Add Python API for importing SUMO ensembles via fmu-sumo package

[Copilot]

ResInsight's direct Elastic search queries to SUMO are fragile. This adds a Python-driven integration path using the fmu-sumo package for metadata access, keeping ResInsight's existing parquet-based data loading intact.

Changes

• RimcProject.h/cpp — New importSummaryEnsembleSumo scriptable PDM method on RimProject. Accepts case_id, case_name, ensemble_name, vector_names, and realization_ids from Python; creates a RimSummarySumoDataSource + RimSummaryEnsembleSumo and registers them in the project. Auto-exposed to Python as project.import_summary_ensemble_sumo(...).

• RimCloudDataSourceCollection — Added addSumoDataSource() to support programmatic data source creation outside the UI workflow.

• sumo_ensemble_import.py — Example script demonstrating the pattern: fmu-sumo handles case/ensemble discovery and metadata (the fragile part), ResInsight handles lazy parquet loading.

Usage

from fmu.sumo.explorer import Explorer
import rips

sumo = Explorer(env="prod")
cases = sumo.cases.filter(asset="Drogon")
case = cases[0]
iteration = case.iterations[0]

realization_ids = [str(r) for r in iteration.realizations]
vector_names = [col for col in iteration.tables.filter(tagname="summary")[0].column_names
                if col not in {"DATE", "REAL", "ENSEMBLE"}]

project = rips.Instance.find().project
ensemble = project.import_summary_ensemble_sumo(
    case_id=case.uuid,
    case_name=case.name,
    ensemble_name=iteration.name,
    vector_names=vector_names,
    realization_ids=realization_ids,
)