Re: [cclib-devel] Separation of parsers and data

[Adam T. <a-t...@st...>]

> Before you do anything, you should realise that this code is tightly
> linked to PyMOlyze, so  unless Adam agrees...

I agree that it makes sense to pass cclibData (or even attributes of  
cclibData) to the calculation methods instead of the parsers. Right  
now, I think it would be best to pass cclibData instead of attributes  
so that the interface for the methods remain as similar as possible.  
For instance: CSPA doesn't need overlaps, while most other  
populations analyses do.

Also, will the methods add the calculated attributes to the  
cclibData? The current implementations allow a FragmentAnalysis  
object to be passed directly to any of the population analysis  
methods, which is quite convenient for determining which fragment MOs  
contribute to the MOs of the entire molecule.

Adam