Replace collection elements with re-run output

This specifically addresses the problem where some jobs of a mapped-over
collection have failed. Instead of filtering the failed collection and
restarting the workflow at this position (involving a lot of copy-paste ...)
the user can now limit the rerun to the problematic jobs and the workflow
should resume from there.
Should fix https://github.com/galaxyproject/galaxy/issues/2235.

This is one possible implementation, it would also be feasible to not
manipulate the original collection, but to copy the HDCA and then to replace
collection elements and replace all references for jobs that depend on the HDCA,
as we do for HDAs. This implementation seems simpler, but let me know if you
see problems with this approach.

lib/galaxy/jobs/__init__.py

@@ -1030,6 +1030,7 @@ class JobWrapper(object, HasResourceParameters):
                 dataset_assoc.dataset.dataset.state = dataset_assoc.dataset.dataset.states.PAUSED
                 dataset_assoc.dataset.info = message
                 self.sa_session.add(dataset_assoc.dataset)
+            log.debug("Pausing Job '%d', %s", job.id, message)
             job.set_state(job.states.PAUSED)
             self.sa_session.add(job)
 

lib/galaxy/model/__init__.py

@@ -3139,6 +3139,13 @@ class DatasetCollection(object, Dictifiable, UsesAnnotations):
         object_session(self).flush()
         return new_collection
 
+    def replace_failed_elements(self, replacements):
+        for element in self.elements:
+            if element.element_object in replacements:
+                if element.element_type == 'hda':
+                    element.hda = replacements[element.element_object]
+                # TODO: handle the case where elements are collections
+
     def set_from_dict(self, new_data):
         # Nothing currently editable in this class.
         return {}

lib/galaxy/tools/actions/__init__.py

@@ -509,6 +509,8 @@ class DefaultToolAction(object):
             # Duplicate PJAs before remap.
             for pjaa in old_job.post_job_actions:
                 current_job.add_post_job_action(pjaa.post_job_action)
+            remapped_hdas = {}
+            input_hdcas = set()
             for jtod in old_job.output_datasets:
                 for (job_to_remap, jtid) in [(jtid.job, jtid) for jtid in jtod.dataset.dependent_jobs]:
                     if (trans.user is not None and job_to_remap.user_id == trans.user.id) or (
@@ -520,6 +522,9 @@ class DefaultToolAction(object):
                                 hda.state = hda.states.NEW
                                 hda.info = None
                         input_values = dict([(p.name, json.loads(p.value)) for p in job_to_remap.parameters])
+                        remapped_hdas[jtod.dataset] = out_data[jtod.name]
+                        for jtidca in job_to_remap.input_dataset_collections:
+                            input_hdcas.add(jtidca.dataset_collection)
                         old_dataset_id = jtod.dataset_id
                         new_dataset_id = out_data[jtod.name].id
                         input_values = update_dataset_ids(input_values, {old_dataset_id: new_dataset_id}, src='hda')
@@ -530,6 +535,12 @@ class DefaultToolAction(object):
                         log.info('Job %s input HDA %s remapped to new HDA %s' % (job_to_remap.id, jtod.dataset.id, jtid.dataset.id))
                         trans.sa_session.add(job_to_remap)
                         trans.sa_session.add(jtid)
+                for hdca in input_hdcas:
+                    hdca.collection.replace_failed_elements(remapped_hdas)
+                    if hdca.implicit_collection_jobs:
+                        for job in hdca.implicit_collection_jobs.jobs:
+                            if job.job_id == old_job.id:
+                                job.job_id = current_job.id
                 jtod.dataset.visible = False
                 trans.sa_session.add(jtod)
         except Exception:

test/api/test_workflows.py

@@ -723,6 +723,54 @@ steps:
                                                                                   assert_ok=False)
             assert unpaused_dataset['state'] == 'ok'
 
+    @skip_without_tool("fail_identifier")
+    @skip_without_tool("identifier_multiple_in_conditional")
+    def test_workflow_resume_with_mapped_over_input(self):
+        with self.dataset_populator.test_history() as history_id:
+            job_summary = self._run_jobs("""
+class: GalaxyWorkflow
+steps:
+  - label: input_datasets
+    type: input_collection
+  - label: fail_identifier_1
+    tool_id: fail_identifier
+    state:
+      input1:
+        $link: input_datasets
+      failbool: true
+  - tool_id: identifier_collection
+    state:
+      input1:
+        $link: fail_identifier_1#out_file1
+test_data:
+  input_datasets:
+    type: list
+    elements:
+      - identifier: fail
+        value: 1.fastq
+        type: File
+      - identifier: success
+        value: 1.fastq
+        type: File
+""", history_id=history_id, assert_ok=False, wait=False)
+            self.wait_for_invocation_and_jobs(history_id, job_summary.workflow_id, job_summary.invocation_id, assert_ok=False)
+            history_contents = self.dataset_populator._get_contents_request(history_id=history_id).json()
+            paused_dataset = history_contents[-1]
+            failed_dataset = self.dataset_populator.get_history_dataset_details(history_id, hid=5, assert_ok=False)
+            assert paused_dataset['state'] == 'paused', paused_dataset
+            assert failed_dataset['state'] == 'error', failed_dataset
+            inputs = {"input1": {'values': [{'src': 'hda',
+                                             'id': history_contents[0]['id']}]
+                                 },
+                      "failbool": "false",
+                      "rerun_remap_job_id": failed_dataset['creating_job']}
+            self.dataset_populator.run_tool(tool_id='fail_identifier',
+                                            inputs=inputs,
+                                            history_id=history_id,
+                                            assert_ok=True)
+            unpaused_dataset = self.dataset_populator.get_history_dataset_details(history_id, wait=True, assert_ok=False)
+            assert unpaused_dataset['state'] == 'ok'
+
     @skip_without_tool("collection_creates_pair")
     def test_workflow_run_output_collection_mapping(self):
         workflow_id = self._upload_yaml_workflow("""

test/base/populators.py

@@ -275,19 +275,19 @@ class BaseDatasetPopulator(object):
         data = {}
         if filename:
             data["filename"] = filename
-        display_response = self.__get_contents_request(history_id, "/%s/display" % dataset_id, data=data)
+        display_response = self._get_contents_request(history_id, "/%s/display" % dataset_id, data=data)
         assert display_response.status_code == 200, display_response.content
         return display_response.content
 
     def get_history_dataset_details(self, history_id, **kwds):
         dataset_id = self.__history_content_id(history_id, **kwds)
-        details_response = self.__get_contents_request(history_id, "/datasets/%s" % dataset_id)
+        details_response = self._get_contents_request(history_id, "/datasets/%s" % dataset_id)
         assert details_response.status_code == 200
         return details_response.json()
 
     def get_history_collection_details(self, history_id, **kwds):
         hdca_id = self.__history_content_id(history_id, **kwds)
-        details_response = self.__get_contents_request(history_id, "/dataset_collections/%s" % hdca_id)
+        details_response = self._get_contents_request(history_id, "/dataset_collections/%s" % hdca_id)
         assert details_response.status_code == 200, details_response.content
         return details_response.json()
 
@@ -320,7 +320,7 @@ class BaseDatasetPopulator(object):
             history_content_id = kwds["dataset"]["id"]
         else:
             hid = kwds.get("hid", None)  # If not hid, just grab last dataset
-            history_contents = self.__get_contents_request(history_id).json()
+            history_contents = self._get_contents_request(history_id).json()
             if hid:
                 history_content_id = None
                 for history_item in history_contents:
@@ -333,7 +333,7 @@ class BaseDatasetPopulator(object):
                 history_content_id = history_contents[-1]["id"]
         return history_content_id
 
-    def __get_contents_request(self, history_id, suffix="", data={}):
+    def _get_contents_request(self, history_id, suffix="", data={}):
         url = "histories/%s/contents" % history_id
         if suffix:
             url = "%s%s" % (url, suffix)

test/functional/tools/fail_identifier.xml

@@ -0,0 +1,21 @@
+<tool id="fail_identifier" name="Fail input with identifier that contains fail">
+    <command><![CDATA[
+        #if $failbool and 'fail' in $input1.element_identifier
+            sh -c "exit 127"
+        #else
+            cp '$input1' '$out_file1'
+        #end if
+    ]]></command>
+    <inputs>
+        <param name="input1" type="data" label="An input file" />
+        <param name="failbool" type="boolean" label="The failure property" checked="false" />
+    </inputs>
+    <outputs>
+        <data name="out_file1" format="data"/>
+    </outputs>
+    <stdio>
+        <exit_code range="127"   level="fatal"   description="Failing exit code." />
+    </stdio>
+    <help>
+    </help>
+</tool>

test/functional/tools/samples_tool_conf.xml

@@ -96,6 +96,7 @@
   <tool file="identifier_multiple_in_repeat.xml" />
   <tool file="identifier_collection.xml" />
   <tool file="identifier_in_actions.xml" />
+  <tool file="fail_identifier.xml" />
   <tool file="tool_directory.xml" />
   <tool file="output_action_change_format.xml" />
   <tool file="collection_paired_test.xml" />