Import kinetics block into nineml.extensions.kinetics

[tclose]

In the files pype9.importer.neuron.nmodl.py I have included a NotImplementedError in the get_component_class method that should be expanded to output a KineticsClass object using the parsed kinetics structure (see commented out line)

[russelljjarvis]

I think it is appropriate for me to comment out or delete the line:
raise NotImplementedError,

uncomment the lines:
for name, (bidirectional, incoming, outgoing,
constraints, compartments) in self.kinetics.iteritems():

So it looks like this:
else:
for name, (bidirectional, incoming, outgoing,
constraints, compartments) in self.kinetics.iteritems():

And then I have to provide an appropriate body to the for loop.

So I should the test the get_component_class method in nmodl.py by attempting to import an NMODL file that contains a Kinetics block like Golgi_SK2.mod. I have located the file: /pype9/test/unittests/test_neuron_import.py which converts all of the files in:
/home/russell/git/pype9/test/data/nmodl, including the file Golgi_SK2.mod

The default argument to get_component_class is flatten_kinetics=False. So when I try to import the NMODL file, the unflattened version should be tested by default.

Then I would try to resolve any errors in the python implementation that result from doing this.

[russelljjarvis]

Regarding the recent code I pushed to the open pull request.

I forgot to mention: In line 200 of nmodl.py, there is probably a more eloquent way of unpacking the value I am indexing as cst[0][0], but I am still trying to figure that out.
cst[0][0] is the first state in the constraint list. If the file was specified in 9ML as opposed to NMODL it would be a user defined value subsituted to make an elimination in the kinetics equation, so I just assume here that its okay to automatically take the first state. Perhaps its not?

cst = Constraint(cst[0][0],ks[0].name)