🔊 Improve error message for MlflowModelTrackingDataset.save when a ru…

…n id is already active (#549)

kedro_mlflow/io/models/mlflow_model_tracking_dataset.py

@@ -102,11 +102,9 @@ def _save(self, model: Any) -> None:
             if mlflow.active_run():
                 # it is not possible to log in a run which is not the current opened one
                 raise DatasetError(
-                    "'run_id' cannot be specified"
-                    " if there is an mlflow active run."
-                    "Run_id mismatch: "
-                    f"\n - 'run_id'={self._run_id}"
-                    f"\n - active_run id={mlflow.active_run().info.run_id}"
+                    f"'run_id' cannot be specified (run_id='{self._run_id}') "
+                    f"if there is an mlflow active run (active run id='{mlflow.active_run().info.run_id}') "
+                    f"See the rationale in this issue: https://github.com/Galileo-Galilei/kedro-mlflow/issues/549."
                 )
             else:
                 # if the run id is specified and there is no opened run,

tests/io/models/test_mlflow_model_tracking_dataset.py

@@ -152,7 +152,7 @@ def test_save_sklearn_flavor_with_run_id_and_already_active_run(tracking_uri):
     with mlflow.start_run():
         with pytest.raises(
             DatasetError,
-            match="'run_id' cannot be specified if there is an mlflow active run.",
+            match="if there is an mlflow active run",
         ):
             mlflow_model_ds.save(linreg_model)