[Scip] Constraint Handler / CONSENFOLP method

[Martin Tieves]

Hello SCIP team,

I implemented a very short version of a constraint handler for the traveling salesman problem. Hereby, I implemented the "conscheck" method (which just says a solution is feasible or not) and the "consenfolp" method.

Now to point of my question: Is it sufficient, just to return feasible or infeasible, in the enfolp method, again (especially not adding the violated inequalities, for saving programming effort - I just have to solve small instances)? If enfolp just returns feasible/infeasible, is the final solution still guaranteed to be optimal? What happens in detail (at the node) in the (mathematical) branch and bound tree, when a solution is recognised as infeasible?

Best Regards

Martin Tieves