Populate Reactions finds different rate from RMG job when there are multiple rate rules

[connie]

RMG will use the rate rule with the higher rank, however PopulateReactions seems to randomly assign the rate rule to the reaction. For instance in my example with Cineole + OH, we the following reaction when running Populate Reactions:

Cineole(1) + OH(2) --> C11H19J(9) + H2O(22)          1.700e+06  1.90   -1.45  
!H_Abstraction exact:   [ C/H/Cs3 , O_pri_rad ] deltaHrxn(T=298K) = -17.40 kcal/mol

The same reaction pops up in an RMG job but gets different kinetics:

Cineole(1)+OH(28)=R5(18)+H2O(99)                     2.570e+06    1.90    1.45  
!H_Abstraction exact:   [ C/H/Cs3 , O_pri_rad ]

Upon closer inspection in the H_abstraction family we see that there exists two rules with different rankings. The RMG job selects the correct rate rule to use, but PopulateReactions does not.

[rwest]

It would be good if they both used the same bit of code to do this, so they always did the same thing. IIRC the RMG-Java project has a lot of code duplication.