XOM API for Processing XML with Java

[Wolfgang Hoschek <whoschek AT lbl.gov>]

I see. However, it seems you can reformulate the check as follows

            else if (this.getParent() != null && child.getChildCount() 
> 0 && isCycle(child, this.getParent())) {
                throw new CycleException(
                  "Cannot add an ancestor as a child");
            }

Now building a tree bottom-up this.getParent() == null and there should 
be now difference to top down building.
Wolfgang.

On Nov 30, 2004, at 1:35 AM, Elliotte Harold wrote:

> Wolfgang Hoschek wrote:
>
> > Can anyone give an example situation where
> >                 throw new CycleException(
> >                   "Cannot add an ancestor as a child");
> > is reachable in the method below? I can't see how it could 
> > theoretically be reached, given that child.getParent() != null has 
> > already been tested, and it's only for cases where child.isElement? 
> > An ancestor always has getParent != null, otherwise it can't be an 
> > ancestor in the first place. If it is indeed unreachable, the whole 
> > block
>
> This one came up because someone did it in their code, and I wrote a 
> unit test to prove  it. See 
> http://www.xom.nu/clover_html/nu/xom/Element.html
> Note the 100% test coverage for Element.java including the referenced 
> line. I believe it's tested by ParentNodeTest.testCycle:
>
>     public void testCycle() {
>
>         Element a = new Element("test");
>         Element b = new Element("test");
>         try {
>             a.appendChild(b);
>             b.appendChild(a);
>             fail("Allowed cycle");
>         }
>         catch (CycleException success) {
>             assertNotNull(success.getMessage());
>         }
>
>     }
>
> --
> Elliotte Rusty Harold  elharo AT metalab.unc.edu
> XML in a Nutshell 3rd Edition Just Published!
> http://www.cafeconleche.org/books/xian3/
> http://www.amazon.com/exec/obidos/ISBN=0596007647/cafeaulaitA/ref=nosim