From 15bde671e772a0ec508cf614f688d1c3e255a419 Mon Sep 17 00:00:00 2001
From: John Chilton <jmchilton@gmail.com>
Date: Fri, 15 Jan 2021 19:49:45 -0500
Subject: [PATCH] Fix certain classes of workflow extraction on copied objects.

---
 lib/galaxy/workflow/extract.py              | 37 +++++++++++++++++++--
 test/unit/workflows/test_extract_summary.py |  1 +
 2 files changed, 35 insertions(+), 3 deletions(-)

diff --git a/lib/galaxy/workflow/extract.py b/lib/galaxy/workflow/extract.py
index cf0f4a96b7b..a6c3354513d 100644
--- a/lib/galaxy/workflow/extract.py
+++ b/lib/galaxy/workflow/extract.py
@@ -129,12 +129,14 @@ def extract_steps(trans, history=None, job_ids=None, dataset_ids=None, dataset_c
             assoc_name = assoc.name
             if ToolOutputCollectionPart.is_named_collection_part_name(assoc_name):
                 continue
+            if assoc_name.startswith("__new_primary_file"):
+                continue
             if job in summary.implicit_map_jobs:
                 hid = None
                 for implicit_pair in jobs[job]:
                     query_assoc_name, dataset_collection = implicit_pair
                     if query_assoc_name == assoc_name or assoc_name.startswith("__new_primary_file_%s|" % query_assoc_name):
-                        hid = dataset_collection.hid
+                        hid = summary.hid(dataset_collection)
                 if hid is None:
                     template = "Failed to find matching implicit job - job id is %s, implicit pairs are %s, assoc_name is %s."
                     message = template % (job.id, jobs[job], assoc_name)
@@ -142,9 +144,12 @@ def extract_steps(trans, history=None, job_ids=None, dataset_ids=None, dataset_c
                     raise Exception("Failed to extract job.")
             else:
                 if hasattr(assoc, "dataset"):
-                    hid = assoc.dataset.hid
+                    has_hid = assoc.dataset
                 else:
-                    hid = assoc.dataset_collection_instance.hid
+                    has_hid = assoc.dataset_collection_instance
+                hid = summary.hid(has_hid)
+            if hid in hid_to_output_pair:
+                log.warning("duplicate hid found in extract_steps [%s]" % hid)
             hid_to_output_pair[hid] = (step, assoc.name)
     return steps
 
@@ -196,8 +201,29 @@ class WorkflowSummary:
         self.implicit_map_jobs = []
         self.collection_types = {}
 
+        self.hda_hid_in_history = {}
+        self.hdca_hid_in_history = {}
+
         self.__summarize()
 
+    def hid(self, object):
+        if object.history_content_type == "dataset_collection":
+            if object.id in self.hdca_hid_in_history:
+                return self.hdca_hid_in_history[object.id]
+            elif object.history == self.history:
+                return object.hid
+            else:
+                log.warning("extraction issue, using hdca hid from outside current history and unmapped")
+                return object.hid
+        else:
+            if object.id in self.hda_hid_in_history:
+                return self.hda_hid_in_history[object.id]
+            elif object.history == self.history:
+                return object.hid
+            else:
+                log.warning("extraction issue, using hda hid from outside current history and unmapped")
+                return object.hid
+
     def __summarize(self):
         # Make a first pass handle all singleton jobs, input dataset and dataset collections
         # just grab the implicitly mapped jobs and handle in second pass. Second pass is
@@ -215,7 +241,10 @@ class WorkflowSummary:
             self.__summarize_dataset(content)
 
     def __summarize_dataset_collection(self, dataset_collection):
+        hid_in_history = dataset_collection.hid
         dataset_collection = self.__original_hdca(dataset_collection)
+        self.hdca_hid_in_history[dataset_collection.id] = hid_in_history
+
         hid = dataset_collection.hid
         self.collection_types[hid] = dataset_collection.collection.collection_type
         cja = dataset_collection.creating_job_associations
@@ -264,7 +293,9 @@ class WorkflowSummary:
         if not self.__check_state(dataset):
             return
 
+        hid_in_history = dataset.hid
         original_hda = self.__original_hda(dataset)
+        self.hda_hid_in_history[original_hda.id] = hid_in_history
 
         if not original_hda.creating_job_associations:
             self.jobs[FakeJob(dataset)] = [(None, dataset)]
diff --git a/test/unit/workflows/test_extract_summary.py b/test/unit/workflows/test_extract_summary.py
index 4b437269421..18be155c675 100644
--- a/test/unit/workflows/test_extract_summary.py
+++ b/test/unit/workflows/test_extract_summary.py
@@ -110,6 +110,7 @@ class MockTrans:
 class MockHda:
 
     def __init__(self, state='ok', output_name='out1', job=None):
+        self.hid = 1
         self.id = 123
         self.state = state
         self.copied_from_history_dataset_association = None