Re: [Cdk-devel] Bonditerator

[John R. <jo...@zf...>]

Kai and Rajarshi,

Thanks for the quick response.  I had been working with a snapshot of 
the CDK development tree that was a couple months old, trying to 
stabilize my own code as I converted it to use the CDK.  I updated to 
the latest revision of CDK, and see that you have already addressed this 
problem.

Regards,
John

Kai Hartmann wrote:

> Hi John.
>
> Egon Willighagen schrieb:
>
>> On Monday 12 February 2007, Christoph Steinbeck wrote:
>>
>>> 1. Am I correct in setting up my radical object as an 
>>> IElectronContainer
>>> that is not an IBond?
>>
>>
>> Correct, radicals are implemented as ISingleElectron objects.
>
>
> Confirmed. IElectronContainer is the superclass of IBond, 
> ISingleElectron and ILonePair.
>
>>> 2. Should BondIterator really return any IElectronContainer?
>>
>>
>> No, only those of type IBond (Kai, please confirm)
>>
>
> Right.
>
>>> 3. If BondIterator should only return IBond objects, then do you 
>>> want me
>>> to fix the BondIterator so it skips over the non-IBond objects (instead
>>> of returning null objects)?  And, if so, do you want to also have an
>>> ElectronContainerIterator and corresponding electronContainers() 
>>> method?
>>
>
> If you want to iterate over all IElectronContainer objects (including 
> radicals and bonds), please use the method 
> IAtomContainer.electronContainers() (already implemented).
>
> The bonds() method should not return any null objects. Please write a 
> junit test and file a bug report for this. At the moment, I don't have 
> any explanation why this happens. Please don't just introduce a skip.
>
> Greetings,
>
> Kai
>
>
>
>