Thread: [cclib-devel] Molpro parser

SourceForge Headquarters 225 Broadway Suite 1600 San Diego, CA 92101 +1 (858) 422-6466

What is the situation with the Molpro parser - is anyone working on that?

Karol

-- 
written by Karol Langner
Tue Feb 13 12:43:15 CET 2007

Mehdi Bounouar started work on that late last year, but there hasn't
been much activity on it since then, except to move it to a branch.

Are you familiar with Molpro? If so, perhaps you know the answers to
the following questions.  I read on the web something about the output
files of Molpro being described in a sort of mini-language in the
input file...do you know anything about this? Rumour has it that
Molpro can write data in XML format...this is the way of the future,
and would make it trivial to parse.

Noel

On 13/02/07, Karol Langner <kar...@kn...> wrote:
> What is the situation with the Molpro parser - is anyone working on that?
>
> Karol
>
> --
> written by Karol Langner
> Tue Feb 13 12:43:15 CET 2007
>
> -------------------------------------------------------------------------
> Using Tomcat but need to do more? Need to support web services, security?
> Get stuff done quickly with pre-integrated technology to make your job easier.
> Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> _______________________________________________
> cclib-devel mailing list
> ccl...@li...
> https://lists.sourceforge.net/lists/listinfo/cclib-devel
>

Hello,

Long time lurker, first time poster here.

On 13 Feb 2007, at 12:49, Noel O'Boyle wrote:
>
> Are you familiar with Molpro? If so, perhaps you know the answers to
> the following questions.  I read on the web something about the output
> files of Molpro being described in a sort of mini-language in the
> input file...do you know anything about this? Rumour has it that
> Molpro can write data in XML format...this is the way of the future,
> and would make it trivial to parse.

Molpro does have a mini scripting-type language that you can use for  
the input. In addition to running loops over user-defined arrays you  
can store calculated information in variables for processing at the  
end of the job.

It can also provide some output in XML format, but I've never really  
played with it.

For reference, I'm interested in contributing to cclib but I'm only  
just picking up the basics of python. If it goes well you may hear  
more from me in the future.

Grant

Hello Grant,

On 13/02/07, Grant Hill <Hi...@ca...> wrote:
> Hello,
>
> Long time lurker, first time poster here.
>
> On 13 Feb 2007, at 12:49, Noel O'Boyle wrote:
> >
> > Are you familiar with Molpro? If so, perhaps you know the answers to
> > the following questions.  I read on the web something about the output
> > files of Molpro being described in a sort of mini-language in the
> > input file...do you know anything about this? Rumour has it that
> > Molpro can write data in XML format...this is the way of the future,
> > and would make it trivial to parse.
>
> Molpro does have a mini scripting-type language that you can use for
> the input. In addition to running loops over user-defined arrays you
> can store calculated information in variables for processing at the
> end of the job.

I wonder does this mean that it's not possible for us to parse these
output files. Rather we would need to parse the 'checkpoint file' or
an XML file or something else (which I think is what Mehdi suggested
but I wasn't really familiar with these problems at that stage).

> It can also provide some output in XML format, but I've never really
> played with it.
I know that PMR (Peter Murray-Rust) is pushing strongly for comp chem
packages to output data in XML, and I had the impression that Molpro
was somehow taking the lead (in the quan chem world at least) in this.
The Molpro website has little information.

> For reference, I'm interested in contributing to cclib but I'm only
> just picking up the basics of python. If it goes well you may hear
> more from me in the future.
We would welcome any contributions you can make. If you have any
questions about our code, how it works, and so on, feel free to ask.

>
>
> Grant
>

On 13 Feb 2007, at 13:24, Noel O'Boyle wrote:

>>
>> Molpro does have a mini scripting-type language that you can use for
>> the input. In addition to running loops over user-defined arrays you
>> can store calculated information in variables for processing at the
>> end of the job.
>
> I wonder does this mean that it's not possible for us to parse these
> output files. Rather we would need to parse the 'checkpoint file' or
> an XML file or something else (which I think is what Mehdi suggested
> but I wasn't really familiar with these problems at that stage).

I would imagine it will certainly make things more difficult, I'm  
guessing it should be somewhat similar to parsing a Gaussian output  
that contains several calculations linked together. But, that is a  
guess.

>> It can also provide some output in XML format, but I've never really
>> played with it.
> I know that PMR (Peter Murray-Rust) is pushing strongly for comp chem
> packages to output data in XML, and I had the impression that Molpro
> was somehow taking the lead (in the quan chem world at least) in this.
> The Molpro website has little information.

The website isn't always the most useful place for information on  
Molpro ;-) I have easy access to the package so I'll try to have a  
play with the xml options. I currently don't even know if it outputs  
CML or conforms to some other custom DTD.

Grant

Hi Grant!

 I can encourage you by saying that I starting helping out with cclib very 
recently. In my case, since that time I have also been using cclib steadily 
more often in my research work.

As to Molpro, I haven't used it too much, but I do have access and sometimes 
run multireference jobs, since they go pretty fast. I haven't even heard 
anything about the XML output, but that would be sooo nice if all programs 
printed xml output :). On a side note, maybe saving xml files of parsed data 
would be a nice feature for cclib in the future?

> > I wonder does this mean that it's not possible for us to parse these
> > output files. Rather we would need to parse the 'checkpoint file' or
> > an XML file or something else (which I think is what Mehdi suggested
> > but I wasn't really familiar with these problems at that stage).
>
> I would imagine it will certainly make things more difficult, I'm
> guessing it should be somewhat similar to parsing a Gaussian output
> that contains several calculations linked together. But, that is a
> guess.
I don't think that means it's not possible to parse these output files, given 
the parsing function is properly contructed. After all, certain blocks in the 
output still have a constant structure, and that's what the parsers use.

Cheers,
Karol

-- 
written by Karol Langner
Wed Feb 14 10:46:09 CET 2007