this.charLength - this.charReceived :
                 buffer.length;
 
     // add the new bytes to the char buffer
-    buffer.copy(this.charBuffer, this.charReceived, offset, i);
-    this.charReceived += (i - offset);
-    offset = i;
+    buffer.copy(this.charBuffer, this.charReceived, 0, available);
+    this.charReceived += available;
 
     if (this.charReceived < this.charLength) {
       // still not enough chars in this buffer? wait for more ...
       return '';
     }
 
+    // remove bytes belonging to the current character from the buffer
+    buffer = buffer.slice(available, buffer.length);
+
     // get the character that was split
     charStr = this.charBuffer.slice(0, this.charLength).toString(this.encoding);
 
-    // lead surrogate (D800-DBFF) is also the incomplete character
+    // CESU-8: lead surrogate (D800-DBFF) is also the incomplete character
     var charCode = charStr.charCodeAt(charStr.length - 1);
     if (charCode >= 0xD800 && charCode <= 0xDBFF) {
       this.charLength += this.surrogateSize;
@@ -89,34 +89,33 @@ StringDecoder.prototype.write = function(buffer) {
     this.charReceived = this.charLength = 0;
 
     // if there are no more bytes in this buffer, just emit our char
-    if (i == buffer.length) return charStr;
-
-    // otherwise cut off the characters end from the beginning of this buffer
-    buffer = buffer.slice(i, buffer.length);
+    if (buffer.length === 0) {
+      return charStr;
+    }
     break;
   }
 
-  var lenIncomplete = this.detectIncompleteChar(buffer);
+  // determine and set charLength / charReceived
+  this.detectIncompleteChar(buffer);
 
   var end = buffer.length;
   if (this.charLength) {
     // buffer the incomplete character bytes we got
-    buffer.copy(this.charBuffer, 0, buffer.length - lenIncomplete, end);
-    this.charReceived = lenIncomplete;
-    end -= lenIncomplete;
+    buffer.copy(this.charBuffer, 0, buffer.length - this.charReceived, end);
+    end -= this.charReceived;
   }
 
   charStr += buffer.toString(this.encoding, 0, end);
 
   var end = charStr.length - 1;
   var charCode = charStr.charCodeAt(end);
-  // lead surrogate (D800-DBFF) is also the incomplete character
+  // CESU-8: lead surrogate (D800-DBFF) is also the incomplete character
   if (charCode >= 0xD800 && charCode <= 0xDBFF) {
     var size = this.surrogateSize;
     this.charLength += size;
     this.charReceived += size;
     this.charBuffer.copy(this.charBuffer, size, 0, size);
-    this.charBuffer.write(charStr.charAt(charStr.length - 1), this.encoding);
+    buffer.copy(this.charBuffer, 0, 0, size);
     return charStr.substring(0, end);
   }
 
@@ -153,8 +152,7 @@ StringDecoder.prototype.detectIncompleteChar = function(buffer) {
       break;
     }
   }
-
-  return i;
+  this.charReceived = i;
 };
 
 StringDecoder.prototype.end = function(buffer) {
@@ -177,13 +175,11 @@ function passThroughWrite(buffer) {
 }
 
 function utf16DetectIncompleteChar(buffer) {
-  var incomplete = this.charReceived = buffer.length % 2;
-  this.charLength = incomplete ? 2 : 0;
-  return incomplete;
+  this.charReceived = buffer.length % 2;
+  this.charLength = this.charReceived ? 2 : 0;
 }
 
 function base64DetectIncompleteChar(buffer) {
-  var incomplete = this.charReceived = buffer.length % 3;
-  this.charLength = incomplete ? 3 : 0;
-  return incomplete;
+  this.charReceived = buffer.length % 3;
+  this.charLength = this.charReceived ? 3 : 0;
 }
