mapping solver parameters

[hugolarzabal]

In cornflow client, we now have a mapping of solvers parameters here:
https://github.com/baobabsoluciones/cornflow/blob/master/libs/client/cornflow_client/constants.py

However, we are still missing a function to use this mapping in order to prepare the options to pass to the solver.

In order to have a standard way to do it, I propose to add a method like the following in ExperimentCore:

def get_solver_config(self, config, lib="pyomo", default_solver="cbc"):
        """
        Format the configuration to use to solver the problem.
        Solver configuration can either be directly in config using cornflow mapping name 
           or in a config["solver_config"] using the solver names.
        Example:
            config = {
                "solver":"milp.cbc",
                "time_limit":60,
                "rel_gap":0.1,
                "solver_config":{"heur":1, "pumpC":0}
            }

        :param config: dict config argument of the solver method
        :param lib: str library used to create the model (pulp or pyomo)
        :param default_solver: str default solver to use if none is present inf config.
        :return: the solver name and the config dict.
        """
        mapping = PARAMETER_SOLVER_TRANSLATING_MAPPING
        solver = config.get("solver", None)
        if "." in solver:
            solver = solver.split(".")[1]
        if solver is None:
            message = f"No solver found in config. Default solver {default_solver} will be used"
            print(message)
            solver = default_solver

        conf = {
            mapping[k, lib, solver]: v for k, v in config if (k, lib, solver) in mapping
        }
        conf.update(config.get("solver_config", {}))

        return solver, conf