get wrong in saveEcmodel

[vivarock233-byte]

Yesterday I tried running the tutorial code with the complete model, and it worked perfectly.
But today it started having issues. The saveECModel function keeps throwing errors, even when I load a model that hasn't been modified at all and immediately save it.

Error using dispEM
The “subSystems” field must be a cell array

Error in checkModelStruct (Line 181)
            dispEM(EM,throwErrors);

Error in writeYAMLmodel (Line 42)
checkModelStruct(model);

Error in saveEcModel (Line 66)
        writeYAMLmodel(ecModel,filename);

the version of RAVEN is 2.10.5

[vivarock233-byte]

I have checked all the subsystems are the cell form, but it still went wrong

[edkerk]

I cannot replicate this bug, but the problem seems to be in the empty subSubsystems entries. A workaround (only works for the tutorial):

protRxns = contains(ecModel.rxns,'prot_');
ecModel.subSystems(protRxns) = ecModel.subSystems(protRxns) = {{'Protein usage reaction'}};

I will change that makeEcModel defines non-empty subsystems for the protein usage reactions.

This also highlights a bug in RAVEN's readYAMLmodel, which loads subSystems as nested cell arrays, while it should be "just" a cell array.

[edkerk]

Transferred to #412

[vivarock233-byte]

I find out a same condition when I used the Raven to construct a model and tried to save as a xml . I just find it will occur when the subSystems got empty rather a 'cell' form .
I use the this code to fix the bug

for i=1:numel(newModel.subSystems)
    if(isempty(newModel.subSystems{i}))
        fprintf("%d\n",i);
        newModel.subSystems{i}={''}
    end
end

I don't sure whether this method is right .