fix a few typing errors

app_runner/src/app_runner/app_runner/runner.py

@@ -46,9 +46,13 @@ def run_process(self, chunk_dir: Path) -> None:
 
     def run_register_outputs(self, chunk_dir: Path, workunit_ref: int | Path, reuse_default_resource: bool) -> None:
         workunit_definition = WorkunitDefinition.from_ref(workunit_ref, client=self._client)
+        registration = workunit_definition.registration
+        if registration is None:
+            msg = "Workunit definition does not provide registration information"
+            raise ValueError(msg)
         register_outputs(
             outputs_yaml=chunk_dir / "outputs.yml",
-            workunit_id=workunit_definition.registration.workunit_id,
+            workunit_id=registration.workunit_id,
             client=self._client,
             ssh_user=self._ssh_user,
             reuse_default_resource=reuse_default_resource,

app_runner/src/app_runner/cli/outputs.py

@@ -1,6 +1,5 @@
 from __future__ import annotations
 
-
 import cyclopts
 
 from bfabric import Bfabric
@@ -32,6 +31,10 @@ def register(
 
     specs_list = OutputsSpec.read_yaml(outputs_yaml)
     workunit = Workunit.find(id=workunit_id, client=client)
+    if workunit is None:
+        msg = f"Workunit with id {workunit_id} not found"
+        raise ValueError(msg)
+
     register_all(
         client=client,
         workunit=workunit,

app_runner/src/app_runner/dispatch/dispatch_individual_resources.py

@@ -107,6 +107,9 @@ def _dispatch_jobs_dataset_flow(self, definition: WorkunitDefinition, params: di
         if config is None:
             raise ValueError("dataset_flow is not configured")
         dataset = Dataset.find(id=definition.execution.dataset, client=self._client)
+        if dataset is None:
+            msg = f"Dataset with id {definition.execution.dataset} not found"
+            raise ValueError(msg)
         dataset_df = dataset.to_polars()
         resources = Resource.find_all(ids=dataset_df[config.resource_column].unique().to_list(), client=self._client)
         paths = []

app_runner/src/app_runner/input_preparation/prepare.py

@@ -1,20 +1,20 @@
 from __future__ import annotations
 
+from typing import TYPE_CHECKING, Literal
 
+from bfabric.entities import Resource, Dataset
 from loguru import logger
 
-from bfabric.entities import Resource, Dataset
+from app_runner.input_preparation.integrity import IntegrityState
+from app_runner.input_preparation.list_inputs import list_input_states
 from app_runner.input_preparation.spec import (
     ResourceSpec,
     DatasetSpec,
     InputSpecType,
     InputsSpec,
 )
-from app_runner.input_preparation.integrity import IntegrityState
-from app_runner.input_preparation.list_inputs import list_input_states
 from app_runner.util.checksums import md5sum
 from app_runner.util.scp import scp
-from typing import TYPE_CHECKING, Literal
 
 if TYPE_CHECKING:
     from pathlib import Path
@@ -55,8 +55,13 @@ def clean_all(self, specs: list[InputSpecType]) -> None:
 
     def prepare_resource(self, spec: ResourceSpec) -> None:
         resource = Resource.find(id=spec.id, client=self._client)
+        if resource is None:
+            msg = f"Resource with id {spec.id} not found"
+            raise ValueError(msg)
 
         # determine path to copy from
+        # TODO as we have seen sometimes a faster approach would be to copy from the NFS mount, but this needs to be
+        #      configured or recognized somehow
         scp_uri = f"{resource.storage.scp_prefix}{resource['relativepath']}"
 
         # determine path to copy to

app_runner/src/app_runner/output_registration/register.py

@@ -127,6 +127,9 @@ def register_outputs(
 
     # register all specs
     workunit = Workunit.find(id=workunit_id, client=client)
+    if workunit is None:
+        msg = f"Workunit with id {workunit_id} not found"
+        raise ValueError(msg)
     register_all(
         client=client,
         workunit=workunit,