Merge pull request #18078 from mvdbeek/better_exception_extract_dataset

[24.0] Improve error message for ``Extract dataset`` tool

lib/galaxy/celery/tasks.py

@@ -288,6 +288,7 @@ def abort_when_job_stops(function: Callable, session: galaxy_scoped_session, job
                 return future.result(timeout=1)
             except TimeoutError:
                 if is_aborted(session, job_id):
+                    future.cancel()
                     return
 
 

lib/galaxy/tools/__init__.py

@@ -3356,9 +3356,15 @@ def produce_outputs(self, trans, out_data, output_collections, incoming, history
         if how == "first":
             extracted_element = collection.first_dataset_element
         elif how == "by_identifier":
-            extracted_element = collection[incoming["which"]["identifier"]]
+            try:
+                extracted_element = collection[incoming["which"]["identifier"]]
+            except KeyError as e:
+                raise exceptions.MessageException(e.args[0])
         elif how == "by_index":
-            extracted_element = collection[int(incoming["which"]["index"])]
+            try:
+                extracted_element = collection[int(incoming["which"]["index"])]
+            except KeyError as e:
+                raise exceptions.MessageException(e.args[0])
         else:
             raise exceptions.MessageException("Invalid tool parameters.")
         extracted = extracted_element.element_object

lib/galaxy/tools/extract_dataset.xml

@@ -19,7 +19,7 @@
       </param>
       <when value="first" />
       <when value="by_identifier">
-        <param name="identifier" label="Element identifier:" type="text">
+        <param name="identifier" label="Element identifier:" type="text" optional="false">
           <sanitizer invalid_char="">
               <valid initial="string.ascii_letters,string.digits">
                   <add value="_" />
@@ -52,9 +52,9 @@ Description
 
 The tool allow extracting datasets based on position (**The first dataset** and **Select by index** options) or name (**Select by element identifier** option). This tool effectively collapses the inner-most collection into a dataset. For nested collections (e.g a list of lists of lists: outer:middle:inner, extracting the inner dataset element) a new list is created where the selected element takes the position of the inner-most collection (so outer:middle, where middle is not a collection but the inner dataset element).
 
-.. class:: warningmark 
+.. class:: warningmark
 
-**Note**: Dataset index (numbering) begins with 0 (zero). 
+**Note**: Dataset index (numbering) begins with 0 (zero).
 
 .. class:: infomark
 

lib/galaxy_test/api/test_tools.py

@@ -701,23 +701,42 @@ def test_database_operation_tool_with_pending_inputs(self):
             hdca1_id = self.dataset_collection_populator.create_list_in_history(
                 history_id, contents=["a\nb\nc\nd", "e\nf\ng\nh"], wait=True
             ).json()["outputs"][0]["id"]
-            self.dataset_populator.run_tool(
+            run_response = self.dataset_populator.run_tool(
                 tool_id="cat_data_and_sleep",
                 inputs={
                     "sleep_time": 15,
                     "input1": {"batch": True, "values": [{"src": "hdca", "id": hdca1_id}]},
                 },
                 history_id=history_id,
             )
+            output_hdca_id = run_response["implicit_collections"][0]["id"]
             run_response = self.dataset_populator.run_tool(
                 tool_id="__EXTRACT_DATASET__",
                 inputs={
-                    "data_collection": {"src": "hdca", "id": hdca1_id},
+                    "data_collection": {"src": "hdca", "id": output_hdca_id},
                 },
                 history_id=history_id,
             )
             assert run_response["outputs"][0]["state"] != "ok"
 
+    @skip_without_tool("__EXTRACT_DATASET__")
+    def test_extract_dataset_invalid_element_identifier(self):
+        with self.dataset_populator.test_history(require_new=False) as history_id:
+            hdca1_id = self.dataset_collection_populator.create_list_in_history(
+                history_id, contents=["a\nb\nc\nd", "e\nf\ng\nh"], wait=True
+            ).json()["outputs"][0]["id"]
+            run_response = self.dataset_populator.run_tool_raw(
+                tool_id="__EXTRACT_DATASET__",
+                inputs={
+                    "data_collection": {"src": "hdca", "id": hdca1_id},
+                    "which": {"which_dataset": "by_index", "index": 100},
+                },
+                history_id=history_id,
+                input_format="21.01",
+            )
+            assert run_response.status_code == 400
+            assert run_response.json()["err_msg"] == "Dataset collection has no element_index with key 100."
+
     @skip_without_tool("__FILTER_FAILED_DATASETS__")
     def test_filter_failed_list(self):
         with self.dataset_populator.test_history(require_new=False) as history_id: