Bug 489820 part 2 - Grow buffers to the worst-case size before tokenizing; fix comments. r=wchen.

parser/html/javasrc/UTF16Buffer.java

@@ -117,6 +117,15 @@ public final class UTF16Buffer {
         return start < end;
     }
 
+    /**
+     * Returns <code>end - start</code>.
+     *
+     * @return <code>end - start</code>
+     */
+    public int getLength() {
+        return end - start;
+    }
+
     /**
      * Adjusts the start index to skip over the first character if it is a line
      * feed and the previous character was a carriage return.