XOM API for Processing XML with Java

[Elliotte Rusty Harold <elharo AT metalab.unc.edu>]

At 11:44 AM +0200 9/25/03, Hallvard Trætteberg wrote:


> I also have a suggestion regarding Element's implementation of
> Node.getValue(). It currently accumulates the childs' values in a
> StringBuffer. In many cases there is only one child, and this could be
> handled as a special case (for speed and reduced memory consumption), as
> follows:
>
>         if (getChildCount() == 1) {
>             Node child = getChild(0);
>             if (child.isText() || child.isElement()) {
>                 return child.getValue();
>             }
>         }
>         StringBuffer result = new StringBuffer();
>         for (int i = 0; i < getChildCount(); i++) {
>                 ...
>         }
>         return result.toString();
>     }
>
> Of course, this complicates the code and may not increase the speed
> much. However, the code is still fairly obvious, so why not?

Do you have any profiling data to indicate that this change is worthwhile?
--

  Elliotte Rusty Harold
  elharo AT metalab.unc.edu
  Processing XML with Java (Addison-Wesley, 2002)
  http://www.cafeconleche.org/books/xmljava
  http://www.amazon.com/exec/obidos/ISBN%3D0201771861/cafeaulaitA