Merge pull request #16937 from mvdbeek/fix_merge_collection_input_state_handling

[23.1] Skip state filtering in ``__MERGE_COLLECTION__`` tool

lib/galaxy/tools/__init__.py

@@ -3383,40 +3383,27 @@ class MergeCollectionTool(DatabaseOperationTool):
         for copy, input_list in enumerate(input_lists):
             for dce in input_list.collection.elements:
                 element = dce.element_object
-                valid = False
+                element_identifier = dce.element_identifier
+                identifier_seen = element_identifier in new_element_structure
+                appearances = identifiers_map[element_identifier]
+                add_suffix = False
+                if dupl_actions == "suffix_every":
+                    add_suffix = True
+                elif dupl_actions == "suffix_conflict" and len(appearances) > 1:
+                    add_suffix = True
+                elif dupl_actions == "suffix_conflict_rest" and len(appearances) > 1 and appearances[0] != copy:
+                    add_suffix = True
 
-                # dealing with a single element
-                if hasattr(element, "is_ok"):
-                    if element.is_ok:
-                        valid = True
-                elif hasattr(element, "dataset_instances"):
-                    # we are probably a list:paired dataset, both need to be in non error state
-                    forward_o, reverse_o = element.dataset_instances
-                    if forward_o.is_ok and reverse_o.is_ok:
-                        valid = True
+                if dupl_actions == "keep_first" and identifier_seen:
+                    continue
 
-                if valid:
-                    element_identifier = dce.element_identifier
-                    identifier_seen = element_identifier in new_element_structure
-                    appearances = identifiers_map[element_identifier]
-                    add_suffix = False
-                    if dupl_actions == "suffix_every":
-                        add_suffix = True
-                    elif dupl_actions == "suffix_conflict" and len(appearances) > 1:
-                        add_suffix = True
-                    elif dupl_actions == "suffix_conflict_rest" and len(appearances) > 1 and appearances[0] != copy:
-                        add_suffix = True
+                if add_suffix and suffix_pattern:
+                    suffix = suffix_pattern.replace("#", str(copy + 1))
+                    effective_identifer = f"{element_identifier}{suffix}"
+                else:
+                    effective_identifer = element_identifier
 
-                    if dupl_actions == "keep_first" and identifier_seen:
-                        continue
-
-                    if add_suffix and suffix_pattern:
-                        suffix = suffix_pattern.replace("#", str(copy + 1))
-                        effective_identifer = f"{element_identifier}{suffix}"
-                    else:
-                        effective_identifer = element_identifier
-
-                    new_element_structure[effective_identifer] = element
+                new_element_structure[effective_identifer] = element
 
         # Don't copy until we know everything is fine and we have the structure of the list ready to go.
         new_elements = {}

lib/galaxy_test/api/test_workflows.py

@@ -2976,6 +2976,47 @@ steps:
             workflow = crate.mainEntity
             assert workflow
 
+    @skip_without_tool("__MERGE_COLLECTION__")
+    def test_merge_collection_scheduling(self, history_id):
+        summary = self._run_workflow(
+            """
+class: GalaxyWorkflow
+inputs:
+  collection:
+    type: collection
+    collection_type: list
+outputs:
+  merge_out:
+    outputSource: merge/output
+steps:
+  sleep:
+    tool_id: cat_data_and_sleep
+    in:
+      input1: collection
+    state:
+      sleep_time: 5
+  merge:
+    tool_id: __MERGE_COLLECTION__
+    in:
+      inputs_1|input: sleep/out_file1
+      inputs_0|input: sleep/out_file1
+test_data:
+  collection:
+    collection_type: list
+    elements:
+      - identifier: 1
+        content: A
+""",
+            history_id=history_id,
+            wait=True,
+            assert_ok=True,
+        )
+        invocation = self.workflow_populator.get_invocation(summary.invocation_id, step_details=True)
+        merge_out_id = invocation["output_collections"]["merge_out"]["id"]
+        merge_out = self.dataset_populator.get_history_collection_details(history_id, content_id=merge_out_id)
+        assert merge_out["element_count"] == 1
+        assert merge_out["elements"][0]["object"]["state"] == "ok"
+
     @skip_without_tool("__MERGE_COLLECTION__")
     @skip_without_tool("cat_collection")
     @skip_without_tool("head")
@@ -5385,6 +5426,39 @@ input1:
             # Also check that we don't overwrite the original HDA's datatype
             assert details2["elements"][0]["object"]["file_ext"] == "fasta"
 
+    @skip_without_tool("__EXTRACT_DATASET__")
+    def test_run_build_list_change_datatype_new_metadata_file_parameter(self):
+        # Regression test for changing datatype to a datatype with a MetadataFileParameter
+        with self.dataset_populator.test_history() as history_id:
+            self._run_workflow(
+                """
+class: GalaxyWorkflow
+inputs:
+  input1: data
+steps:
+  build_list:
+    tool_id: __BUILD_LIST__
+    in:
+      datasets_0|input: input1
+  extract_dataset:
+    tool_id: __EXTRACT_DATASET__
+    in:
+      input: build_list/output
+    outputs:
+      output:
+        change_datatype: vcf_bgzip
+""",
+                test_data="""
+input1:
+  value: test.vcf.gz
+  type: File
+  file_type: vcf_bgzip
+""",
+                history_id=history_id,
+                assert_ok=True,
+                wait=True,
+            )
+
     @skip_without_tool("__BUILD_LIST__")
     def test_run_build_list_rename_collection_output(self):
         with self.dataset_populator.test_history() as history_id: