Re: [Cdk-devel] state of the QSAR molecular descriptor...

[Rajarshi G. <rg...@in...>]

On Dec 20, 2007, at 10:40 AM, Egon Willighagen wrote:

> Hi all,
>
> got some more reading for you, this time on the state of the QSAR
> descriptors in the CDK. If you use them, or written any one of them,
> you should read:
>
> http://chem-bla-ics.blogspot.com/2007/12/molecular-qsar-descriptors- 
> in-cdk.html

One other issue, not directly related to the QSAR stuff, is the  
coverage tester (that checks for untested methods).

I think that automatically generating names based on method args and  
looking for those test methods is not a great idea. On the other  
hand, we had previously discussed the use of annotations and I think  
that's the way we should go.

Specifically:

1. A given class is annotated with the name of its corresponding test  
class
2. Each method of the class is annotated with the name of the method  
in the test class

Coverage checker checks

1. A class has a test class annotation
2. All public methods have annotations
3. All method annotations exists in the test class annotation

This would be much more flexible and general.

An example is the annotations in the pcore module and the associated  
CoverageAnnotationTest.jaav file and its subclasses.

-------------------------------------------------------------------
Rajarshi Guha  <rg...@in...>
GPG Fingerprint: 0CCA 8EE2 2EEB 25E2 AB04  06F7 1BB9 E634 9B87 56EE
-------------------------------------------------------------------
Q:  What's purple and commutes?
A:  An abelian grape.