issues with ODE -> Reaction converter

[fbergmann]

I've tried with a basic brusselator example, converting the file first to ODE's then trying to run the converter to restore the reactions:

Original Reaction Scheme:

	#	Name	Reaction	Rate Law	Flux [mmol/s]	Number Flux [1/(s)]	Noise Expression
	1	R1	A -> X	Mass action (irreversible)	nan	nan	
	2	R2	2 * X + Y -> 3 * X	Mass action (irreversible)	nan	nan	
	3	R3	X + B -> Y + D	Mass action (irreversible)	nan	nan	
	4	R4	X -> E	Mass action (irreversible)	nan	nan	
		New Reaction					

after running the conversion to ode and back the resulting scheme was:

	#	Name	Reaction	Rate Law	Flux [mmol/s]	Number Flux [1/(s)]	Noise Expression
	1	J1	 -> X;  A	Function for J1	nan	nan	
	2	J2	X + Y -> X + Y	Function for J2	nan	nan	
	3	J3	X -> X + Y;  B	Function for J3	nan	nan	
	4	J4	X -> X	Function for J4	nan	nan	
		New Reaction					

Originally X and Y were floating, A, B, D and E boundary.

• The stoichiometries look strange to me. (since it was just mass action, it is fine that D and E do not appear in the reaction sche. I'm even fine with A and B being modifier rather than substrates.
• it is however problematic, that in the resulting model X and Y are also boundary.

I'll attach the ode version of the model.

BrusselatorODE.xml.txt