Also set extension and metadata on copies of job outputs when finishi…

…ng job

lib/galaxy/jobs/__init__.py

@@ -67,6 +67,7 @@
     Task,
 )
 from galaxy.model.base import transaction
+from galaxy.model.store import copy_dataset_instance_metadata_attributes
 from galaxy.model.store.discover import MaxDiscoveredFilesExceededError
 from galaxy.objectstore import ObjectStorePopulator
 from galaxy.structured_app import MinimalManagerApp
@@ -1949,9 +1950,7 @@ def fail(message=job.info, exception=None):
                 ]
 
             for dataset_assoc in output_dataset_associations:
-                if getattr(dataset_assoc.dataset, "discovered", False):
-                    # skip outputs that have been discovered
-                    continue
+                is_discovered_dataset = getattr(dataset_assoc.dataset, "discovered", False)
                 context = self.get_dataset_finish_context(job_context, dataset_assoc)
                 # should this also be checking library associations? - can a library item be added from a history before the job has ended? -
                 # lets not allow this to occur
@@ -1960,6 +1959,12 @@ def fail(message=job.info, exception=None):
                     dataset_assoc.dataset.dataset.history_associations
                     + dataset_assoc.dataset.dataset.library_associations
                 ):
+                    if is_discovered_dataset:
+                        if dataset is dataset_assoc.dataset:
+                            continue
+                        elif dataset.extension == dataset_assoc.dataset.extension or dataset.extension == "auto":
+                            copy_dataset_instance_metadata_attributes(dataset_assoc.dataset, dataset)
+                            continue
                     output_name = dataset_assoc.name
 
                     # Handles retry internally on error for instance...

lib/galaxy/model/store/__init__.py

@@ -478,12 +478,11 @@ def handle_dataset_object_edit(dataset_instance, dataset_attrs):
                         if (
                             dataset_association is not dataset_instance
                             and dataset_association.extension == dataset_instance.extension
+                            or dataset_association.extension == "auto"
                         ):
-                            dataset_association.metadata = dataset_instance.metadata
-                            dataset_association.blurb = dataset_instance.blurb
-                            dataset_association.peek = dataset_instance.peek
-                            dataset_association.info = dataset_instance.info
-                            dataset_association.tool_version = dataset_instance.tool_version
+                            copy_dataset_instance_metadata_attributes(
+                                source=dataset_instance, target=dataset_association
+                            )
                 if job:
                     dataset_instance.dataset.job_id = job.id
 
@@ -3085,3 +3084,12 @@ def payload_to_source_uri(payload) -> str:
             dump(store_dict, f)
         source_uri = f"file://{import_json}"
     return source_uri
+
+
+def copy_dataset_instance_metadata_attributes(source: model.DatasetInstance, target: model.DatasetInstance) -> None:
+    target.metadata = source.metadata
+    target.blurb = source.blurb
+    target.peek = source.peek
+    target.info = source.info
+    target.tool_version = source.tool_version
+    target.extension = source.extension

lib/galaxy/tools/__init__.py

@@ -2875,7 +2875,7 @@ def exec_after_process(self, app, inp_data, out_data, param_dict, job=None, **kw
                         break
                 if copy_object is None:
                     raise exceptions.MessageException("Failed to find dataset output.")
-                out_data[key].copy_from(copy_object)
+                out_data[key].copy_from(copy_object, include_metadata=True)
 
     def parse_environment_variables(self, tool_source):
         """Setup environment variable for inputs file."""