Various possible running changes & ideas for minor bugfixes

[exaexa]

converted from @stelmo's issue:

More small improvements, don't address these, I will when I am done adding to this list. However, feel free to comment on the changes, maybe I am wrong...

• Reaction has fields called lb and ub, change these to lower_bound and upper_bound to match the more verbose style we have adopted. ⚠️ this would be a breaking change
• make a set_objective! function
• change_constraint let it be possible to use kwargs to only change a specific bound
• flux_summary should not error when nothing is input, but rather display a message saying the solver didn't return a solution
• metabolite_fluxes why ignore zero flux metabolites? change the function to drop the dependence on _constants.tolerance
• change documentation for change_objective to reflect the fact that MOI is not necessarily imported e.g. change_objective(...; sense=MOI.MIN_SENSE) will error if JuMP is not imported, rather have the docstring read: change_objective(...; sense=COBREXA.MOI.MIN_SENSE)
• flux_summary should not error when there is no biomass or import or export reactions
• use @kwdef (see Add names field to genes, metabolites and reaction types. #566)

[stelmo]

I think this discussion is basically done, only kwdefs is really left. That could be something that goes into 2.0

[exaexa]

true, certainly will come up when redoing StdModel