Re: yet another graph question

[Feng QIAN <fqian@sable.mcgill.ca>]

Hi Florian,

I am not so sure the full reason of such design, but following comment in
AbstractTrap.java may proof that BlockGraph does right thing.

    /** The unit just before the last unit being trapped. */
    protected UnitBox endUnitBox;

Cheers,
===========================================================
Feng Qian                             fqian@sable.mcgill.ca

On Tue, 12 Mar 2002, Florian Loitsch wrote:

> hi,
> sorry to bother you again with some code of this BlockGraph.java file...
>
> in BlockGraph.java (line 120) there's the following code snippet:
> ====
>   // Get the leaders that bound exception contexts.
>   if(type == ZONED) {
>     List predList = new ArrayList();
>     predList.add(mUnits.getPredOf(someTrap.getBeginUnit()));
>     leaders.put(someTrap.getBeginUnit(), predList);
>
>     predList = new ArrayList();
>     predList.add(mUnits.getPredOf(someTrap.getEndUnit()));
>     leaders.put(someTrap.getEndUnit(), predList);
>   }
> ====
> according to the java-doc of Trap.java the getEndUnit()-element is still in
> the exception-scope.
> my question is now:
> why is the trap.getEndUnit()-element the header of a new block?
> wouldn't it be more logical to say that the successor (if any) of the
> trap.getEndUnit()-element should be a new block?
> mfg
>  florian loitsch
> --
> People who think they know everything are very irritating to those of us who do.
> - Anon.
>