"Steve Hickman" wrote:

> The problem may not be with toArray(). I've noticed that ArrayList will
> occasionally be improperly constructed, with nextItem being null when in fact
> I'm not at the end of the list. In these situations, the length of the
> ArrayList as shown in the .NET debugger shows more items than I can navigate
> to by opening nextItem. However, if I open prevItem and then the nextItem of
> the prevItem, I can always (so far at least) get to all the items on the
> list. This construction problem, when I've seen it, has always happened
> between the first and second items in the ArrayList.
>
> I haven't yet figured out the exact cause of this. The problem is showing up
> consistently for me right now, but the code where it shows up works the first
> 4 times it executes, then consistently fails the 5th time with an improperly
> constructed list. I'd provide a simple example if I could, but I don't have
> one at this time.
>
> "Keith McLeod" wrote:
>
> > We are writing a GIS application in C# and using Vivid Solutions’ Java
> > Topology Suite converted to J# for our Computational Geometry library. I am
> > receiving a null reference exception on the following line of code…
> >
> > object [] list = spatialIndex.query(new Envelope( minX + (ix * tileWidth),
> > minX + ((ix + 1) * tileWidth), maxY - (iy * tileHeight), maxY - ((iy + 1) *
> > tileHeight))).toArray();
> >
> > This is the method being called…
> > protected List query(Object searchBounds) {
> > if (!built) { build(); }
> > ArrayList matches = new ArrayList();
> > if (itemBoundables.isEmpty()) {
> > Assert.isTrue(root.getBounds() == null);
> > return matches;
> > }
> > if (getIntersectsOp().intersects(root.getBounds(), searchBounds)) {
> > query(searchBounds, root, matches);
> > }
> > return matches;
> > }
> >
> > I’ve stepped through the code and get all the way to the second "return
> > matches" statement. The matches object is indeed instantiated and has objects
> > (in this case they are ints) in it so it is definitely not null. This method
> > is used throughout the rest of the application and works fine. It just seems
> > to be the toArray method that is giving us a problem. I’m wondering if this
> > is a known bug in the J# runtime. We are using the J# redistributable package
> > 1.1 along with the J# supplemental UI library.
> >
> > One more bit of information, if I step through this code more than twice,
> > the exception then becomes a System.EngineException and I have to reboot my
> > computer.
> >