class HistoryContentsApiTestCase(ApiTestCase): def setUp(self): super().setUp() self.dataset_populator = DatasetPopulator(self.galaxy_interactor) self.dataset_collection_populator = DatasetCollectionPopulator( self.galaxy_interactor) self.library_populator = LibraryPopulator(self.galaxy_interactor) self.history_id = self.dataset_populator.new_history() def test_index_hda_summary(self): hda1 = self.dataset_populator.new_dataset(self.history_id) contents_response = self._get("histories/%s/contents" % self.history_id) hda_summary = self.__check_for_hda(contents_response, hda1) assert "display_types" not in hda_summary # Quick summary, not full details def test_make_private_and_public(self): hda1 = self._wait_for_new_hda() update_url = "histories/{}/contents/{}/permissions".format( self.history_id, hda1["id"]) role_id = self.dataset_populator.user_private_role_id() # Give manage permission to the user. payload = { "access": [], "manage": [role_id], } update_response = self._update_permissions(update_url, payload, admin=True) self._assert_status_code_is(update_response, 200) self._assert_other_user_can_access(hda1["id"]) # Then we restrict access. payload = { "action": "make_private", } update_response = self._update_permissions(update_url, payload) self._assert_status_code_is(update_response, 200) self._assert_other_user_cannot_access(hda1["id"]) # Then we restrict access. payload = { "action": "remove_restrictions", } update_response = self._update_permissions(update_url, payload) self._assert_status_code_is(update_response, 200) self._assert_other_user_can_access(hda1["id"]) def test_set_permissions_add_admin_history_contents(self): self._verify_dataset_permissions("history_contents") def test_set_permissions_add_admin_datasets(self): self._verify_dataset_permissions("dataset") def _verify_dataset_permissions(self, api_endpoint): hda1 = self._wait_for_new_hda() hda_id = hda1["id"] if api_endpoint == "history_contents": update_url = f"histories/{self.history_id}/contents/{hda_id}/permissions" else: update_url = "datasets/%s/permissions" % hda_id role_id = self.dataset_populator.user_private_role_id() payload = { "access": [role_id], "manage": [role_id], } # Other users cannot modify permissions. with self._different_user(): update_response = self._update_permissions(update_url, payload) self._assert_status_code_is(update_response, 403) # First the details render for another user. self._assert_other_user_can_access(hda_id) # Then we restrict access. update_response = self._update_permissions(update_url, payload, admin=True) self._assert_status_code_is(update_response, 200) # Finally the details don't render. self._assert_other_user_cannot_access(hda_id) # But they do for the original user. contents_response = self._get( f"histories/{self.history_id}/contents/{hda_id}").json() assert "name" in contents_response update_response = self._update_permissions(update_url, payload) self._assert_status_code_is(update_response, 200) payload = { "access": [role_id], "manage": [role_id], } update_response = self._update_permissions(update_url, payload) self._assert_status_code_is(update_response, 200) self._assert_other_user_cannot_access(hda_id) user_id = self.dataset_populator.user_id() with self._different_user(): different_user_id = self.dataset_populator.user_id() combined_user_role = self.dataset_populator.create_role( [user_id, different_user_id], description="role for testing permissions") payload = { "access": [combined_user_role["id"]], "manage": [role_id], } update_response = self._update_permissions(update_url, payload) self._assert_status_code_is(update_response, 200) # Now other user can see dataset again with access permission. self._assert_other_user_can_access(hda_id) # access doesn't imply management though... with self._different_user(): update_response = self._update_permissions(update_url, payload) self._assert_status_code_is(update_response, 403) def _assert_other_user_cannot_access(self, history_content_id): with self._different_user(): contents_response = self.dataset_populator.get_history_dataset_details_raw( history_id=self.history_id, dataset_id=history_content_id) assert contents_response.status_code == 403 def _assert_other_user_can_access(self, history_content_id): with self._different_user(): contents_response = self.dataset_populator.get_history_dataset_details_raw( history_id=self.history_id, dataset_id=history_content_id) contents_response.raise_for_status() assert "name" in contents_response.json() def test_index_hda_all_details(self): hda1 = self.dataset_populator.new_dataset(self.history_id) contents_response = self._get("histories/%s/contents?details=all" % self.history_id) hda_details = self.__check_for_hda(contents_response, hda1) self.__assert_hda_has_full_details(hda_details) def test_index_hda_detail_by_id(self): hda1 = self.dataset_populator.new_dataset(self.history_id) contents_response = self._get( "histories/{}/contents?details={}".format(self.history_id, hda1["id"])) hda_details = self.__check_for_hda(contents_response, hda1) self.__assert_hda_has_full_details(hda_details) def test_show_hda(self): hda1 = self.dataset_populator.new_dataset(self.history_id) show_response = self.__show(hda1) self._assert_status_code_is(show_response, 200) self.__assert_matches_hda(hda1, show_response.json()) def test_hda_copy(self): hda1 = self.dataset_populator.new_dataset(self.history_id) create_data = dict( source='hda', content=hda1["id"], ) second_history_id = self.dataset_populator.new_history() assert self.__count_contents(second_history_id) == 0 create_response = self._post( "histories/%s/contents" % second_history_id, create_data) self._assert_status_code_is(create_response, 200) assert self.__count_contents(second_history_id) == 1 def test_library_copy(self): ld = self.library_populator.new_library_dataset("lda_test_library") create_data = dict( source='library', content=ld["id"], ) assert self.__count_contents(self.history_id) == 0 create_response = self._post("histories/%s/contents" % self.history_id, create_data) self._assert_status_code_is(create_response, 200) assert self.__count_contents(self.history_id) == 1 def test_update(self): hda1 = self._wait_for_new_hda() assert str(hda1["deleted"]).lower() == "false" update_response = self._raw_update(hda1["id"], dict(deleted=True)) self._assert_status_code_is(update_response, 200) show_response = self.__show(hda1) assert str(show_response.json()["deleted"]).lower() == "true" update_response = self._raw_update(hda1["id"], dict(name="Updated Name")) assert self.__show(hda1).json()["name"] == "Updated Name" update_response = self._raw_update(hda1["id"], dict(name="Updated Name")) assert self.__show(hda1).json()["name"] == "Updated Name" unicode_name = 'ржевский сапоги' update_response = self._raw_update(hda1["id"], dict(name=unicode_name)) updated_hda = self.__show(hda1).json() assert updated_hda["name"] == unicode_name, updated_hda quoted_name = '"Mooo"' update_response = self._raw_update(hda1["id"], dict(name=quoted_name)) updated_hda = self.__show(hda1).json() assert updated_hda["name"] == quoted_name, quoted_name data = { "dataset_id": hda1["id"], "name": "moocow", "dbkey": "?", "annotation": None, "info": "my info is", "operation": "attributes" } update_response = self._set_edit_update(data) # No key or anything supplied, expect a permission problem. # A bit questionable but I think this is a 400 instead of a 403 so that # we don't distinguish between this is a valid ID you don't have access to # and this is an invalid ID. assert update_response.status_code == 400, update_response.content def test_update_batch(self): hda1 = self._wait_for_new_hda() assert str(hda1["deleted"]).lower() == "false" payload = dict(items=[{ "history_content_type": "dataset", "id": hda1["id"] }], deleted=True) update_response = self._raw_update_batch(payload) objects = update_response.json() assert objects[0]["deleted"] def test_update_type_failures(self): hda1 = self._wait_for_new_hda() update_response = self._raw_update(hda1["id"], dict(deleted='not valid')) self._assert_status_code_is(update_response, 400) def _wait_for_new_hda(self): hda1 = self.dataset_populator.new_dataset(self.history_id) self.dataset_populator.wait_for_history(self.history_id) return hda1 def _set_edit_update(self, json): set_edit_url = "%s/dataset/set_edit" % self.url update_response = put(set_edit_url, json=json) return update_response def _raw_update(self, item_id, data, admin=False, history_id=None): history_id = history_id or self.history_id key_param = "use_admin_key" if admin else "use_key" update_url = self._api_url( f"histories/{history_id}/contents/{item_id}", **{key_param: True}) update_response = put(update_url, json=data) return update_response def _update_permissions(self, url, data, admin=False): key_param = "use_admin_key" if admin else "use_key" update_url = self._api_url(url, **{key_param: True}) update_response = put(update_url, json=data) return update_response def _raw_update_batch(self, data): update_url = self._api_url("histories/%s/contents" % (self.history_id), use_key=True) update_response = put(update_url, json=data) return update_response def test_delete(self): hda1 = self.dataset_populator.new_dataset(self.history_id) self.dataset_populator.wait_for_history(self.history_id) assert str(self.__show(hda1).json()["deleted"]).lower() == "false" delete_response = self._delete("histories/{}/contents/{}".format( self.history_id, hda1["id"])) assert delete_response.status_code < 300 # Something in the 200s :). assert str(self.__show(hda1).json()["deleted"]).lower() == "true" def test_purge(self): hda1 = self.dataset_populator.new_dataset(self.history_id) self.dataset_populator.wait_for_history(self.history_id) assert str(self.__show(hda1).json()["deleted"]).lower() == "false" assert str(self.__show(hda1).json()["purged"]).lower() == "false" data = {'purge': True} delete_response = self._delete("histories/{}/contents/{}".format( self.history_id, hda1["id"]), data=data) assert delete_response.status_code < 300 # Something in the 200s :). assert str(self.__show(hda1).json()["deleted"]).lower() == "true" assert str(self.__show(hda1).json()["purged"]).lower() == "true" def test_dataset_collection_creation_on_contents(self): payload = self.dataset_collection_populator.create_pair_payload( self.history_id, type="dataset_collection") endpoint = "histories/%s/contents" % self.history_id self._check_pair_creation(endpoint, payload) def test_dataset_collection_creation_on_typed_contents(self): payload = self.dataset_collection_populator.create_pair_payload( self.history_id, ) endpoint = "histories/%s/contents/dataset_collections" % self.history_id self._check_pair_creation(endpoint, payload) def test_dataset_collection_create_from_exisiting_datasets_with_new_tags( self): with self.dataset_populator.test_history() as history_id: hda_id = self.dataset_populator.new_dataset(history_id, content="1 2 3")['id'] hda2_id = self.dataset_populator.new_dataset(history_id, content="1 2 3")['id'] update_response = self._raw_update(hda2_id, dict(tags=['existing:tag']), history_id=history_id).json() assert update_response['tags'] == ['existing:tag'] creation_payload = { 'collection_type': 'list', 'history_id': history_id, 'element_identifiers': json.dumps([{ 'id': hda_id, 'src': 'hda', 'name': 'element_id1', 'tags': ['my_new_tag'] }, { 'id': hda2_id, 'src': 'hda', 'name': 'element_id2', 'tags': ['another_new_tag'] }]), 'type': 'dataset_collection', 'copy_elements': True } r = self._post("histories/%s/contents" % self.history_id, creation_payload).json() assert r['elements'][0]['object'][ 'id'] != hda_id, "HDA has not been copied" assert len(r['elements'][0]['object']['tags']) == 1 assert r['elements'][0]['object']['tags'][0] == 'my_new_tag' assert len(r['elements'][1]['object'] ['tags']) == 2, r['elements'][1]['object']['tags'] original_hda = self.dataset_populator.get_history_dataset_details( history_id=history_id, dataset_id=hda_id) assert len(original_hda['tags']) == 0, original_hda['tags'] def _check_pair_creation(self, endpoint, payload): pre_collection_count = self.__count_contents(type="dataset_collection") pre_dataset_count = self.__count_contents(type="dataset") pre_combined_count = self.__count_contents( type="dataset,dataset_collection") dataset_collection_response = self._post(endpoint, payload) dataset_collection = self.__check_create_collection_response( dataset_collection_response) post_collection_count = self.__count_contents( type="dataset_collection") post_dataset_count = self.__count_contents(type="dataset") post_combined_count = self.__count_contents( type="dataset,dataset_collection") # Test filtering types with index. assert pre_collection_count == 0 assert post_collection_count == 1 assert post_combined_count == pre_dataset_count + 1 assert post_combined_count == pre_combined_count + 1 assert pre_dataset_count == post_dataset_count # Test show dataset colleciton. collection_url = "histories/{}/contents/dataset_collections/{}".format( self.history_id, dataset_collection["id"]) show_response = self._get(collection_url) self._assert_status_code_is(show_response, 200) dataset_collection = show_response.json() self._assert_has_keys(dataset_collection, "url", "name", "deleted") assert not dataset_collection["deleted"] delete_response = delete(self._api_url(collection_url, use_key=True)) self._assert_status_code_is(delete_response, 200) show_response = self._get(collection_url) dataset_collection = show_response.json() assert dataset_collection["deleted"] @skip_without_tool("collection_creates_list") def test_jobs_summary_simple_hdca(self): create_response = self.dataset_collection_populator.create_list_in_history( self.history_id, contents=["a\nb\nc\nd", "e\nf\ng\nh"]) hdca_id = create_response.json()["id"] run = self.dataset_populator.run_collection_creates_list( self.history_id, hdca_id) collections = run['output_collections'] collection = collections[0] jobs_summary_url = "histories/{}/contents/dataset_collections/{}/jobs_summary".format( self.history_id, collection["id"]) jobs_summary_response = self._get(jobs_summary_url) self._assert_status_code_is(jobs_summary_response, 200) jobs_summary = jobs_summary_response.json() self._assert_has_keys(jobs_summary, "populated_state", "states") @skip_without_tool("cat1") def test_jobs_summary_implicit_hdca(self): create_response = self.dataset_collection_populator.create_pair_in_history( self.history_id, contents=["123", "456"]) hdca_id = create_response.json()["id"] inputs = { "input1": { 'batch': True, 'values': [{ 'src': 'hdca', 'id': hdca_id }] }, } run = self.dataset_populator.run_tool("cat1", inputs=inputs, history_id=self.history_id) self.dataset_populator.wait_for_history_jobs(self.history_id) collections = run['implicit_collections'] collection = collections[0] jobs_summary_url = "histories/{}/contents/dataset_collections/{}/jobs_summary".format( self.history_id, collection["id"]) jobs_summary_response = self._get(jobs_summary_url) self._assert_status_code_is(jobs_summary_response, 200) jobs_summary = jobs_summary_response.json() self._assert_has_keys(jobs_summary, "populated_state", "states") states = jobs_summary["states"] assert states.get("ok") == 2, states def test_dataset_collection_hide_originals(self): payload = self.dataset_collection_populator.create_pair_payload( self.history_id, type="dataset_collection") payload["hide_source_items"] = True dataset_collection_response = self._post( "histories/%s/contents" % self.history_id, payload) self.__check_create_collection_response(dataset_collection_response) contents_response = self._get("histories/%s/contents" % self.history_id) datasets = [ d for d in contents_response.json() if d["history_content_type"] == "dataset" and d["hid"] in [1, 2] ] # Assert two datasets in source were hidden. assert len(datasets) == 2 assert not datasets[0]["visible"] assert not datasets[1]["visible"] def test_update_dataset_collection(self): payload = self.dataset_collection_populator.create_pair_payload( self.history_id, type="dataset_collection") dataset_collection_response = self._post( "histories/%s/contents" % self.history_id, payload) self._assert_status_code_is(dataset_collection_response, 200) hdca = dataset_collection_response.json() update_url = self._api_url( "histories/{}/contents/dataset_collections/{}".format( self.history_id, hdca["id"]), use_key=True) # Awkward json.dumps required here because of https://trello.com/c/CQwmCeG6 body = json.dumps(dict(name="newnameforpair")) update_response = put(update_url, data=body) self._assert_status_code_is(update_response, 200) show_response = self.__show(hdca) assert str(show_response.json()["name"]) == "newnameforpair" def test_hdca_copy(self): hdca = self.dataset_collection_populator.create_pair_in_history( self.history_id).json() hdca_id = hdca["id"] second_history_id = self.dataset_populator.new_history() create_data = dict( source='hdca', content=hdca_id, ) assert len( self._get("histories/%s/contents/dataset_collections" % second_history_id).json()) == 0 create_response = self._post( "histories/%s/contents/dataset_collections" % second_history_id, create_data) self.__check_create_collection_response(create_response) contents = self._get("histories/%s/contents/dataset_collections" % second_history_id).json() assert len(contents) == 1 new_forward, _ = self.__get_paired_response_elements(contents[0]) self._assert_has_keys(new_forward, "history_id") assert new_forward["history_id"] == self.history_id def test_hdca_copy_with_new_dbkey(self): hdca = self.dataset_collection_populator.create_pair_in_history( self.history_id).json() hdca_id = hdca["id"] assert hdca["elements"][0]["object"]["metadata_dbkey"] == "?" assert hdca["elements"][0]["object"]["genome_build"] == "?" create_data = {'source': 'hdca', 'content': hdca_id, 'dbkey': 'hg19'} create_response = self._post( f"histories/{self.history_id}/contents/dataset_collections", create_data) collection = self.__check_create_collection_response(create_response) new_forward = collection['elements'][0]['object'] assert new_forward["metadata_dbkey"] == "hg19" assert new_forward["genome_build"] == "hg19" def test_hdca_copy_and_elements(self): hdca = self.dataset_collection_populator.create_pair_in_history( self.history_id).json() hdca_id = hdca["id"] second_history_id = self.dataset_populator.new_history() create_data = dict( source='hdca', content=hdca_id, copy_elements=True, ) assert len( self._get("histories/%s/contents/dataset_collections" % second_history_id).json()) == 0 create_response = self._post( "histories/%s/contents/dataset_collections" % second_history_id, create_data) self.__check_create_collection_response(create_response) contents = self._get("histories/%s/contents/dataset_collections" % second_history_id).json() assert len(contents) == 1 new_forward, _ = self.__get_paired_response_elements(contents[0]) self._assert_has_keys(new_forward, "history_id") assert new_forward["history_id"] == second_history_id def __get_paired_response_elements(self, contents): hdca = self.__show(contents).json() self._assert_has_keys(hdca, "name", "deleted", "visible", "elements") elements = hdca["elements"] assert len(elements) == 2 element0 = elements[0] element1 = elements[1] self._assert_has_keys(element0, "object") self._assert_has_keys(element1, "object") return element0["object"], element1["object"] def test_hdca_from_library_datasets(self): ld = self.library_populator.new_library_dataset("el1") ldda_id = ld["ldda_id"] element_identifiers = [{"name": "el1", "src": "ldda", "id": ldda_id}] history_id = self.dataset_populator.new_history() create_data = dict( history_id=history_id, type="dataset_collection", name="Test From Library", element_identifiers=json.dumps(element_identifiers), collection_type="list", ) create_response = self._post( "histories/%s/contents/dataset_collections" % history_id, create_data) hdca = self.__check_create_collection_response(create_response) elements = hdca["elements"] assert len(elements) == 1 hda = elements[0]["object"] assert hda["hda_ldda"] == "hda" assert hda["history_content_type"] == "dataset" assert hda["copied_from_ldda_id"] == ldda_id assert hda['history_id'] == history_id def test_hdca_from_inaccessible_library_datasets(self): library, library_dataset = self.library_populator.new_library_dataset_in_private_library( "HDCACreateInaccesibleLibrary") ldda_id = library_dataset["id"] element_identifiers = [{"name": "el1", "src": "ldda", "id": ldda_id}] create_data = dict( history_id=self.history_id, type="dataset_collection", name="Test From Library", element_identifiers=json.dumps(element_identifiers), collection_type="list", ) with self._different_user(): second_history_id = self.dataset_populator.new_history() create_response = self._post( "histories/%s/contents/dataset_collections" % second_history_id, create_data) self._assert_status_code_is(create_response, 403) def __check_create_collection_response(self, response): self._assert_status_code_is(response, 200) dataset_collection = response.json() self._assert_has_keys(dataset_collection, "url", "name", "deleted", "visible", "elements") return dataset_collection def __show(self, contents): show_response = self._get("histories/{}/contents/{}s/{}".format( self.history_id, contents["history_content_type"], contents["id"])) return show_response def __count_contents(self, history_id=None, **kwds): if history_id is None: history_id = self.history_id contents_response = self._get("histories/%s/contents" % history_id, kwds) return len(contents_response.json()) def __assert_hda_has_full_details(self, hda_details): self._assert_has_keys(hda_details, "display_types", "display_apps") def __check_for_hda(self, contents_response, hda): self._assert_status_code_is(contents_response, 200) contents = contents_response.json() assert len(contents) == 1 hda_summary = contents[0] self.__assert_matches_hda(hda, hda_summary) return hda_summary def __assert_matches_hda(self, input_hda, query_hda): self._assert_has_keys(query_hda, "id", "name") assert input_hda["name"] == query_hda["name"] assert input_hda["id"] == query_hda["id"] def test_job_state_summary_field(self): create_response = self.dataset_collection_populator.create_pair_in_history( self.history_id, contents=["123", "456"]) self._assert_status_code_is(create_response, 200) contents_response = self._get( "histories/%s/contents?v=dev&keys=job_state_summary&view=summary" % self.history_id) self._assert_status_code_is(contents_response, 200) contents = contents_response.json() for c in filter( lambda c: c['history_content_type'] == 'dataset_collection', contents): assert isinstance(c, dict) assert 'job_state_summary' in c assert isinstance(c['job_state_summary'], dict) def _get_content(self, history_id, update_time): return self._get( f"/api/histories/{history_id}/contents/near/100/100?update_time-ge={update_time}" ).json() def test_history_contents_near_with_update_time(self): with self.dataset_populator.test_history() as history_id: first_time = datetime.utcnow().isoformat() assert len(self._get_content(history_id, update_time=first_time)) == 0 self.dataset_collection_populator.create_list_in_history( history_id=history_id) assert len(self._get_content( history_id, update_time=first_time)) == 4 # 3 datasets self.dataset_populator.wait_for_history(history_id) all_datasets_finished = first_time = datetime.utcnow().isoformat() assert len( self._get_content(history_id, update_time=all_datasets_finished)) == 0 @skip_without_tool('cat_data_and_sleep') def test_history_contents_near_with_update_time_implicit_collection(self): with self.dataset_populator.test_history() as history_id: hdca_id = self.dataset_collection_populator.create_list_in_history( history_id=history_id).json()['id'] self.dataset_populator.wait_for_history(history_id) inputs = { "input1": { 'batch': True, 'values': [{ "src": "hdca", "id": hdca_id }] }, "sleep_time": 2, } response = self.dataset_populator.run_tool( "cat_data_and_sleep", inputs, history_id, assert_ok=False, ).json() update_time = datetime.utcnow().isoformat() collection_id = response['implicit_collections'][0]['id'] for _ in range(20): time.sleep(1) update = self._get_content(history_id, update_time=update_time) if any(c for c in update if c['history_content_type'] == 'dataset_collection' and c['job_state_summary']['ok'] == 3): return raise Exception( f"History content update time query did not include final update for implicit collection {collection_id}" ) @skip_without_tool('collection_creates_dynamic_nested') def test_history_contents_near_with_update_time_explicit_collection(self): with self.dataset_populator.test_history() as history_id: inputs = {'foo': 'bar', 'sleep_time': 2} response = self.dataset_populator.run_tool( "collection_creates_dynamic_nested", inputs, history_id, assert_ok=False, ).json() update_time = datetime.utcnow().isoformat() collection_id = response['output_collections'][0]['id'] for _ in range(20): time.sleep(1) update = self._get_content(history_id, update_time=update_time) if any(c for c in update if c['history_content_type'] == 'dataset_collection' and c['populated_state'] == 'ok'): return raise Exception( f"History content update time query did not include populated_state update for dynamic nested collection {collection_id}" )
class DatasetCollectionApiTestCase(ApiTestCase): def setUp(self): super().setUp() self.dataset_populator = DatasetPopulator(self.galaxy_interactor) self.dataset_collection_populator = DatasetCollectionPopulator(self.galaxy_interactor) self.history_id = self.dataset_populator.new_history() def test_create_pair_from_history(self): payload = self.dataset_collection_populator.create_pair_payload( self.history_id, instance_type="history", ) create_response = self._post("dataset_collections", payload) dataset_collection = self._check_create_response(create_response) returned_datasets = dataset_collection["elements"] assert len(returned_datasets) == 2, dataset_collection def test_create_list_from_history(self): element_identifiers = self.dataset_collection_populator.list_identifiers(self.history_id) payload = dict( instance_type="history", history_id=self.history_id, element_identifiers=json.dumps(element_identifiers), collection_type="list", ) create_response = self._post("dataset_collections", payload) dataset_collection = self._check_create_response(create_response) returned_datasets = dataset_collection["elements"] assert len(returned_datasets) == 3, dataset_collection def test_create_list_of_existing_pairs(self): pair_payload = self.dataset_collection_populator.create_pair_payload( self.history_id, instance_type="history", ) pair_create_response = self._post("dataset_collections", pair_payload) dataset_collection = self._check_create_response(pair_create_response) hdca_id = dataset_collection["id"] element_identifiers = [ dict(name="test1", src="hdca", id=hdca_id) ] payload = dict( instance_type="history", history_id=self.history_id, element_identifiers=json.dumps(element_identifiers), collection_type="list", ) create_response = self._post("dataset_collections", payload) dataset_collection = self._check_create_response(create_response) returned_collections = dataset_collection["elements"] assert len(returned_collections) == 1, dataset_collection def test_create_list_of_new_pairs(self): identifiers = self.dataset_collection_populator.nested_collection_identifiers(self.history_id, "list:paired") payload = dict( collection_type="list:paired", instance_type="history", history_id=self.history_id, name="a nested collection", element_identifiers=json.dumps(identifiers), ) create_response = self._post("dataset_collections", payload) dataset_collection = self._check_create_response(create_response) assert dataset_collection["collection_type"] == "list:paired" assert dataset_collection["name"] == "a nested collection" returned_collections = dataset_collection["elements"] assert len(returned_collections) == 1, dataset_collection pair_1_element = returned_collections[0] self._assert_has_keys(pair_1_element, "element_identifier", "element_index", "object") assert pair_1_element["element_identifier"] == "test_level_1", pair_1_element assert pair_1_element["element_index"] == 0, pair_1_element pair_1_object = pair_1_element["object"] self._assert_has_keys(pair_1_object, "collection_type", "elements", "element_count") self.assertEqual(pair_1_object["collection_type"], "paired") self.assertEqual(pair_1_object["populated"], True) pair_elements = pair_1_object["elements"] assert len(pair_elements) == 2 pair_1_element_1 = pair_elements[0] assert pair_1_element_1["element_index"] == 0 def test_list_download(self): fetch_response = self.dataset_collection_populator.create_list_in_history(self.history_id, direct_upload=True).json() dataset_collection = self.dataset_collection_populator.wait_for_fetched_collection(fetch_response) returned_dce = dataset_collection["elements"] assert len(returned_dce) == 3, dataset_collection create_response = self._download_dataset_collection(history_id=self.history_id, hdca_id=dataset_collection['id']) self._assert_status_code_is(create_response, 200) archive = zipfile.ZipFile(BytesIO(create_response.content)) namelist = archive.namelist() assert len(namelist) == 3, "Expected 3 elements in [%s]" % namelist collection_name = dataset_collection['name'] for element, zip_path in zip(returned_dce, namelist): assert f"{collection_name}/{element['element_identifier']}.{element['object']['file_ext']}" == zip_path def test_pair_download(self): fetch_response = self.dataset_collection_populator.create_pair_in_history(self.history_id, direct_upload=True).json() dataset_collection = self.dataset_collection_populator.wait_for_fetched_collection(fetch_response) returned_dce = dataset_collection["elements"] assert len(returned_dce) == 2, dataset_collection hdca_id = dataset_collection['id'] create_response = self._download_dataset_collection(history_id=self.history_id, hdca_id=hdca_id) self._assert_status_code_is(create_response, 200) archive = zipfile.ZipFile(BytesIO(create_response.content)) namelist = archive.namelist() assert len(namelist) == 2, "Expected 2 elements in [%s]" % namelist collection_name = dataset_collection['name'] for element, zip_path in zip(returned_dce, namelist): assert "{}/{}.{}".format(collection_name, element['element_identifier'], element['object']['file_ext']) == zip_path def test_list_pair_download(self): fetch_response = self.dataset_collection_populator.create_list_of_pairs_in_history(self.history_id).json() dataset_collection = self.dataset_collection_populator.wait_for_fetched_collection(fetch_response) returned_dce = dataset_collection["elements"] assert len(returned_dce) == 1, dataset_collection list_collection_name = dataset_collection['name'] pair = returned_dce[0] create_response = self._download_dataset_collection(history_id=self.history_id, hdca_id=dataset_collection['id']) self._assert_status_code_is(create_response, 200) archive = zipfile.ZipFile(BytesIO(create_response.content)) namelist = archive.namelist() assert len(namelist) == 2, "Expected 2 elements in [%s]" % namelist pair_collection_name = pair['element_identifier'] for element, zip_path in zip(pair['object']['elements'], namelist): assert "{}/{}/{}.{}".format(list_collection_name, pair_collection_name, element['element_identifier'], element['object']['file_ext']) == zip_path def test_list_list_download(self): dataset_collection = self.dataset_collection_populator.create_list_of_list_in_history(self.history_id).json() self.dataset_collection_populator.wait_for_dataset_collection(dataset_collection, assert_ok=True) returned_dce = dataset_collection["elements"] assert len(returned_dce) == 1, dataset_collection create_response = self._download_dataset_collection(history_id=self.history_id, hdca_id=dataset_collection['id']) self._assert_status_code_is(create_response, 200) archive = zipfile.ZipFile(BytesIO(create_response.content)) namelist = archive.namelist() assert len(namelist) == 3, "Expected 3 elements in [%s]" % namelist def test_list_list_list_download(self): dataset_collection = self.dataset_collection_populator.create_list_of_list_in_history(self.history_id, collection_type='list:list:list').json() self.dataset_collection_populator.wait_for_dataset_collection(dataset_collection, assert_ok=True) returned_dce = dataset_collection["elements"] assert len(returned_dce) == 1, dataset_collection create_response = self._download_dataset_collection(history_id=self.history_id, hdca_id=dataset_collection['id']) self._assert_status_code_is(create_response, 200) archive = zipfile.ZipFile(BytesIO(create_response.content)) namelist = archive.namelist() assert len(namelist) == 3, "Expected 3 elements in [%s]" % namelist def test_hda_security(self): element_identifiers = self.dataset_collection_populator.pair_identifiers(self.history_id) self.dataset_populator.make_private(self.history_id, element_identifiers[0]["id"]) with self._different_user(): history_id = self.dataset_populator.new_history() payload = dict( instance_type="history", history_id=history_id, element_identifiers=json.dumps(element_identifiers), collection_type="paired", ) create_response = self._post("dataset_collections", payload) self._assert_status_code_is(create_response, 403) def test_enforces_unique_names(self): element_identifiers = self.dataset_collection_populator.list_identifiers(self.history_id) element_identifiers[2]["name"] = element_identifiers[0]["name"] payload = dict( instance_type="history", history_id=self.history_id, element_identifiers=json.dumps(element_identifiers), collection_type="list", ) create_response = self._post("dataset_collections", payload) self._assert_status_code_is(create_response, 400) def test_upload_collection(self): elements = [{"src": "files", "dbkey": "hg19", "info": "my cool bed", "tags": ["name:data1", "group:condition:treated", "machine:illumina"]}] targets = [{ "destination": {"type": "hdca"}, "elements": elements, "collection_type": "list", "name": "Test upload", "tags": ["name:collection1"] }] payload = { "history_id": self.history_id, "targets": json.dumps(targets), "__files": {"files_0|file_data": open(self.test_data_resolver.get_filename("4.bed"))}, } self.dataset_populator.fetch(payload) hdca = self._assert_one_collection_created_in_history() self.assertEqual(hdca["name"], "Test upload") hdca_tags = hdca["tags"] assert len(hdca_tags) == 1 assert "name:collection1" in hdca_tags assert len(hdca["elements"]) == 1, hdca element0 = hdca["elements"][0] assert element0["element_identifier"] == "4.bed" dataset0 = element0["object"] assert dataset0["file_size"] == 61 dataset_tags = dataset0["tags"] assert len(dataset_tags) == 3, dataset0 def test_upload_nested(self): elements = [{"name": "samp1", "elements": [{"src": "files", "dbkey": "hg19", "info": "my cool bed"}]}] targets = [{ "destination": {"type": "hdca"}, "elements": elements, "collection_type": "list:list", "name": "Test upload", }] payload = { "history_id": self.history_id, "targets": json.dumps(targets), "__files": {"files_0|file_data": open(self.test_data_resolver.get_filename("4.bed"))}, } self.dataset_populator.fetch(payload) hdca = self._assert_one_collection_created_in_history() self.assertEqual(hdca["name"], "Test upload") assert len(hdca["elements"]) == 1, hdca element0 = hdca["elements"][0] assert element0["element_identifier"] == "samp1" @skip_if_github_down def test_upload_collection_from_url(self): elements = [{"src": "url", "url": "https://raw.githubusercontent.com/galaxyproject/galaxy/dev/test-data/4.bed", "info": "my cool bed"}] targets = [{ "destination": {"type": "hdca"}, "elements": elements, "collection_type": "list", }] payload = { "history_id": self.history_id, "targets": json.dumps(targets), "__files": {"files_0|file_data": open(self.test_data_resolver.get_filename("4.bed"))}, } self.dataset_populator.fetch(payload) hdca = self._assert_one_collection_created_in_history() assert len(hdca["elements"]) == 1, hdca element0 = hdca["elements"][0] assert element0["element_identifier"] == "4.bed" assert element0["object"]["file_size"] == 61 @skip_if_github_down def test_upload_collection_failed_expansion_url(self): targets = [{ "destination": {"type": "hdca"}, "elements_from": "bagit", "collection_type": "list", "src": "url", "url": "https://raw.githubusercontent.com/galaxyproject/galaxy/dev/test-data/4.bed", }] payload = { "history_id": self.history_id, "targets": json.dumps(targets), "__files": {"files_0|file_data": open(self.test_data_resolver.get_filename("4.bed"))}, } self.dataset_populator.fetch(payload, assert_ok=False, wait=True) hdca = self._assert_one_collection_created_in_history() assert hdca["populated"] is False assert "bagit.txt" in hdca["populated_state_message"], hdca def _assert_one_collection_created_in_history(self): contents_response = self._get("histories/%s/contents/dataset_collections" % self.history_id) self._assert_status_code_is(contents_response, 200) contents = contents_response.json() assert len(contents) == 1 hdca = contents[0] assert hdca["history_content_type"] == "dataset_collection" hdca_id = hdca["id"] collection_response = self._get(f"histories/{self.history_id}/contents/dataset_collections/{hdca_id}") self._assert_status_code_is(collection_response, 200) return collection_response.json() def _check_create_response(self, create_response): self._assert_status_code_is(create_response, 200) dataset_collection = create_response.json() self._assert_has_keys(dataset_collection, "elements", "url", "name", "collection_type", "element_count") return dataset_collection def _download_dataset_collection(self, history_id, hdca_id): return self._get(f"histories/{history_id}/contents/dataset_collections/{hdca_id}/download") def test_collection_contents_security(self): # request contents on an hdca that doesn't belong to user hdca, contents_url = self._create_collection_contents_pair() with self._different_user(): contents_response = self._get(contents_url) self._assert_status_code_is(contents_response, 403) def test_collection_contents_invalid_collection(self): # request an invalid collection from a valid hdca, should get 404 hdca, contents_url = self._create_collection_contents_pair() response = self._get(contents_url) self._assert_status_code_is(response, 200) fake_collection_id = '5d7db0757a2eb7ef' fake_contents_url = '/api/dataset_collections/{}/contents/{}'.format(hdca['id'], fake_collection_id) error_response = self._get(fake_contents_url) assert_object_id_error(error_response) def test_show_dataset_collection_contents(self): # Get contents_url from history contents, use it to show the first level # of collection contents in the created HDCA, then use it again to drill # down into the nested collection contents hdca = self.dataset_collection_populator.create_list_of_list_in_history(self.history_id).json() root_contents_url = self._get_contents_url_for_hdca(hdca) # check root contents for this collection root_contents = self._get(root_contents_url).json() assert len(root_contents) == len(hdca['elements']) self._compare_collection_contents_elements(root_contents, hdca['elements']) # drill down, retrieve nested collection contents assert 'object' in root_contents[0] assert 'contents_url' in root_contents[0]['object'] drill_contents_url = root_contents[0]['object']['contents_url'] drill_contents = self._get(drill_contents_url).json() assert len(drill_contents) == len(hdca['elements'][0]['object']['elements']) self._compare_collection_contents_elements(drill_contents, hdca['elements'][0]['object']['elements']) def test_collection_contents_limit_offset(self): # check limit/offset params for collection contents endpoint hdca, root_contents_url = self._create_collection_contents_pair() # check limit limited_contents = self._get(root_contents_url + '?limit=1').json() assert len(limited_contents) == 1 assert limited_contents[0]['element_index'] == 0 # check offset offset_contents = self._get(root_contents_url + '?offset=1').json() assert len(offset_contents) == 1 assert offset_contents[0]['element_index'] == 1 def _compare_collection_contents_elements(self, contents_elements, hdca_elements): # compare collection api results to existing hdca element contents fields = ['element_identifier', 'element_index', 'element_type', 'id', 'model_class'] for (content_element, hdca_element) in zip(contents_elements, hdca_elements): for f in fields: assert content_element[f] == hdca_element[f] def _create_collection_contents_pair(self): # Create a simple collection, return hdca and contents_url payload = self.dataset_collection_populator.create_pair_payload(self.history_id, instance_type="history") create_response = self._post("dataset_collections", payload) hdca = self._check_create_response(create_response) root_contents_url = self._get_contents_url_for_hdca(hdca) return hdca, root_contents_url def _get_contents_url_for_hdca(self, hdca): # look up the history contents using optional serialization key history_contents_url = "histories/%s/contents?v=dev&view=summary&keys=contents_url" % (self.history_id) json = self._get(history_contents_url).json() # filter out the collection we just made id = hdca.id # make sure the contents_url appears def find_hdca(c): return c['history_content_type'] == 'dataset_collection' and c['id'] == hdca['id'] matches = list(filter(find_hdca, json)) assert len(matches) == 1 assert 'contents_url' in matches[0] return matches[0]['contents_url']
class DatasetCollectionApiTestCase(ApiTestCase): history_id: str def setUp(self): super().setUp() self.dataset_populator = DatasetPopulator(self.galaxy_interactor) self.dataset_collection_populator = DatasetCollectionPopulator( self.galaxy_interactor) self.history_id = self.dataset_populator.new_history() def test_create_pair_from_history(self): payload = self.dataset_collection_populator.create_pair_payload( self.history_id, instance_type="history", ) create_response = self._post("dataset_collections", payload, json=True) dataset_collection = self._check_create_response(create_response) returned_datasets = dataset_collection["elements"] assert len(returned_datasets) == 2, dataset_collection def test_create_list_from_history(self): element_identifiers = self.dataset_collection_populator.list_identifiers( self.history_id) payload = dict( instance_type="history", history_id=self.history_id, element_identifiers=element_identifiers, collection_type="list", ) create_response = self._post("dataset_collections", payload, json=True) dataset_collection = self._check_create_response(create_response) returned_datasets = dataset_collection["elements"] assert len(returned_datasets) == 3, dataset_collection def test_create_list_of_existing_pairs(self): pair_payload = self.dataset_collection_populator.create_pair_payload( self.history_id, instance_type="history", ) pair_create_response = self._post("dataset_collections", pair_payload, json=True) dataset_collection = self._check_create_response(pair_create_response) hdca_id = dataset_collection["id"] element_identifiers = [dict(name="test1", src="hdca", id=hdca_id)] payload = dict( instance_type="history", history_id=self.history_id, element_identifiers=element_identifiers, collection_type="list", ) create_response = self._post("dataset_collections", payload, json=True) dataset_collection = self._check_create_response(create_response) returned_collections = dataset_collection["elements"] assert len(returned_collections) == 1, dataset_collection def test_create_list_of_new_pairs(self): identifiers = self.dataset_collection_populator.nested_collection_identifiers( self.history_id, "list:paired") payload = dict( collection_type="list:paired", instance_type="history", history_id=self.history_id, name="a nested collection", element_identifiers=identifiers, ) create_response = self._post("dataset_collections", payload, json=True) dataset_collection = self._check_create_response(create_response) assert dataset_collection["collection_type"] == "list:paired" assert dataset_collection["name"] == "a nested collection" returned_collections = dataset_collection["elements"] assert len(returned_collections) == 1, dataset_collection pair_1_element = returned_collections[0] self._assert_has_keys(pair_1_element, "element_identifier", "element_index", "object") assert pair_1_element[ "element_identifier"] == "test_level_1", pair_1_element assert pair_1_element["element_index"] == 0, pair_1_element pair_1_object = pair_1_element["object"] self._assert_has_keys(pair_1_object, "collection_type", "elements", "element_count") self.assertEqual(pair_1_object["collection_type"], "paired") self.assertEqual(pair_1_object["populated"], True) pair_elements = pair_1_object["elements"] assert len(pair_elements) == 2 pair_1_element_1 = pair_elements[0] assert pair_1_element_1["element_index"] == 0 def test_list_download(self): fetch_response = self.dataset_collection_populator.create_list_in_history( self.history_id, direct_upload=True).json() dataset_collection = self.dataset_collection_populator.wait_for_fetched_collection( fetch_response) returned_dce = dataset_collection["elements"] assert len(returned_dce) == 3, dataset_collection create_response = self._download_dataset_collection( history_id=self.history_id, hdca_id=dataset_collection['id']) self._assert_status_code_is(create_response, 200) archive = zipfile.ZipFile(BytesIO(create_response.content)) namelist = archive.namelist() assert len(namelist) == 3, f"Expected 3 elements in [{namelist}]" collection_name = dataset_collection['name'] for element, zip_path in zip(returned_dce, namelist): assert f"{collection_name}/{element['element_identifier']}.{element['object']['file_ext']}" == zip_path def test_pair_download(self): fetch_response = self.dataset_collection_populator.create_pair_in_history( self.history_id, direct_upload=True).json() dataset_collection = self.dataset_collection_populator.wait_for_fetched_collection( fetch_response) returned_dce = dataset_collection["elements"] assert len(returned_dce) == 2, dataset_collection hdca_id = dataset_collection['id'] create_response = self._download_dataset_collection( history_id=self.history_id, hdca_id=hdca_id) self._assert_status_code_is(create_response, 200) archive = zipfile.ZipFile(BytesIO(create_response.content)) namelist = archive.namelist() assert len(namelist) == 2, f"Expected 2 elements in [{namelist}]" collection_name = dataset_collection['name'] for element, zip_path in zip(returned_dce, namelist): assert f"{collection_name}/{element['element_identifier']}.{element['object']['file_ext']}" == zip_path def test_list_pair_download(self): fetch_response = self.dataset_collection_populator.create_list_of_pairs_in_history( self.history_id).json() dataset_collection = self.dataset_collection_populator.wait_for_fetched_collection( fetch_response) returned_dce = dataset_collection["elements"] assert len(returned_dce) == 1, dataset_collection list_collection_name = dataset_collection['name'] pair = returned_dce[0] create_response = self._download_dataset_collection( history_id=self.history_id, hdca_id=dataset_collection['id']) self._assert_status_code_is(create_response, 200) archive = zipfile.ZipFile(BytesIO(create_response.content)) namelist = archive.namelist() assert len(namelist) == 2, f"Expected 2 elements in [{namelist}]" pair_collection_name = pair['element_identifier'] for element, zip_path in zip(pair['object']['elements'], namelist): assert f"{list_collection_name}/{pair_collection_name}/{element['element_identifier']}.{element['object']['file_ext']}" == zip_path def test_list_list_download(self): dataset_collection = self.dataset_collection_populator.create_list_of_list_in_history( self.history_id).json() self.dataset_collection_populator.wait_for_dataset_collection( dataset_collection, assert_ok=True) returned_dce = dataset_collection["elements"] assert len(returned_dce) == 1, dataset_collection create_response = self._download_dataset_collection( history_id=self.history_id, hdca_id=dataset_collection['id']) self._assert_status_code_is(create_response, 200) archive = zipfile.ZipFile(BytesIO(create_response.content)) namelist = archive.namelist() assert len(namelist) == 3, f"Expected 3 elements in [{namelist}]" def test_list_list_list_download(self): dataset_collection = self.dataset_collection_populator.create_list_of_list_in_history( self.history_id, collection_type='list:list:list').json() self.dataset_collection_populator.wait_for_dataset_collection( dataset_collection, assert_ok=True) returned_dce = dataset_collection["elements"] assert len(returned_dce) == 1, dataset_collection create_response = self._download_dataset_collection( history_id=self.history_id, hdca_id=dataset_collection['id']) self._assert_status_code_is(create_response, 200) archive = zipfile.ZipFile(BytesIO(create_response.content)) namelist = archive.namelist() assert len(namelist) == 3, f"Expected 3 elements in [{namelist}]" def test_hda_security(self): element_identifiers = self.dataset_collection_populator.pair_identifiers( self.history_id) self.dataset_populator.make_private(self.history_id, element_identifiers[0]["id"]) with self._different_user(): history_id = self.dataset_populator.new_history() payload = dict( instance_type="history", history_id=history_id, element_identifiers=element_identifiers, collection_type="paired", ) create_response = self._post("dataset_collections", payload, json=True) self._assert_status_code_is(create_response, 403) def test_enforces_unique_names(self): element_identifiers = self.dataset_collection_populator.list_identifiers( self.history_id) element_identifiers[2]["name"] = element_identifiers[0]["name"] payload = dict( instance_type="history", history_id=self.history_id, element_identifiers=element_identifiers, collection_type="list", ) create_response = self._post("dataset_collections", payload, json=True) self._assert_status_code_is(create_response, 400) def test_upload_collection(self): elements = [{ "src": "files", "dbkey": "hg19", "info": "my cool bed", "tags": ["name:data1", "group:condition:treated", "machine:illumina"] }] targets = [{ "destination": { "type": "hdca" }, "elements": elements, "collection_type": "list", "name": "Test upload", "tags": ["name:collection1"] }] payload = { "history_id": self.history_id, "targets": json.dumps(targets), "__files": { "files_0|file_data": open(self.test_data_resolver.get_filename("4.bed")) }, } self.dataset_populator.fetch(payload) hdca = self._assert_one_collection_created_in_history() self.assertEqual(hdca["name"], "Test upload") hdca_tags = hdca["tags"] assert len(hdca_tags) == 1 assert "name:collection1" in hdca_tags assert len(hdca["elements"]) == 1, hdca element0 = hdca["elements"][0] assert element0["element_identifier"] == "4.bed" dataset0 = element0["object"] assert dataset0["file_size"] == 61 dataset_tags = dataset0["tags"] assert len(dataset_tags) == 3, dataset0 def test_upload_nested(self): elements = [{ "name": "samp1", "elements": [{ "src": "files", "dbkey": "hg19", "info": "my cool bed" }] }] targets = [{ "destination": { "type": "hdca" }, "elements": elements, "collection_type": "list:list", "name": "Test upload", }] payload = { "history_id": self.history_id, "targets": json.dumps(targets), "__files": { "files_0|file_data": open(self.test_data_resolver.get_filename("4.bed")) }, } self.dataset_populator.fetch(payload) hdca = self._assert_one_collection_created_in_history() self.assertEqual(hdca["name"], "Test upload") assert len(hdca["elements"]) == 1, hdca element0 = hdca["elements"][0] assert element0["element_identifier"] == "samp1" @skip_if_github_down def test_upload_collection_from_url(self): elements = [{ "src": "url", "url": "https://raw.githubusercontent.com/galaxyproject/galaxy/dev/test-data/4.bed", "info": "my cool bed" }] targets = [{ "destination": { "type": "hdca" }, "elements": elements, "collection_type": "list", }] payload = { "history_id": self.history_id, "targets": json.dumps(targets), "__files": { "files_0|file_data": open(self.test_data_resolver.get_filename("4.bed")) }, } self.dataset_populator.fetch(payload) hdca = self._assert_one_collection_created_in_history() assert len(hdca["elements"]) == 1, hdca element0 = hdca["elements"][0] assert element0["element_identifier"] == "4.bed" assert element0["object"]["file_size"] == 61 @skip_if_github_down def test_upload_collection_failed_expansion_url(self): targets = [{ "destination": { "type": "hdca" }, "elements_from": "bagit", "collection_type": "list", "src": "url", "url": "https://raw.githubusercontent.com/galaxyproject/galaxy/dev/test-data/4.bed", }] payload = { "history_id": self.history_id, "targets": json.dumps(targets), "__files": { "files_0|file_data": open(self.test_data_resolver.get_filename("4.bed")) }, } self.dataset_populator.fetch(payload, assert_ok=False, wait=True) hdca = self._assert_one_collection_created_in_history() assert hdca["populated"] is False assert "bagit.txt" in hdca["populated_state_message"], hdca def _assert_one_collection_created_in_history(self): contents_response = self._get( f"histories/{self.history_id}/contents/dataset_collections") self._assert_status_code_is(contents_response, 200) contents = contents_response.json() assert len(contents) == 1 hdca = contents[0] assert hdca["history_content_type"] == "dataset_collection" hdca_id = hdca["id"] collection_response = self._get( f"histories/{self.history_id}/contents/dataset_collections/{hdca_id}" ) self._assert_status_code_is(collection_response, 200) return collection_response.json() def _check_create_response(self, create_response): self._assert_status_code_is(create_response, 200) dataset_collection = create_response.json() self._assert_has_keys(dataset_collection, "elements", "url", "name", "collection_type", "element_count") return dataset_collection def _download_dataset_collection(self, history_id, hdca_id): return self._get( f"histories/{history_id}/contents/dataset_collections/{hdca_id}/download" ) def test_collection_contents_security(self): # request contents on an hdca that doesn't belong to user hdca, contents_url = self._create_collection_contents_pair() with self._different_user(): contents_response = self._get(contents_url) self._assert_status_code_is(contents_response, 403) def test_collection_contents_invalid_collection(self): # request an invalid collection from a valid hdca, should get 404 hdca, contents_url = self._create_collection_contents_pair() response = self._get(contents_url) self._assert_status_code_is(response, 200) fake_collection_id = '5d7db0757a2eb7ef' fake_contents_url = f"/api/dataset_collections/{hdca['id']}/contents/{fake_collection_id}" error_response = self._get(fake_contents_url) assert_object_id_error(error_response) def test_show_dataset_collection(self): fetch_response = self.dataset_collection_populator.create_list_in_history( self.history_id, direct_upload=True).json() dataset_collection = self.dataset_collection_populator.wait_for_fetched_collection( fetch_response) returned_dce = dataset_collection["elements"] assert len(returned_dce) == 3, dataset_collection hdca_id = dataset_collection['id'] dataset_collection_url = f"/api/dataset_collections/{hdca_id}" dataset_collection = self._get(dataset_collection_url).json() assert dataset_collection['id'] == hdca_id assert dataset_collection['collection_type'] == 'list' def test_show_dataset_collection_contents(self): # Get contents_url from history contents, use it to show the first level # of collection contents in the created HDCA, then use it again to drill # down into the nested collection contents hdca = self.dataset_collection_populator.create_list_of_list_in_history( self.history_id).json() root_contents_url = self._get_contents_url_for_hdca(hdca) # check root contents for this collection root_contents = self._get(root_contents_url).json() assert len(root_contents) == len(hdca['elements']) self._compare_collection_contents_elements(root_contents, hdca['elements']) # drill down, retrieve nested collection contents assert 'object' in root_contents[0] assert 'contents_url' in root_contents[0]['object'] drill_contents_url = root_contents[0]['object']['contents_url'] drill_contents = self._get(drill_contents_url).json() assert len(drill_contents) == len( hdca['elements'][0]['object']['elements']) self._compare_collection_contents_elements( drill_contents, hdca['elements'][0]['object']['elements']) def test_collection_contents_limit_offset(self): # check limit/offset params for collection contents endpoint hdca, root_contents_url = self._create_collection_contents_pair() # check limit limited_contents = self._get(f"{root_contents_url}?limit=1").json() assert len(limited_contents) == 1 assert limited_contents[0]['element_index'] == 0 # check offset offset_contents = self._get(f"{root_contents_url}?offset=1").json() assert len(offset_contents) == 1 assert offset_contents[0]['element_index'] == 1 def test_get_suitable_converters_single_datatype(self): response = self.dataset_collection_populator.upload_collection( self.history_id, "list:paired", elements=[{ "name": "test0", "elements": [ { "src": "pasted", "paste_content": "123\n", "name": "forward", "ext": "bed" }, { "src": "pasted", "paste_content": "456\n", "name": "reverse", "ext": "bed" }, ] }, { "name": "test1", "elements": [ { "src": "pasted", "paste_content": "789\n", "name": "forward", "ext": "bed" }, { "src": "pasted", "paste_content": "0ab\n", "name": "reverse", "ext": "bed" }, ] }]) self._assert_status_code_is(response, 200) hdca_list_id = response.json()["outputs"][0]["id"] converters = self._get("dataset_collections/" + hdca_list_id + "/suitable_converters") expected = [ # This list is subject to change, but it's unlikely we'll be removing converters 'CONVERTER_bed_to_fli_0', 'CONVERTER_bed_gff_or_vcf_to_bigwig_0', 'CONVERTER_bed_to_gff_0', 'CONVERTER_interval_to_bgzip_0', 'tabular_to_csv', 'CONVERTER_interval_to_bed6_0', 'CONVERTER_interval_to_bedstrict_0', 'CONVERTER_interval_to_tabix_0', 'CONVERTER_interval_to_bed12_0' ] actual = [] for converter in converters.json(): actual.append(converter["tool_id"]) missing_expected_converters = set(expected) - set(actual) assert not missing_expected_converters, f"Expected converter(s) {', '.join(missing_expected_converters)} missing from response" def test_get_suitable_converters_different_datatypes_matches(self): response = self.dataset_collection_populator.upload_collection( self.history_id, "list:paired", elements=[{ "name": "test0", "elements": [ { "src": "pasted", "paste_content": "123\n", "name": "forward", "ext": "bed" }, { "src": "pasted", "paste_content": "456\n", "name": "reverse", "ext": "bed" }, ] }, { "name": "test1", "elements": [ { "src": "pasted", "paste_content": "789\n", "name": "forward", "ext": "tabular" }, { "src": "pasted", "paste_content": "0ab\n", "name": "reverse", "ext": "tabular" }, ] }]) self._assert_status_code_is(response, 200) hdca_list_id = response.json()["outputs"][0]["id"] converters = self._get("dataset_collections/" + hdca_list_id + "/suitable_converters") expected = 'tabular_to_csv' actual = [] for converter in converters.json(): actual.append(converter["tool_id"]) assert expected in actual def test_get_suitable_converters_different_datatypes_no_matches(self): response = self.dataset_collection_populator.upload_collection( self.history_id, "list:paired", elements=[{ "name": "test0", "elements": [ { "src": "pasted", "paste_content": "123\n", "name": "forward", "ext": "bed" }, { "src": "pasted", "paste_content": "456\n", "name": "reverse", "ext": "bed" }, ] }, { "name": "test1", "elements": [ { "src": "pasted", "paste_content": "789\n", "name": "forward", "ext": "fasta" }, { "src": "pasted", "paste_content": "0ab\n", "name": "reverse", "ext": "fasta" }, ] }]) self._assert_status_code_is(response, 200) hdca_list_id = response.json()["outputs"][0]["id"] converters = self._get("dataset_collections/" + hdca_list_id + "/suitable_converters") actual: List[str] = [] for converter in converters.json(): actual.append(converter["tool_id"]) assert actual == [] def test_collection_tools_tag_propagation(self): elements = [{"src": "files", "tags": ["name:element_tag"]}] targets = [{ "destination": { "type": "hdca" }, "elements": elements, "collection_type": "list", "name": "Test collection", "tags": ["name:collection_tag"] }] payload = { "history_id": self.history_id, "targets": json.dumps(targets), "__files": { "files_0|file_data": open(self.test_data_resolver.get_filename("4.bed")) }, } hdca_id = self.dataset_populator.fetch( payload).json()['output_collections'][0]['id'] inputs = { "input": { "batch": False, "src": "hdca", "id": hdca_id }, } payload = self.dataset_populator.run_tool_payload( tool_id='__FILTER_FAILED_DATASETS__', inputs=inputs, history_id=self.history_id, input_format='legacy', ) response = self._post("tools", payload).json() self.dataset_populator.wait_for_history(self.history_id, assert_ok=False) output_collection = response["output_collections"][0] # collection should not inherit tags from input collection elements, only parent collection assert output_collection['tags'] == ["name:collection_tag"] element = output_collection['elements'][0] # new element hda should have tags copied from old hda assert element['object']['tags'] == ['name:element_tag'] def _compare_collection_contents_elements(self, contents_elements, hdca_elements): # compare collection api results to existing hdca element contents fields = [ 'element_identifier', 'element_index', 'element_type', 'id', 'model_class' ] for (content_element, hdca_element) in zip(contents_elements, hdca_elements): for f in fields: assert content_element[f] == hdca_element[f] def _create_collection_contents_pair(self): # Create a simple collection, return hdca and contents_url payload = self.dataset_collection_populator.create_pair_payload( self.history_id, instance_type="history") create_response = self._post("dataset_collections", payload, json=True) hdca = self._check_create_response(create_response) root_contents_url = self._get_contents_url_for_hdca(hdca) return hdca, root_contents_url def _get_contents_url_for_hdca(self, hdca): # look up the history contents using optional serialization key history_contents_url = f"histories/{self.history_id}/contents?v=dev&view=summary&keys=contents_url" json = self._get(history_contents_url).json() # filter out the collection we just made id = hdca.id # make sure the contents_url appears def find_hdca(c): return c['history_content_type'] == 'dataset_collection' and c[ 'id'] == hdca['id'] matches = list(filter(find_hdca, json)) assert len(matches) == 1 assert 'contents_url' in matches[0] return matches[0]['contents_url']
class JobsApiTestCase(ApiTestCase): def setUp(self): super(JobsApiTestCase, self).setUp() self.dataset_populator = DatasetPopulator(self.galaxy_interactor) self.dataset_collection_populator = DatasetCollectionPopulator( self.galaxy_interactor) @uses_test_history(require_new=True) def test_index(self, history_id): # Create HDA to ensure at least one job exists... self.__history_with_new_dataset(history_id) jobs = self.__jobs_index() assert "upload1" in map(itemgetter("tool_id"), jobs) @uses_test_history(require_new=True) def test_system_details_admin_only(self, history_id): self.__history_with_new_dataset(history_id) jobs = self.__jobs_index(admin=False) job = jobs[0] self._assert_not_has_keys(job, "command_line", "external_id") jobs = self.__jobs_index(admin=True) job = jobs[0] self._assert_has_keys(job, "command_line", "external_id") @uses_test_history(require_new=True) def test_index_state_filter(self, history_id): # Initial number of ok jobs original_count = len(self.__uploads_with_state("ok")) # Run through dataset upload to ensure num uplaods at least greater # by 1. self.__history_with_ok_dataset(history_id) # Verify number of ok jobs is actually greater. count_increased = False for _ in range(10): new_count = len(self.__uploads_with_state("ok")) if original_count < new_count: count_increased = True break time.sleep(.1) if not count_increased: template = "Jobs in ok state did not increase (was %d, now %d)" message = template % (original_count, new_count) raise AssertionError(message) @uses_test_history(require_new=True) def test_index_date_filter(self, history_id): self.__history_with_new_dataset(history_id) two_weeks_ago = (datetime.datetime.utcnow() - datetime.timedelta(7)).isoformat() last_week = (datetime.datetime.utcnow() - datetime.timedelta(7)).isoformat() next_week = (datetime.datetime.utcnow() + datetime.timedelta(7)).isoformat() today = datetime.datetime.utcnow().isoformat() tomorrow = (datetime.datetime.utcnow() + datetime.timedelta(1)).isoformat() jobs = self.__jobs_index(data={ "date_range_min": today[0:10], "date_range_max": tomorrow[0:10] }) assert len(jobs) > 0 today_job_id = jobs[0]["id"] jobs = self.__jobs_index(data={ "date_range_min": two_weeks_ago, "date_range_max": last_week }) assert today_job_id not in map(itemgetter("id"), jobs) jobs = self.__jobs_index(data={ "date_range_min": last_week, "date_range_max": next_week }) assert today_job_id in map(itemgetter("id"), jobs) @uses_test_history(require_new=True) def test_index_history(self, history_id): self.__history_with_new_dataset(history_id) jobs = self.__jobs_index(data={"history_id": history_id}) assert len(jobs) > 0 with self.dataset_populator.test_history() as other_history_id: jobs = self.__jobs_index(data={"history_id": other_history_id}) assert len(jobs) == 0 @uses_test_history(require_new=True) def test_index_multiple_states_filter(self, history_id): # Initial number of ok jobs original_count = len(self.__uploads_with_state("ok", "new")) # Run through dataset upload to ensure num uplaods at least greater # by 1. self.__history_with_ok_dataset(history_id) # Verify number of ok jobs is actually greater. new_count = len(self.__uploads_with_state("new", "ok")) assert original_count < new_count, new_count @uses_test_history(require_new=True) def test_show(self, history_id): # Create HDA to ensure at least one job exists... self.__history_with_new_dataset(history_id) jobs_response = self._get("jobs") first_job = jobs_response.json()[0] self._assert_has_key(first_job, 'id', 'state', 'exit_code', 'update_time', 'create_time') job_id = first_job["id"] show_jobs_response = self._get("jobs/%s" % job_id) self._assert_status_code_is(show_jobs_response, 200) job_details = show_jobs_response.json() self._assert_has_key(job_details, 'id', 'state', 'exit_code', 'update_time', 'create_time') show_jobs_response = self._get("jobs/%s" % job_id, {"full": True}) self._assert_status_code_is(show_jobs_response, 200) job_details = show_jobs_response.json() self._assert_has_key(job_details, 'id', 'state', 'exit_code', 'update_time', 'create_time', 'stdout', 'stderr', 'job_messages') @uses_test_history(require_new=True) def test_show_security(self, history_id): self.__history_with_new_dataset(history_id) jobs_response = self._get("jobs", data={"history_id": history_id}) job = jobs_response.json()[0] job_id = job["id"] show_jobs_response = self._get("jobs/%s" % job_id, admin=False) self._assert_not_has_keys(show_jobs_response.json(), "command_line", "external_id") # TODO: Re-activate test case when API accepts privacy settings # with self._different_user(): # show_jobs_response = self._get( "jobs/%s" % job_id, admin=False ) # self._assert_status_code_is( show_jobs_response, 200 ) show_jobs_response = self._get("jobs/%s" % job_id, admin=True) self._assert_has_keys(show_jobs_response.json(), "command_line", "external_id") def _run_detect_errors(self, history_id, inputs): payload = self.dataset_populator.run_tool_payload( tool_id='detect_errors_aggressive', inputs=inputs, history_id=history_id, ) return self._post("tools", data=payload).json() @skip_without_tool("detect_errors_aggressive") def test_unhide_on_error(self): with self.dataset_populator.test_history() as history_id: inputs = {'error_bool': 'true'} run_response = self._run_detect_errors(history_id=history_id, inputs=inputs) job_id = run_response['jobs'][0]["id"] self.dataset_populator.wait_for_job(job_id) job = self.dataset_populator.get_job_details(job_id).json() assert job['state'] == 'error' dataset = self.dataset_populator.get_history_dataset_details( history_id=history_id, dataset_id=run_response['outputs'][0]['id'], assert_ok=False) assert dataset['visible'] @skip_without_tool("detect_errors_aggressive") def test_no_unhide_on_error_if_mapped_over(self): with self.dataset_populator.test_history() as history_id: hdca1 = self.dataset_collection_populator.create_list_in_history( history_id, contents=[("sample1-1", "1 2 3")]).json() inputs = { 'error_bool': 'true', 'dataset': { 'batch': True, 'values': [{ 'src': 'hdca', 'id': hdca1['id'] }], } } run_response = self._run_detect_errors(history_id=history_id, inputs=inputs) job_id = run_response['jobs'][0]["id"] self.dataset_populator.wait_for_job(job_id) job = self.dataset_populator.get_job_details(job_id).json() assert job['state'] == 'error' dataset = self.dataset_populator.get_history_dataset_details( history_id=history_id, dataset_id=run_response['outputs'][0]['id'], assert_ok=False) assert not dataset['visible'] @skip_without_tool('empty_output') def test_common_problems(self): with self.dataset_populator.test_history() as history_id: empty_run_response = self.dataset_populator.run_tool( tool_id='empty_output', inputs={}, history_id=history_id, ) empty_hda = empty_run_response["outputs"][0] cat_empty_twice_run_response = self.dataset_populator.run_tool( tool_id='cat1', inputs={ 'input1': { 'src': 'hda', 'id': empty_hda['id'] }, 'queries_0|input2': { 'src': 'hda', 'id': empty_hda['id'] } }, history_id=history_id, ) empty_output_job = empty_run_response["jobs"][0] cat_empty_job = cat_empty_twice_run_response["jobs"][0] empty_output_common_problems_response = self._get( 'jobs/%s/common_problems' % empty_output_job["id"]).json() cat_empty_common_problems_response = self._get( 'jobs/%s/common_problems' % cat_empty_job["id"]).json() self._assert_has_keys(empty_output_common_problems_response, "has_empty_inputs", "has_duplicate_inputs") self._assert_has_keys(cat_empty_common_problems_response, "has_empty_inputs", "has_duplicate_inputs") assert not empty_output_common_problems_response["has_empty_inputs"] assert cat_empty_common_problems_response["has_empty_inputs"] assert not empty_output_common_problems_response[ "has_duplicate_inputs"] assert cat_empty_common_problems_response["has_duplicate_inputs"] @skip_without_tool('detect_errors_aggressive') def test_report_error(self): with self.dataset_populator.test_history() as history_id: payload = self.dataset_populator.run_tool_payload( tool_id='detect_errors_aggressive', inputs={'error_bool': 'true'}, history_id=history_id, ) run_response = self._post("tools", data=payload).json() job_id = run_response['jobs'][0]["id"] self.dataset_populator.wait_for_job(job_id) dataset_id = run_response['outputs'][0]['id'] response = self._post('jobs/%s/error' % job_id, data={'dataset_id': dataset_id}) assert response.status_code == 200, response.text @skip_without_tool('detect_errors_aggressive') def test_report_error_anon(self): # Need to get a cookie and use that for anonymous tool runs cookies = requests.get(self.url).cookies payload = json.dumps({ "tool_id": "detect_errors_aggressive", "inputs": { "error_bool": "true" } }) run_response = requests.post("%s/tools" % self.galaxy_interactor.api_url, data=payload, cookies=cookies).json() job_id = run_response['jobs'][0]["id"] dataset_id = run_response['outputs'][0]['id'] response = requests.post('%s/jobs/%s/error' % (self.galaxy_interactor.api_url, job_id), data={ 'email': '*****@*****.**', 'dataset_id': dataset_id }, cookies=cookies) assert response.status_code == 200, response.text @uses_test_history(require_new=True) def test_deleting_output_keep_running_until_all_deleted(self, history_id): job_state, outputs = self._setup_running_two_output_job( history_id, 120) self._hack_to_skip_test_if_state_ok(job_state) # Delete one of the two outputs and make sure the job is still running. self._raw_update_history_item(history_id, outputs[0]["id"], {"deleted": True}) self._hack_to_skip_test_if_state_ok(job_state) time.sleep(1) self._hack_to_skip_test_if_state_ok(job_state) state = job_state().json()["state"] assert state == "running", state # Delete the second output and make sure the job is cancelled. self._raw_update_history_item(history_id, outputs[1]["id"], {"deleted": True}) final_state = wait_on_state(job_state, assert_ok=False, timeout=15) assert final_state in ["deleted_new", "deleted"], final_state @uses_test_history(require_new=True) def test_purging_output_keep_running_until_all_purged(self, history_id): job_state, outputs = self._setup_running_two_output_job( history_id, 120) # Pretty much right away after the job is running, these paths should be populated - # if they are grab them and make sure they are deleted at the end of the job. dataset_1 = self._get_history_item_as_admin(history_id, outputs[0]["id"]) dataset_2 = self._get_history_item_as_admin(history_id, outputs[1]["id"]) if "file_name" in dataset_1: output_dataset_paths = [ dataset_1["file_name"], dataset_2["file_name"] ] # This may or may not exist depending on if the test is local or not. output_dataset_paths_exist = os.path.exists( output_dataset_paths[0]) else: output_dataset_paths = [] output_dataset_paths_exist = False self._hack_to_skip_test_if_state_ok(job_state) current_state = job_state().json()["state"] assert current_state == "running", current_state # Purge one of the two outputs and make sure the job is still running. self._raw_update_history_item(history_id, outputs[0]["id"], {"purged": True}) time.sleep(1) self._hack_to_skip_test_if_state_ok(job_state) current_state = job_state().json()["state"] assert current_state == "running", current_state # Purge the second output and make sure the job is cancelled. self._raw_update_history_item(history_id, outputs[1]["id"], {"purged": True}) final_state = wait_on_state(job_state, assert_ok=False, timeout=15) assert final_state in ["deleted_new", "deleted"], final_state def paths_deleted(): if not os.path.exists( output_dataset_paths[0]) and not os.path.exists( output_dataset_paths[1]): return True if output_dataset_paths_exist: wait_on(paths_deleted, "path deletion") @uses_test_history(require_new=True) def test_purging_output_cleaned_after_ok_run(self, history_id): job_state, outputs = self._setup_running_two_output_job(history_id, 10) # Pretty much right away after the job is running, these paths should be populated - # if they are grab them and make sure they are deleted at the end of the job. dataset_1 = self._get_history_item_as_admin(history_id, outputs[0]["id"]) dataset_2 = self._get_history_item_as_admin(history_id, outputs[1]["id"]) if "file_name" in dataset_1: output_dataset_paths = [ dataset_1["file_name"], dataset_2["file_name"] ] # This may or may not exist depending on if the test is local or not. output_dataset_paths_exist = os.path.exists( output_dataset_paths[0]) else: output_dataset_paths = [] output_dataset_paths_exist = False if not output_dataset_paths_exist: # Given this Galaxy configuration - there is nothing more to be tested here. # Consider throwing a skip instead. return # Purge one of the two outputs and wait for the job to complete. self._raw_update_history_item(history_id, outputs[0]["id"], {"purged": True}) wait_on_state(job_state, assert_ok=True) if output_dataset_paths_exist: time.sleep(.5) # Make sure the non-purged dataset is on disk and the purged one is not. assert os.path.exists(output_dataset_paths[1]) assert not os.path.exists(output_dataset_paths[0]) def _hack_to_skip_test_if_state_ok(self, job_state): from nose.plugins.skip import SkipTest if job_state().json()["state"] == "ok": message = "Job state switch from running to ok too quickly - the rest of the test requires the job to be in a running state. Skipping test." raise SkipTest(message) def _setup_running_two_output_job(self, history_id, sleep_time): payload = self.dataset_populator.run_tool_payload( tool_id='create_2', inputs=dict(sleep_time=sleep_time, ), history_id=history_id, ) run_response = self._post("tools", data=payload).json() outputs = run_response["outputs"] jobs = run_response["jobs"] assert len(outputs) == 2 assert len(jobs) == 1 def job_state(): jobs_response = self._get("jobs/%s" % jobs[0]["id"]) return jobs_response # Give job some time to get up and running. time.sleep(2) running_state = wait_on_state(job_state, skip_states=["queued", "new"], assert_ok=False, timeout=15) assert running_state == "running", running_state def job_state(): jobs_response = self._get("jobs/%s" % jobs[0]["id"]) return jobs_response return job_state, outputs def _raw_update_history_item(self, history_id, item_id, data): update_url = self._api_url("histories/%s/contents/%s" % (history_id, item_id), use_key=True) update_response = requests.put(update_url, json=data) assert_status_code_is_ok(update_response) return update_response @skip_without_tool("cat_data_and_sleep") @uses_test_history(require_new=True) def test_resume_job(self, history_id): hda1 = self.dataset_populator.new_dataset( history_id, content="samp1\t10.0\nsamp2\t20.0\n") hda2 = self.dataset_populator.new_dataset( history_id, content="samp1\t30.0\nsamp2\t40.0\n") # Submit first job payload = self.dataset_populator.run_tool_payload( tool_id='cat_data_and_sleep', inputs={ 'sleep_time': 15, 'input1': { 'src': 'hda', 'id': hda2['id'] }, 'queries_0|input2': { 'src': 'hda', 'id': hda2['id'] } }, history_id=history_id, ) run_response = self._post("tools", data=payload).json() output = run_response["outputs"][0] # Submit second job that waits on job1 payload = self.dataset_populator.run_tool_payload( tool_id='cat1', inputs={ 'input1': { 'src': 'hda', 'id': hda1['id'] }, 'queries_0|input2': { 'src': 'hda', 'id': output['id'] } }, history_id=history_id, ) run_response = self._post("tools", data=payload).json() job_id = run_response['jobs'][0]['id'] output = run_response["outputs"][0] # Delete second jobs input while second job is waiting for first job delete_response = self._delete("histories/%s/contents/%s" % (history_id, hda1['id'])) self._assert_status_code_is(delete_response, 200) self.dataset_populator.wait_for_history_jobs(history_id, assert_ok=False) dataset_details = self._get("histories/%s/contents/%s" % (history_id, output['id'])).json() assert dataset_details['state'] == 'paused' # Undelete input dataset undelete_response = self._put("histories/%s/contents/%s" % (history_id, hda1['id']), data=json.dumps({'deleted': False})) self._assert_status_code_is(undelete_response, 200) resume_response = self._put("jobs/%s/resume" % job_id) self._assert_status_code_is(resume_response, 200) self.dataset_populator.wait_for_history_jobs(history_id, assert_ok=True) dataset_details = self._get("histories/%s/contents/%s" % (history_id, output['id'])).json() assert dataset_details['state'] == 'ok' def _get_history_item_as_admin(self, history_id, item_id): response = self._get("histories/%s/contents/%s?view=detailed" % (history_id, item_id), admin=True) assert_status_code_is_ok(response) return response.json() @uses_test_history(require_new=True) def test_search(self, history_id): dataset_id = self.__history_with_ok_dataset(history_id) # We first copy the datasets, so that the update time is lower than the job creation time new_history_id = self.dataset_populator.new_history() copy_payload = { "content": dataset_id, "source": "hda", "type": "dataset" } copy_response = self._post("histories/%s/contents" % new_history_id, data=copy_payload) self._assert_status_code_is(copy_response, 200) inputs = json.dumps({'input1': {'src': 'hda', 'id': dataset_id}}) self._job_search(tool_id='cat1', history_id=history_id, inputs=inputs) # We test that a job can be found even if the dataset has been copied to another history new_dataset_id = copy_response.json()['id'] copied_inputs = json.dumps( {'input1': { 'src': 'hda', 'id': new_dataset_id }}) search_payload = self._search_payload(history_id=history_id, tool_id='cat1', inputs=copied_inputs) self._search(search_payload, expected_search_count=1) # Now we delete the original input HDA that was used -- we should still be able to find the job delete_respone = self._delete("histories/%s/contents/%s" % (history_id, dataset_id)) self._assert_status_code_is(delete_respone, 200) self._search(search_payload, expected_search_count=1) # Now we also delete the copy -- we shouldn't find a job delete_respone = self._delete("histories/%s/contents/%s" % (new_history_id, new_dataset_id)) self._assert_status_code_is(delete_respone, 200) self._search(search_payload, expected_search_count=0) @uses_test_history(require_new=True) def test_search_handle_identifiers(self, history_id): # Test that input name and element identifier of a jobs' output must match for a job to be returned. dataset_id = self.__history_with_ok_dataset(history_id) inputs = json.dumps({'input1': {'src': 'hda', 'id': dataset_id}}) self._job_search(tool_id='identifier_single', history_id=history_id, inputs=inputs) dataset_details = self._get("histories/%s/contents/%s" % (history_id, dataset_id)).json() dataset_details['name'] = 'Renamed Test Dataset' dataset_update_response = self._put( "histories/%s/contents/%s" % (history_id, dataset_id), data=dict(name='Renamed Test Dataset')) self._assert_status_code_is(dataset_update_response, 200) assert dataset_update_response.json()['name'] == 'Renamed Test Dataset' search_payload = self._search_payload(history_id=history_id, tool_id='identifier_single', inputs=inputs) self._search(search_payload, expected_search_count=0) @uses_test_history(require_new=True) def test_search_delete_outputs(self, history_id): dataset_id = self.__history_with_ok_dataset(history_id) inputs = json.dumps({'input1': {'src': 'hda', 'id': dataset_id}}) tool_response = self._job_search(tool_id='cat1', history_id=history_id, inputs=inputs) output_id = tool_response.json()['outputs'][0]['id'] delete_respone = self._delete("histories/%s/contents/%s" % (history_id, output_id)) self._assert_status_code_is(delete_respone, 200) search_payload = self._search_payload(history_id=history_id, tool_id='cat1', inputs=inputs) self._search(search_payload, expected_search_count=0) @uses_test_history(require_new=True) def test_search_with_hdca_list_input(self, history_id): list_id_a = self.__history_with_ok_collection(collection_type='list', history_id=history_id) list_id_b = self.__history_with_ok_collection(collection_type='list', history_id=history_id) inputs = json.dumps({ 'f1': { 'src': 'hdca', 'id': list_id_a }, 'f2': { 'src': 'hdca', 'id': list_id_b }, }) tool_response = self._job_search(tool_id='multi_data_param', history_id=history_id, inputs=inputs) # We switch the inputs, this should not return a match inputs_switched = json.dumps({ 'f2': { 'src': 'hdca', 'id': list_id_a }, 'f1': { 'src': 'hdca', 'id': list_id_b }, }) search_payload = self._search_payload(history_id=history_id, tool_id='multi_data_param', inputs=inputs_switched) self._search(search_payload, expected_search_count=0) # We delete the ouput (this is a HDA, as multi_data_param reduces collections) # and use the correct input job definition, the job should not be found output_id = tool_response.json()['outputs'][0]['id'] delete_respone = self._delete("histories/%s/contents/%s" % (history_id, output_id)) self._assert_status_code_is(delete_respone, 200) search_payload = self._search_payload(history_id=history_id, tool_id='multi_data_param', inputs=inputs) self._search(search_payload, expected_search_count=0) @uses_test_history(require_new=True) def test_search_delete_hdca_output(self, history_id): list_id_a = self.__history_with_ok_collection(collection_type='list', history_id=history_id) inputs = json.dumps({ 'input1': { 'src': 'hdca', 'id': list_id_a }, }) tool_response = self._job_search(tool_id='collection_creates_list', history_id=history_id, inputs=inputs) output_id = tool_response.json()['outputs'][0]['id'] # We delete a single tool output, no job should be returned delete_respone = self._delete("histories/%s/contents/%s" % (history_id, output_id)) self._assert_status_code_is(delete_respone, 200) search_payload = self._search_payload( history_id=history_id, tool_id='collection_creates_list', inputs=inputs) self._search(search_payload, expected_search_count=0) tool_response = self._job_search(tool_id='collection_creates_list', history_id=history_id, inputs=inputs) output_collection_id = tool_response.json( )['output_collections'][0]['id'] # We delete a collection output, no job should be returned delete_respone = self._delete( "histories/%s/contents/dataset_collections/%s" % (history_id, output_collection_id)) self._assert_status_code_is(delete_respone, 200) search_payload = self._search_payload( history_id=history_id, tool_id='collection_creates_list', inputs=inputs) self._search(search_payload, expected_search_count=0) @uses_test_history(require_new=True) def test_search_with_hdca_pair_input(self, history_id): list_id_a = self.__history_with_ok_collection(collection_type='pair', history_id=history_id) inputs = json.dumps({ 'f1': { 'src': 'hdca', 'id': list_id_a }, 'f2': { 'src': 'hdca', 'id': list_id_a }, }) self._job_search(tool_id='multi_data_param', history_id=history_id, inputs=inputs) # We test that a job can be found even if the collection has been copied to another history new_history_id = self.dataset_populator.new_history() copy_payload = { "content": list_id_a, "source": "hdca", "type": "dataset_collection" } copy_response = self._post("histories/%s/contents" % new_history_id, data=copy_payload) self._assert_status_code_is(copy_response, 200) new_list_a = copy_response.json()['id'] copied_inputs = json.dumps({ 'f1': { 'src': 'hdca', 'id': new_list_a }, 'f2': { 'src': 'hdca', 'id': new_list_a }, }) search_payload = self._search_payload(history_id=new_history_id, tool_id='multi_data_param', inputs=copied_inputs) self._search(search_payload, expected_search_count=1) # Now we delete the original input HDCA that was used -- we should still be able to find the job delete_respone = self._delete( "histories/%s/contents/dataset_collections/%s" % (history_id, list_id_a)) self._assert_status_code_is(delete_respone, 200) self._search(search_payload, expected_search_count=1) # Now we also delete the copy -- we shouldn't find a job delete_respone = self._delete( "histories/%s/contents/dataset_collections/%s" % (history_id, new_list_a)) self._assert_status_code_is(delete_respone, 200) self._search(search_payload, expected_search_count=0) @uses_test_history(require_new=True) def test_search_with_hdca_list_pair_input(self, history_id): list_id_a = self.__history_with_ok_collection( collection_type='list:pair', history_id=history_id) inputs = json.dumps({ 'f1': { 'src': 'hdca', 'id': list_id_a }, 'f2': { 'src': 'hdca', 'id': list_id_a }, }) self._job_search(tool_id='multi_data_param', history_id=history_id, inputs=inputs) def _job_search(self, tool_id, history_id, inputs): search_payload = self._search_payload(history_id=history_id, tool_id=tool_id, inputs=inputs) empty_search_response = self._post("jobs/search", data=search_payload) self._assert_status_code_is(empty_search_response, 200) self.assertEqual(len(empty_search_response.json()), 0) tool_response = self._post("tools", data=search_payload) self.dataset_populator.wait_for_tool_run(history_id, run_response=tool_response) self._search(search_payload, expected_search_count=1) return tool_response def _search_payload(self, history_id, tool_id, inputs, state='ok'): search_payload = dict(tool_id=tool_id, inputs=inputs, history_id=history_id, state=state) return search_payload def _search(self, payload, expected_search_count=1): # in case job and history aren't updated at exactly the same # time give time to wait for _ in range(5): search_count = self._search_count(payload) if search_count == expected_search_count: break time.sleep(1) assert search_count == expected_search_count, "expected to find %d jobs, got %d jobs" % ( expected_search_count, search_count) return search_count def _search_count(self, search_payload): search_response = self._post("jobs/search", data=search_payload) self._assert_status_code_is(search_response, 200) search_json = search_response.json() return len(search_json) def __uploads_with_state(self, *states): jobs_response = self._get("jobs", data=dict(state=states)) self._assert_status_code_is(jobs_response, 200) jobs = jobs_response.json() assert not [j for j in jobs if not j['state'] in states] return [j for j in jobs if j['tool_id'] == 'upload1'] def __history_with_new_dataset(self, history_id): dataset_id = self.dataset_populator.new_dataset(history_id)["id"] return dataset_id def __history_with_ok_dataset(self, history_id): dataset_id = self.dataset_populator.new_dataset(history_id, wait=True)["id"] return dataset_id def __history_with_ok_collection(self, collection_type='list', history_id=None): if not history_id: history_id = self.dataset_populator.new_history() if collection_type == 'list': fetch_response = self.dataset_collection_populator.create_list_in_history( history_id, direct_upload=True).json() elif collection_type == 'pair': fetch_response = self.dataset_collection_populator.create_pair_in_history( history_id, direct_upload=True).json() elif collection_type == 'list:pair': fetch_response = self.dataset_collection_populator.create_list_of_pairs_in_history( history_id).json() self.dataset_collection_populator.wait_for_fetched_collection( fetch_response) return fetch_response["outputs"][0]['id'] def __jobs_index(self, **kwds): jobs_response = self._get("jobs", **kwds) self._assert_status_code_is(jobs_response, 200) jobs = jobs_response.json() assert isinstance(jobs, list) return jobs
class HistoryContentsApiTestCase(ApiTestCase): def setUp(self): super().setUp() self.dataset_populator = DatasetPopulator(self.galaxy_interactor) self.dataset_collection_populator = DatasetCollectionPopulator( self.galaxy_interactor) self.library_populator = LibraryPopulator(self.galaxy_interactor) self.history_id = self.dataset_populator.new_history() def test_index_hda_summary(self): hda1 = self.dataset_populator.new_dataset(self.history_id) contents_response = self._get(f"histories/{self.history_id}/contents") hda_summary = self.__check_for_hda(contents_response, hda1) assert "display_types" not in hda_summary # Quick summary, not full details def test_make_private_and_public(self): hda1 = self._wait_for_new_hda() update_url = f"histories/{self.history_id}/contents/{hda1['id']}/permissions" role_id = self.dataset_populator.user_private_role_id() # Give manage permission to the user. payload = { "access": [], "manage": [role_id], } update_response = self._update_permissions(update_url, payload, admin=True) self._assert_status_code_is(update_response, 200) self._assert_other_user_can_access(hda1["id"]) # Then we restrict access. payload = { "action": "make_private", } update_response = self._update_permissions(update_url, payload) self._assert_status_code_is(update_response, 200) self._assert_other_user_cannot_access(hda1["id"]) # Then we restrict access. payload = { "action": "remove_restrictions", } update_response = self._update_permissions(update_url, payload) self._assert_status_code_is(update_response, 200) self._assert_other_user_can_access(hda1["id"]) def test_set_permissions_add_admin_history_contents(self): self._verify_dataset_permissions("history_contents") def test_set_permissions_add_admin_datasets(self): self._verify_dataset_permissions("dataset") def _verify_dataset_permissions(self, api_endpoint): hda1 = self._wait_for_new_hda() hda_id = hda1["id"] if api_endpoint == "history_contents": update_url = f"histories/{self.history_id}/contents/{hda_id}/permissions" else: update_url = f"datasets/{hda_id}/permissions" role_id = self.dataset_populator.user_private_role_id() payload = { "access": [role_id], "manage": [role_id], } # Other users cannot modify permissions. with self._different_user(): update_response = self._update_permissions(update_url, payload) self._assert_status_code_is(update_response, 403) # First the details render for another user. self._assert_other_user_can_access(hda_id) # Then we restrict access. update_response = self._update_permissions(update_url, payload, admin=True) self._assert_status_code_is(update_response, 200) # Finally the details don't render. self._assert_other_user_cannot_access(hda_id) # But they do for the original user. contents_response = self._get( f"histories/{self.history_id}/contents/{hda_id}").json() assert "name" in contents_response update_response = self._update_permissions(update_url, payload) self._assert_status_code_is(update_response, 200) payload = { "access": [role_id], "manage": [role_id], } update_response = self._update_permissions(update_url, payload) self._assert_status_code_is(update_response, 200) self._assert_other_user_cannot_access(hda_id) user_id = self.dataset_populator.user_id() with self._different_user(): different_user_id = self.dataset_populator.user_id() combined_user_role = self.dataset_populator.create_role( [user_id, different_user_id], description="role for testing permissions") payload = { "access": [combined_user_role["id"]], "manage": [role_id], } update_response = self._update_permissions(update_url, payload) self._assert_status_code_is(update_response, 200) # Now other user can see dataset again with access permission. self._assert_other_user_can_access(hda_id) # access doesn't imply management though... with self._different_user(): update_response = self._update_permissions(update_url, payload) self._assert_status_code_is(update_response, 403) def _assert_other_user_cannot_access(self, history_content_id): with self._different_user(): contents_response = self.dataset_populator.get_history_dataset_details_raw( history_id=self.history_id, dataset_id=history_content_id) assert contents_response.status_code == 403 def _assert_other_user_can_access(self, history_content_id): with self._different_user(): contents_response = self.dataset_populator.get_history_dataset_details_raw( history_id=self.history_id, dataset_id=history_content_id) contents_response.raise_for_status() assert "name" in contents_response.json() def test_index_hda_all_details(self): hda1 = self.dataset_populator.new_dataset(self.history_id) contents_response = self._get( f"histories/{self.history_id}/contents?details=all") hda_details = self.__check_for_hda(contents_response, hda1) self.__assert_hda_has_full_details(hda_details) def test_index_hda_detail_by_id(self): hda1 = self.dataset_populator.new_dataset(self.history_id) contents_response = self._get( f"histories/{self.history_id}/contents?details={hda1['id']}") hda_details = self.__check_for_hda(contents_response, hda1) self.__assert_hda_has_full_details(hda_details) def test_show_hda(self): hda1 = self.dataset_populator.new_dataset(self.history_id) show_response = self.__show(hda1) self._assert_status_code_is(show_response, 200) self.__assert_matches_hda(hda1, show_response.json()) def _create_copy(self): hda1 = self.dataset_populator.new_dataset(self.history_id) create_data = dict( source='hda', content=hda1["id"], ) second_history_id = self.dataset_populator.new_history() assert self.__count_contents(second_history_id) == 0 create_response = self._post(f"histories/{second_history_id}/contents", create_data, json=True) self._assert_status_code_is(create_response, 200) return create_response.json() def test_hda_copy(self): response = self._create_copy() assert self.__count_contents(response['history_id']) == 1 def test_inheritance_chain(self): response = self._create_copy() inheritance_chain_response = self._get( f"datasets/{response['id']}/inheritance_chain") self._assert_status_code_is_ok(inheritance_chain_response) inheritance_chain = inheritance_chain_response.json() assert len(inheritance_chain) == 1 def test_library_copy(self): ld = self.library_populator.new_library_dataset("lda_test_library") create_data = dict( source='library', content=ld["id"], ) assert self.__count_contents(self.history_id) == 0 create_response = self._post(f"histories/{self.history_id}/contents", create_data, json=True) self._assert_status_code_is(create_response, 200) assert self.__count_contents(self.history_id) == 1 def test_update(self): hda1 = self._wait_for_new_hda() assert str(hda1["deleted"]).lower() == "false" update_response = self._update(hda1["id"], dict(deleted=True)) self._assert_status_code_is(update_response, 200) show_response = self.__show(hda1) assert str(show_response.json()["deleted"]).lower() == "true" update_response = self._update(hda1["id"], dict(name="Updated Name")) assert self.__show(hda1).json()["name"] == "Updated Name" update_response = self._update(hda1["id"], dict(name="Updated Name")) assert self.__show(hda1).json()["name"] == "Updated Name" unicode_name = 'ржевский сапоги' update_response = self._update(hda1["id"], dict(name=unicode_name)) updated_hda = self.__show(hda1).json() assert updated_hda["name"] == unicode_name, updated_hda quoted_name = '"Mooo"' update_response = self._update(hda1["id"], dict(name=quoted_name)) updated_hda = self.__show(hda1).json() assert updated_hda["name"] == quoted_name, quoted_name data = { "dataset_id": hda1["id"], "name": "moocow", "dbkey": "?", "annotation": None, "info": "my info is", "operation": "attributes" } update_response = self._set_edit_update(data) # No key or anything supplied, expect a permission problem. # A bit questionable but I think this is a 400 instead of a 403 so that # we don't distinguish between this is a valid ID you don't have access to # and this is an invalid ID. assert update_response.status_code == 400, update_response.content def test_update_batch(self): hda1 = self._wait_for_new_hda() assert str(hda1["deleted"]).lower() == "false" assert str(hda1["visible"]).lower() == "true" # update deleted flag => true payload = dict(items=[{ "history_content_type": "dataset", "id": hda1["id"] }], deleted=True) update_response = self._update_batch(payload) objects = update_response.json() assert objects[0]["deleted"] is True assert objects[0]["visible"] is True # update visibility flag => false payload = dict(items=[{ "history_content_type": "dataset", "id": hda1["id"] }], visible=False) update_response = self._update_batch(payload) objects = update_response.json() assert objects[0]["deleted"] is True assert objects[0]["visible"] is False # update both flags payload = dict(items=[{ "history_content_type": "dataset", "id": hda1["id"] }], deleted=False, visible=True) update_response = self._update_batch(payload) objects = update_response.json() assert objects[0]["deleted"] is False assert objects[0]["visible"] is True def test_update_batch_collections(self): hdca = self._create_pair_collection() assert hdca["deleted"] is False assert hdca["visible"] is True # update deleted flag => true payload = dict(items=[{ "history_content_type": "dataset_collection", "id": hdca["id"] }], deleted=True) update_response = self._update_batch(payload) objects = update_response.json() assert objects[0]["deleted"] is True assert objects[0]["visible"] is True # update visibility flag => false payload = dict(items=[{ "history_content_type": "dataset_collection", "id": hdca["id"] }], visible=False) update_response = self._update_batch(payload) objects = update_response.json() assert objects[0]["deleted"] is True assert objects[0]["visible"] is False # update both flags payload = dict(items=[{ "history_content_type": "dataset_collection", "id": hdca["id"] }], deleted=False, visible=True) update_response = self._update_batch(payload) objects = update_response.json() assert objects[0]["deleted"] is False assert objects[0]["visible"] is True def test_update_type_failures(self): hda1 = self._wait_for_new_hda() update_response = self._update(hda1["id"], dict(deleted='not valid')) self._assert_status_code_is(update_response, 400) def _wait_for_new_hda(self): hda1 = self.dataset_populator.new_dataset(self.history_id) self.dataset_populator.wait_for_history(self.history_id) return hda1 def _set_edit_update(self, data): update_response = self._put(urllib.parse.urljoin( self.url, "dataset/set_edit"), data=data, json=True) return update_response def _update(self, item_id, data, admin=False, history_id=None): history_id = history_id or self.history_id update_response = self._put( f"histories/{history_id}/contents/{item_id}", data=data, json=True, admin=admin) return update_response def _update_permissions(self, url, data, admin=False): update_response = self._put(url, data=data, json=True, admin=admin) return update_response def _update_batch(self, data): update_response = self._put(f"histories/{self.history_id}/contents", data=data, json=True) return update_response def test_delete(self): hda1 = self.dataset_populator.new_dataset(self.history_id) self.dataset_populator.wait_for_history(self.history_id) assert str(self.__show(hda1).json()["deleted"]).lower() == "false" delete_response = self._delete( f"histories/{self.history_id}/contents/{hda1['id']}") assert delete_response.status_code < 300 # Something in the 200s :). assert str(self.__show(hda1).json()["deleted"]).lower() == "true" def test_delete_anon(self): with self._different_user(anon=True): history_id = self._get( urllib.parse.urljoin( self.url, "history/current_history_json")).json()['id'] hda1 = self.dataset_populator.new_dataset(history_id) self.dataset_populator.wait_for_history(history_id) assert str(self.__show(hda1).json()["deleted"]).lower() == "false" delete_response = self._delete( f"histories/{history_id}/contents/{hda1['id']}") assert delete_response.status_code < 300 # Something in the 200s :). assert str(self.__show(hda1).json()["deleted"]).lower() == "true" def test_delete_permission_denied(self): hda1 = self.dataset_populator.new_dataset(self.history_id) with self._different_user(anon=True): delete_response = self._delete( f"histories/{self.history_id}/contents/{hda1['id']}") assert delete_response.status_code == 403 assert delete_response.json( )['err_msg'] == 'HistoryDatasetAssociation is not owned by user' def test_purge(self): hda1 = self.dataset_populator.new_dataset(self.history_id) self.dataset_populator.wait_for_history(self.history_id) assert str(self.__show(hda1).json()["deleted"]).lower() == "false" assert str(self.__show(hda1).json()["purged"]).lower() == "false" data = {'purge': True} delete_response = self._delete( f"histories/{self.history_id}/contents/{hda1['id']}", data=data, json=True) assert delete_response.status_code < 300 # Something in the 200s :). assert str(self.__show(hda1).json()["deleted"]).lower() == "true" assert str(self.__show(hda1).json()["purged"]).lower() == "true" def test_dataset_collection_creation_on_contents(self): payload = self.dataset_collection_populator.create_pair_payload( self.history_id, type="dataset_collection") endpoint = f"histories/{self.history_id}/contents" self._check_pair_creation(endpoint, payload) def test_dataset_collection_creation_on_typed_contents(self): payload = self.dataset_collection_populator.create_pair_payload( self.history_id, ) endpoint = f"histories/{self.history_id}/contents/dataset_collections" self._check_pair_creation(endpoint, payload) def test_dataset_collection_create_from_exisiting_datasets_with_new_tags( self): with self.dataset_populator.test_history() as history_id: hda_id = self.dataset_populator.new_dataset(history_id, content="1 2 3")['id'] hda2_id = self.dataset_populator.new_dataset(history_id, content="1 2 3")['id'] update_response = self._update(hda2_id, dict(tags=['existing:tag']), history_id=history_id).json() assert update_response['tags'] == ['existing:tag'] creation_payload = { 'collection_type': 'list', 'history_id': history_id, 'element_identifiers': [{ 'id': hda_id, 'src': 'hda', 'name': 'element_id1', 'tags': ['my_new_tag'] }, { 'id': hda2_id, 'src': 'hda', 'name': 'element_id2', 'tags': ['another_new_tag'] }], 'type': 'dataset_collection', 'copy_elements': True } r = self._post(f"histories/{self.history_id}/contents", creation_payload, json=True).json() assert r['elements'][0]['object'][ 'id'] != hda_id, "HDA has not been copied" assert len(r['elements'][0]['object']['tags']) == 1 assert r['elements'][0]['object']['tags'][0] == 'my_new_tag' assert len(r['elements'][1]['object'] ['tags']) == 2, r['elements'][1]['object']['tags'] original_hda = self.dataset_populator.get_history_dataset_details( history_id=history_id, dataset_id=hda_id) assert len(original_hda['tags']) == 0, original_hda['tags'] def _check_pair_creation(self, endpoint, payload): pre_collection_count = self.__count_contents(type="dataset_collection") pre_dataset_count = self.__count_contents(type="dataset") pre_combined_count = self.__count_contents( type="dataset,dataset_collection") dataset_collection_response = self._post(endpoint, payload, json=True) dataset_collection = self.__check_create_collection_response( dataset_collection_response) post_collection_count = self.__count_contents( type="dataset_collection") post_dataset_count = self.__count_contents(type="dataset") post_combined_count = self.__count_contents( type="dataset,dataset_collection") # Test filtering types with index. assert pre_collection_count == 0 assert post_collection_count == 1 assert post_combined_count == pre_dataset_count + 1 assert post_combined_count == pre_combined_count + 1 assert pre_dataset_count == post_dataset_count # Test show dataset colleciton. collection_url = f"histories/{self.history_id}/contents/dataset_collections/{dataset_collection['id']}" show_response = self._get(collection_url) self._assert_status_code_is(show_response, 200) dataset_collection = show_response.json() self._assert_has_keys(dataset_collection, "url", "name", "deleted") assert not dataset_collection["deleted"] delete_response = self._delete(collection_url) self._assert_status_code_is(delete_response, 200) show_response = self._get(collection_url) dataset_collection = show_response.json() assert dataset_collection["deleted"] @skip_without_tool("collection_creates_list") def test_jobs_summary_simple_hdca(self): create_response = self.dataset_collection_populator.create_list_in_history( self.history_id, contents=["a\nb\nc\nd", "e\nf\ng\nh"]) hdca_id = create_response.json()["id"] run = self.dataset_populator.run_collection_creates_list( self.history_id, hdca_id) collections = run['output_collections'] collection = collections[0] jobs_summary_url = f"histories/{self.history_id}/contents/dataset_collections/{collection['id']}/jobs_summary" jobs_summary_response = self._get(jobs_summary_url) self._assert_status_code_is(jobs_summary_response, 200) jobs_summary = jobs_summary_response.json() self._assert_has_keys(jobs_summary, "populated_state", "states") @skip_without_tool("cat1") def test_jobs_summary_implicit_hdca(self): create_response = self.dataset_collection_populator.create_pair_in_history( self.history_id, contents=["123", "456"]) hdca_id = create_response.json()["id"] inputs = { "input1": { 'batch': True, 'values': [{ 'src': 'hdca', 'id': hdca_id }] }, } run = self.dataset_populator.run_tool("cat1", inputs=inputs, history_id=self.history_id) self.dataset_populator.wait_for_history_jobs(self.history_id) collections = run['implicit_collections'] collection = collections[0] jobs_summary_url = f"histories/{self.history_id}/contents/dataset_collections/{collection['id']}/jobs_summary" jobs_summary_response = self._get(jobs_summary_url) self._assert_status_code_is(jobs_summary_response, 200) jobs_summary = jobs_summary_response.json() self._assert_has_keys(jobs_summary, "populated_state", "states") states = jobs_summary["states"] assert states.get("ok") == 2, states def test_dataset_collection_hide_originals(self): payload = self.dataset_collection_populator.create_pair_payload( self.history_id, type="dataset_collection") payload["hide_source_items"] = True dataset_collection_response = self._post( f"histories/{self.history_id}/contents", payload, json=True) self.__check_create_collection_response(dataset_collection_response) contents_response = self._get(f"histories/{self.history_id}/contents") datasets = [ d for d in contents_response.json() if d["history_content_type"] == "dataset" and d["hid"] in [1, 2] ] # Assert two datasets in source were hidden. assert len(datasets) == 2 assert not datasets[0]["visible"] assert not datasets[1]["visible"] def test_update_dataset_collection(self): hdca = self._create_pair_collection() body = dict(name="newnameforpair") update_response = self._put( f"histories/{self.history_id}/contents/dataset_collections/{hdca['id']}", data=body, json=True) self._assert_status_code_is(update_response, 200) show_response = self.__show(hdca) assert str(show_response.json()["name"]) == "newnameforpair" def test_update_batch_dataset_collection(self): hdca = self._create_pair_collection() body = { "items": [{ "history_content_type": "dataset_collection", "id": hdca["id"] }], "name": "newnameforpair" } update_response = self._put(f"histories/{self.history_id}/contents", data=body, json=True) self._assert_status_code_is(update_response, 200) show_response = self.__show(hdca) assert str(show_response.json()["name"]) == "newnameforpair" def _create_pair_collection(self): payload = self.dataset_collection_populator.create_pair_payload( self.history_id, type="dataset_collection") dataset_collection_response = self._post( f"histories/{self.history_id}/contents", payload, json=True) self._assert_status_code_is(dataset_collection_response, 200) hdca = dataset_collection_response.json() return hdca def test_hdca_copy(self): hdca = self.dataset_collection_populator.create_pair_in_history( self.history_id).json() hdca_id = hdca["id"] second_history_id = self.dataset_populator.new_history() create_data = dict( source='hdca', content=hdca_id, ) assert len( self._get( f"histories/{second_history_id}/contents/dataset_collections"). json()) == 0 create_response = self._post( f"histories/{second_history_id}/contents/dataset_collections", create_data, json=True) self.__check_create_collection_response(create_response) contents = self._get( f"histories/{second_history_id}/contents/dataset_collections" ).json() assert len(contents) == 1 new_forward, _ = self.__get_paired_response_elements(contents[0]) self._assert_has_keys(new_forward, "history_id") assert new_forward["history_id"] == self.history_id def test_hdca_copy_with_new_dbkey(self): hdca = self.dataset_collection_populator.create_pair_in_history( self.history_id).json() hdca_id = hdca["id"] assert hdca["elements"][0]["object"]["metadata_dbkey"] == "?" assert hdca["elements"][0]["object"]["genome_build"] == "?" create_data = {'source': 'hdca', 'content': hdca_id, 'dbkey': 'hg19'} create_response = self._post( f"histories/{self.history_id}/contents/dataset_collections", create_data, json=True) collection = self.__check_create_collection_response(create_response) new_forward = collection['elements'][0]['object'] assert new_forward["metadata_dbkey"] == "hg19" assert new_forward["genome_build"] == "hg19" def test_hdca_copy_and_elements(self): hdca = self.dataset_collection_populator.create_pair_in_history( self.history_id).json() hdca_id = hdca["id"] second_history_id = self.dataset_populator.new_history() create_data = dict( source='hdca', content=hdca_id, copy_elements=True, ) assert len( self._get( f"histories/{second_history_id}/contents/dataset_collections"). json()) == 0 create_response = self._post( f"histories/{second_history_id}/contents/dataset_collections", create_data, json=True) self.__check_create_collection_response(create_response) contents = self._get( f"histories/{second_history_id}/contents/dataset_collections" ).json() assert len(contents) == 1 new_forward, _ = self.__get_paired_response_elements(contents[0]) self._assert_has_keys(new_forward, "history_id") assert new_forward["history_id"] == second_history_id def __get_paired_response_elements(self, contents): hdca = self.__show(contents).json() self._assert_has_keys(hdca, "name", "deleted", "visible", "elements") elements = hdca["elements"] assert len(elements) == 2 element0 = elements[0] element1 = elements[1] self._assert_has_keys(element0, "object") self._assert_has_keys(element1, "object") return element0["object"], element1["object"] def test_hdca_from_library_datasets(self): ld = self.library_populator.new_library_dataset("el1") ldda_id = ld["ldda_id"] element_identifiers = [{"name": "el1", "src": "ldda", "id": ldda_id}] history_id = self.dataset_populator.new_history() create_data = dict( history_id=history_id, type="dataset_collection", name="Test From Library", element_identifiers=element_identifiers, collection_type="list", ) create_response = self._post( f"histories/{history_id}/contents/dataset_collections", create_data, json=True) hdca = self.__check_create_collection_response(create_response) elements = hdca["elements"] assert len(elements) == 1 hda = elements[0]["object"] assert hda["hda_ldda"] == "hda" assert hda["history_content_type"] == "dataset" assert hda["copied_from_ldda_id"] == ldda_id assert hda['history_id'] == history_id def test_hdca_from_inaccessible_library_datasets(self): library, library_dataset = self.library_populator.new_library_dataset_in_private_library( "HDCACreateInaccesibleLibrary") ldda_id = library_dataset["id"] element_identifiers = [{"name": "el1", "src": "ldda", "id": ldda_id}] create_data = dict( history_id=self.history_id, type="dataset_collection", name="Test From Library", element_identifiers=element_identifiers, collection_type="list", ) with self._different_user(): second_history_id = self.dataset_populator.new_history() create_response = self._post( f"histories/{second_history_id}/contents/dataset_collections", create_data, json=True) self._assert_status_code_is(create_response, 403) def __check_create_collection_response(self, response): self._assert_status_code_is(response, 200) dataset_collection = response.json() self._assert_has_keys(dataset_collection, "url", "name", "deleted", "visible", "elements") return dataset_collection def __show(self, contents): show_response = self._get( f"histories/{self.history_id}/contents/{contents['history_content_type']}s/{contents['id']}" ) return show_response def __count_contents(self, history_id=None, **kwds): if history_id is None: history_id = self.history_id contents_response = self._get(f"histories/{history_id}/contents", kwds) return len(contents_response.json()) def __assert_hda_has_full_details(self, hda_details): self._assert_has_keys(hda_details, "display_types", "display_apps") def __check_for_hda(self, contents_response, hda): self._assert_status_code_is(contents_response, 200) contents = contents_response.json() assert len(contents) == 1 hda_summary = contents[0] self.__assert_matches_hda(hda, hda_summary) return hda_summary def __assert_matches_hda(self, input_hda, query_hda): self._assert_has_keys(query_hda, "id", "name") assert input_hda["name"] == query_hda["name"] assert input_hda["id"] == query_hda["id"] def test_job_state_summary_field(self): create_response = self.dataset_collection_populator.create_pair_in_history( self.history_id, contents=["123", "456"]) self._assert_status_code_is(create_response, 200) contents_response = self._get( f"histories/{self.history_id}/contents?v=dev&keys=job_state_summary&view=summary" ) self._assert_status_code_is(contents_response, 200) contents = contents_response.json() for c in filter( lambda c: c['history_content_type'] == 'dataset_collection', contents): assert isinstance(c, dict) assert 'job_state_summary' in c assert isinstance(c['job_state_summary'], dict) def _get_content(self, history_id, update_time): return self._get( f"/api/histories/{history_id}/contents/near/100/100?update_time-gt={update_time}" ).json() def test_history_contents_near_with_update_time(self): with self.dataset_populator.test_history() as history_id: first_time = datetime.utcnow().isoformat() assert len(self._get_content(history_id, update_time=first_time)) == 0 self.dataset_collection_populator.create_list_in_history( history_id=history_id) assert len(self._get_content( history_id, update_time=first_time)) == 4 # 3 datasets self.dataset_populator.wait_for_history(history_id) all_datasets_finished = first_time = datetime.utcnow().isoformat() assert len( self._get_content(history_id, update_time=all_datasets_finished)) == 0 def test_history_contents_near_with_since(self): with self.dataset_populator.test_history() as history_id: original_history = self._get(f"/api/histories/{history_id}").json() original_history_stamp = original_history['update_time'] # check empty contents, with no since flag, should return an empty 200 result history_contents = self._get( f"/api/histories/{history_id}/contents/near/100/100") assert history_contents.status_code == 200 assert len(history_contents.json()) == 0 # adding a since parameter, should return a 204 if history has not changed at all history_contents = self._get( f"/api/histories/{history_id}/contents/near/100/100?since={original_history_stamp}" ) assert history_contents.status_code == 204 # add some stuff self.dataset_collection_populator.create_list_in_history( history_id=history_id) self.dataset_populator.wait_for_history(history_id) # check to make sure the added stuff is there changed_history_contents = self._get( f"/api/histories/{history_id}/contents/near/100/100") assert changed_history_contents.status_code == 200 assert len(changed_history_contents.json()) == 4 # check to make sure the history date has actually changed due to changing the contents changed_history = self._get(f"/api/histories/{history_id}").json() changed_history_stamp = changed_history['update_time'] assert original_history_stamp != changed_history_stamp # a repeated contents request with since=original_history_stamp should now return data # because we have added datasets and the update_time should have been changed changed_content = self._get( f"/api/histories/{history_id}/contents/near/100/100?since={original_history_stamp}" ) assert changed_content.status_code == 200 assert len(changed_content.json()) == 4 def test_history_contents_near_since_with_standard_iso8601_date(self): with self.dataset_populator.test_history() as history_id: original_history = self._get(f"/api/histories/{history_id}").json() original_history_stamp = original_history['update_time'] # this is the standard date format that javascript will emit using .toISOString(), it # should be the expected date format for any modern api # https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Date/toISOString # checking to make sure that the same exact history.update_time returns a "not changed" # result after date parsing valid_iso8601_date = original_history_stamp + 'Z' encoded_valid_date = urllib.parse.quote_plus(valid_iso8601_date) history_contents = self._get( f"/api/histories/{history_id}/contents/near/100/100?since={encoded_valid_date}" ) assert history_contents.status_code == 204 # test parsing for other standard is08601 formats sample_formats = [ '2021-08-26T15:53:02+00:00', '2021-08-26T15:53:02Z', '2002-10-10T12:00:00-05:00' ] for date_str in sample_formats: encoded_date = urllib.parse.quote_plus( date_str) # handles pluses, minuses history_contents = self._get( f"/api/histories/{history_id}/contents/near/100/100?since={encoded_date}" ) self._assert_status_code_is_ok(history_contents) @skip_without_tool('cat_data_and_sleep') def test_history_contents_near_with_update_time_implicit_collection(self): with self.dataset_populator.test_history() as history_id: hdca_id = self.dataset_collection_populator.create_list_in_history( history_id=history_id).json()['id'] self.dataset_populator.wait_for_history(history_id) inputs = { "input1": { 'batch': True, 'values': [{ "src": "hdca", "id": hdca_id }] }, "sleep_time": 2, } response = self.dataset_populator.run_tool( "cat_data_and_sleep", inputs, history_id, ) update_time = datetime.utcnow().isoformat() collection_id = response['implicit_collections'][0]['id'] for _ in range(20): time.sleep(1) update = self._get_content(history_id, update_time=update_time) if any(c for c in update if c['history_content_type'] == 'dataset_collection' and c['job_state_summary']['ok'] == 3): return raise Exception( f"History content update time query did not include final update for implicit collection {collection_id}" ) @skip_without_tool('collection_creates_dynamic_nested') def test_history_contents_near_with_update_time_explicit_collection(self): with self.dataset_populator.test_history() as history_id: inputs = {'foo': 'bar', 'sleep_time': 2} response = self.dataset_populator.run_tool( "collection_creates_dynamic_nested", inputs, history_id, ) update_time = datetime.utcnow().isoformat() collection_id = response['output_collections'][0]['id'] for _ in range(20): time.sleep(1) update = self._get_content(history_id, update_time=update_time) if any(c for c in update if c['history_content_type'] == 'dataset_collection' and c['populated_state'] == 'ok'): return raise Exception( f"History content update time query did not include populated_state update for dynamic nested collection {collection_id}" ) def test_index_filter_by_type(self): history_id = self.dataset_populator.new_history() self.dataset_populator.new_dataset(history_id) self.dataset_collection_populator.create_list_in_history( history_id=history_id) contents_response = self._get( f"histories/{history_id}/contents").json() num_items = len(contents_response) expected_num_collections = 1 expected_num_datasets = num_items - expected_num_collections contents_response = self._get( f"histories/{history_id}/contents?types=dataset").json() assert len(contents_response) == expected_num_datasets contents_response = self._get( f"histories/{history_id}/contents?types=dataset_collection").json( ) assert len(contents_response) == expected_num_collections def test_elements_datatypes_field(self): history_id = self.dataset_populator.new_history() collection_name = "homogeneous" expected_datatypes = ["txt"] elements = [ # List with all elements of txt datatype (homogeneous) { "name": "test1", "src": "pasted", "paste_content": "abc", "ext": "txt" }, { "name": "test2", "src": "pasted", "paste_content": "abc", "ext": "txt" }, ] self._upload_collection_list_with_elements(history_id, collection_name, elements) self._assert_collection_has_expected_elements_datatypes( history_id, collection_name, expected_datatypes) collection_name = "heterogeneous" expected_datatypes = ["txt", "tabular"] elements = [ # List with txt and tabular datatype (heterogeneous) { "name": "test2", "src": "pasted", "paste_content": "abc", "ext": "txt" }, { "name": "test3", "src": "pasted", "paste_content": "a,b,c\n", "ext": "tabular" }, ] self._upload_collection_list_with_elements(history_id, collection_name, elements) self._assert_collection_has_expected_elements_datatypes( history_id, collection_name, expected_datatypes) def _upload_collection_list_with_elements(self, history_id: str, collection_name: str, elements: List[Any]): create_homogeneous_response = self.dataset_collection_populator.upload_collection( history_id, "list", elements=elements, name=collection_name) self._assert_status_code_is_ok(create_homogeneous_response) def _assert_collection_has_expected_elements_datatypes( self, history_id, collection_name, expected_datatypes): contents_response = self._get( f"histories/{history_id}/contents?v=dev&view=betawebclient&q=name-eq&qv={collection_name}" ) self._assert_status_code_is(contents_response, 200) collection = contents_response.json()[0] self.assertCountEqual(collection["elements_datatypes"], expected_datatypes)
class JobsApiTestCase(ApiTestCase, TestsTools): def setUp(self): super().setUp() self.workflow_populator = WorkflowPopulator(self.galaxy_interactor) self.dataset_populator = DatasetPopulator(self.galaxy_interactor) self.dataset_collection_populator = DatasetCollectionPopulator( self.galaxy_interactor) @uses_test_history(require_new=True) def test_index(self, history_id): # Create HDA to ensure at least one job exists... self.__history_with_new_dataset(history_id) jobs = self.__jobs_index() assert "upload1" in map(itemgetter("tool_id"), jobs) @uses_test_history(require_new=True) def test_system_details_admin_only(self, history_id): self.__history_with_new_dataset(history_id) jobs = self.__jobs_index(admin=False) job = jobs[0] self._assert_not_has_keys(job, "external_id") jobs = self.__jobs_index(admin=True) job = jobs[0] self._assert_has_keys(job, "command_line", "external_id") @uses_test_history(require_new=True) def test_admin_job_list(self, history_id): self.__history_with_new_dataset(history_id) jobs_response = self._get("jobs?view=admin_job_list", admin=False) assert jobs_response.status_code == 403 assert jobs_response.json( )['err_msg'] == 'Only admins can use the admin_job_list view' jobs = self._get("jobs?view=admin_job_list", admin=True).json() job = jobs[0] self._assert_has_keys(job, "command_line", "external_id", 'handler') @uses_test_history(require_new=True) def test_index_state_filter(self, history_id): # Initial number of ok jobs original_count = len(self.__uploads_with_state("ok")) # Run through dataset upload to ensure num uplaods at least greater # by 1. self.__history_with_ok_dataset(history_id) # Verify number of ok jobs is actually greater. count_increased = False for _ in range(10): new_count = len(self.__uploads_with_state("ok")) if original_count < new_count: count_increased = True break time.sleep(.1) if not count_increased: template = "Jobs in ok state did not increase (was %d, now %d)" message = template % (original_count, new_count) raise AssertionError(message) @uses_test_history(require_new=True) def test_index_date_filter(self, history_id): self.__history_with_new_dataset(history_id) two_weeks_ago = (datetime.datetime.utcnow() - datetime.timedelta(14)).isoformat() last_week = (datetime.datetime.utcnow() - datetime.timedelta(7)).isoformat() next_week = (datetime.datetime.utcnow() + datetime.timedelta(7)).isoformat() today = datetime.datetime.utcnow().isoformat() tomorrow = (datetime.datetime.utcnow() + datetime.timedelta(1)).isoformat() jobs = self.__jobs_index(data={ "date_range_min": today[0:10], "date_range_max": tomorrow[0:10] }) assert len(jobs) > 0 today_job_id = jobs[0]["id"] jobs = self.__jobs_index(data={ "date_range_min": two_weeks_ago, "date_range_max": last_week }) assert today_job_id not in map(itemgetter("id"), jobs) jobs = self.__jobs_index(data={ "date_range_min": last_week, "date_range_max": next_week }) assert today_job_id in map(itemgetter("id"), jobs) @uses_test_history(require_new=True) def test_index_history(self, history_id): self.__history_with_new_dataset(history_id) jobs = self.__jobs_index(data={"history_id": history_id}) assert len(jobs) > 0 with self.dataset_populator.test_history() as other_history_id: jobs = self.__jobs_index(data={"history_id": other_history_id}) assert len(jobs) == 0 @uses_test_history(require_new=True) def test_index_workflow_and_invocation_filter(self, history_id): workflow_simple = """ class: GalaxyWorkflow name: Simple Workflow inputs: input1: data outputs: wf_output_1: outputSource: first_cat/out_file1 steps: first_cat: tool_id: cat1 in: input1: input1 """ summary = self.workflow_populator.run_workflow( workflow_simple, history_id=history_id, test_data={"input1": "hello world"}) invocation_id = summary.invocation_id workflow_id = self._get( f"invocations/{invocation_id}").json()['workflow_id'] self.workflow_populator.wait_for_invocation(workflow_id, invocation_id) jobs1 = self.__jobs_index(data={"workflow_id": workflow_id}) assert len(jobs1) == 1 jobs2 = self.__jobs_index(data={"invocation_id": invocation_id}) assert len(jobs2) == 1 assert jobs1 == jobs2 @uses_test_history(require_new=True) def test_index_workflow_filter_implicit_jobs(self, history_id): workflow_id = self.workflow_populator.upload_yaml_workflow(""" class: GalaxyWorkflow inputs: input_datasets: collection steps: multi_data_optional: tool_id: multi_data_optional in: input1: input_datasets """) hdca_id = self.dataset_collection_populator.create_list_of_list_in_history( history_id).json() self.dataset_populator.wait_for_history(history_id, assert_ok=True) inputs = { '0': self.dataset_populator.ds_entry(hdca_id), } invocation_id = self.workflow_populator.invoke_workflow_and_wait( workflow_id, history_id=history_id, inputs=inputs, assert_ok=True) jobs1 = self.__jobs_index(data={"workflow_id": workflow_id}) jobs2 = self.__jobs_index(data={"invocation_id": invocation_id}) assert len(jobs1) == len(jobs2) == 1 second_invocation_id = self.workflow_populator.invoke_workflow_and_wait( workflow_id, history_id=history_id, inputs=inputs, assert_ok=True) workflow_jobs = self.__jobs_index(data={"workflow_id": workflow_id}) second_invocation_jobs = self.__jobs_index( data={"invocation_id": second_invocation_id}) assert len(workflow_jobs) == 2 assert len(second_invocation_jobs) == 1 @uses_test_history(require_new=True) def test_index_limit_and_offset_filter(self, history_id): self.__history_with_new_dataset(history_id) jobs = self.__jobs_index(data={"history_id": history_id}) assert len(jobs) > 0 length = len(jobs) jobs = self.__jobs_index(data={"history_id": history_id, "offset": 1}) assert len(jobs) == length - 1 jobs = self.__jobs_index(data={"history_id": history_id, "limit": 0}) assert len(jobs) == 0 @uses_test_history(require_new=True) def test_index_user_filter(self, history_id): test_user_email = "*****@*****.**" user = self._setup_user(test_user_email) with self._different_user(email=test_user_email): # User should be able to jobs for their own ID. jobs = self.__jobs_index(data={"user_id": user["id"]}) assert jobs == [] # Admin should be able to see jobs of another user. jobs = self.__jobs_index(data={"user_id": user["id"]}, admin=True) assert jobs == [] # Normal user should not be able to see jobs of another user. jobs_response = self._get("jobs", data={"user_id": user["id"]}) self._assert_status_code_is(jobs_response, 403) assert jobs_response.json() == { "err_msg": "Only admins can index the jobs of others", "err_code": 403006 } @uses_test_history(require_new=True) def test_index_multiple_states_filter(self, history_id): # Initial number of ok jobs original_count = len(self.__uploads_with_state("ok", "new")) # Run through dataset upload to ensure num uplaods at least greater # by 1. self.__history_with_ok_dataset(history_id) # Verify number of ok jobs is actually greater. new_count = len(self.__uploads_with_state("new", "ok")) assert original_count < new_count, new_count @uses_test_history(require_new=True) def test_show(self, history_id): job_properties_tool_run = self.dataset_populator.run_tool( tool_id="job_properties", inputs={}, history_id=history_id, ) first_job = self.__jobs_index()[0] self._assert_has_key(first_job, 'id', 'state', 'exit_code', 'update_time', 'create_time') job_id = job_properties_tool_run["jobs"][0]["id"] show_jobs_response = self.dataset_populator.get_job_details(job_id) self._assert_status_code_is(show_jobs_response, 200) job_details = show_jobs_response.json() self._assert_has_key(job_details, 'id', 'state', 'exit_code', 'update_time', 'create_time') show_jobs_response = self.dataset_populator.get_job_details(job_id, full=True) self._assert_status_code_is(show_jobs_response, 200) job_details = show_jobs_response.json() self._assert_has_key( job_details, "create_time", "exit_code", "id", "job_messages", "job_stderr", "job_stdout", "state", "stderr", "stdout", "tool_stderr", "tool_stdout", "update_time", ) self.dataset_populator.wait_for_job(job_id, assert_ok=True) show_jobs_response = self.dataset_populator.get_job_details(job_id, full=True) job_details = show_jobs_response.json() assert "The bool is not true\n" not in job_details["job_stdout"] assert "The bool is very not true\n" not in job_details["job_stderr"] assert job_details["tool_stdout"] == "The bool is not true\n" assert job_details["tool_stderr"] == "The bool is very not true\n" assert "The bool is not true\n" in job_details["stdout"] assert "The bool is very not true\n" in job_details["stderr"] @uses_test_history(require_new=True) def test_show_security(self, history_id): self.__history_with_new_dataset(history_id) jobs_response = self._get("jobs", data={"history_id": history_id}) job = jobs_response.json()[0] job_id = job["id"] job_lock_response = self._get("job_lock", admin=True) job_lock_response.raise_for_status() assert not job_lock_response.json()["active"] show_jobs_response = self._get(f"jobs/{job_id}", admin=False) self._assert_not_has_keys(show_jobs_response.json(), "external_id") # TODO: Re-activate test case when API accepts privacy settings # with self._different_user(): # show_jobs_response = self._get( "jobs/%s" % job_id, admin=False ) # self._assert_status_code_is( show_jobs_response, 200 ) show_jobs_response = self._get(f"jobs/{job_id}", admin=True) self._assert_has_keys(show_jobs_response.json(), "command_line", "external_id") def _run_detect_errors(self, history_id, inputs): payload = self.dataset_populator.run_tool_payload( tool_id='detect_errors_aggressive', inputs=inputs, history_id=history_id, ) return self._post("tools", data=payload).json() @skip_without_tool("detect_errors_aggressive") def test_unhide_on_error(self): with self.dataset_populator.test_history() as history_id: inputs = {'error_bool': 'true'} run_response = self._run_detect_errors(history_id=history_id, inputs=inputs) job_id = run_response['jobs'][0]["id"] self.dataset_populator.wait_for_job(job_id) job = self.dataset_populator.get_job_details(job_id).json() assert job['state'] == 'error' dataset = self.dataset_populator.get_history_dataset_details( history_id=history_id, dataset_id=run_response['outputs'][0]['id'], assert_ok=False) assert dataset['visible'] def _run_map_over_error(self, history_id): hdca1 = self.dataset_collection_populator.create_list_in_history( history_id, contents=[("sample1-1", "1 2 3")]).json() inputs = { 'error_bool': 'true', 'dataset': { 'batch': True, 'values': [{ 'src': 'hdca', 'id': hdca1['id'] }], } } return self._run_detect_errors(history_id=history_id, inputs=inputs) @skip_without_tool("detect_errors_aggressive") def test_no_unhide_on_error_if_mapped_over(self): with self.dataset_populator.test_history() as history_id: run_response = self._run_map_over_error(history_id) job_id = run_response['jobs'][0]["id"] self.dataset_populator.wait_for_job(job_id) job = self.dataset_populator.get_job_details(job_id).json() assert job['state'] == 'error' dataset = self.dataset_populator.get_history_dataset_details( history_id=history_id, dataset_id=run_response['outputs'][0]['id'], assert_ok=False) assert not dataset['visible'] def test_no_hide_on_rerun(self): with self.dataset_populator.test_history() as history_id: run_response = self._run_map_over_error(history_id) job_id = run_response['jobs'][0]["id"] self.dataset_populator.wait_for_job(job_id) failed_hdca = self.dataset_populator.get_history_collection_details( history_id=history_id, content_id=run_response['implicit_collections'][0]['id'], assert_ok=False, ) first_update_time = failed_hdca['update_time'] assert failed_hdca['visible'] rerun_params = self._get(f"jobs/{job_id}/build_for_rerun").json() inputs = rerun_params['state_inputs'] inputs['rerun_remap_job_id'] = job_id rerun_response = self._run_detect_errors(history_id=history_id, inputs=inputs) rerun_job_id = rerun_response['jobs'][0]["id"] self.dataset_populator.wait_for_job(rerun_job_id) # Verify source hdca is still visible hdca = self.dataset_populator.get_history_collection_details( history_id=history_id, content_id=run_response['implicit_collections'][0]['id'], assert_ok=False, ) assert hdca['visible'] assert isoparse( hdca['update_time']) > (isoparse(first_update_time)) @skip_without_tool('empty_output') def test_common_problems(self): with self.dataset_populator.test_history() as history_id: empty_run_response = self.dataset_populator.run_tool( tool_id='empty_output', inputs={}, history_id=history_id, ) empty_hda = empty_run_response["outputs"][0] cat_empty_twice_run_response = self.dataset_populator.run_tool( tool_id='cat1', inputs={ 'input1': { 'src': 'hda', 'id': empty_hda['id'] }, 'queries_0|input2': { 'src': 'hda', 'id': empty_hda['id'] } }, history_id=history_id, ) empty_output_job = empty_run_response["jobs"][0] cat_empty_job = cat_empty_twice_run_response["jobs"][0] empty_output_common_problems_response = self._get( f"jobs/{empty_output_job['id']}/common_problems").json() cat_empty_common_problems_response = self._get( f"jobs/{cat_empty_job['id']}/common_problems").json() self._assert_has_keys(empty_output_common_problems_response, "has_empty_inputs", "has_duplicate_inputs") self._assert_has_keys(cat_empty_common_problems_response, "has_empty_inputs", "has_duplicate_inputs") assert not empty_output_common_problems_response["has_empty_inputs"] assert cat_empty_common_problems_response["has_empty_inputs"] assert not empty_output_common_problems_response[ "has_duplicate_inputs"] assert cat_empty_common_problems_response["has_duplicate_inputs"] @skip_without_tool('detect_errors_aggressive') def test_report_error(self): with self.dataset_populator.test_history() as history_id: self._run_error_report(history_id) @skip_without_tool('detect_errors_aggressive') def test_report_error_anon(self): with self._different_user(anon=True): history_id = self._get( urllib.parse.urljoin( self.url, "history/current_history_json")).json()['id'] self._run_error_report(history_id) def _run_error_report(self, history_id): payload = self.dataset_populator.run_tool_payload( tool_id='detect_errors_aggressive', inputs={'error_bool': 'true'}, history_id=history_id, ) run_response = self._post("tools", data=payload).json() job_id = run_response['jobs'][0]["id"] self.dataset_populator.wait_for_job(job_id) dataset_id = run_response['outputs'][0]['id'] response = self._post(f'jobs/{job_id}/error', data={'dataset_id': dataset_id}) assert response.status_code == 200, response.text @skip_without_tool('detect_errors_aggressive') def test_report_error_bootstrap_admin(self): with self.dataset_populator.test_history() as history_id: payload = self.dataset_populator.run_tool_payload( tool_id='detect_errors_aggressive', inputs={'error_bool': 'true'}, history_id=history_id, ) run_response = self._post("tools", data=payload, key=self.master_api_key) self._assert_status_code_is(run_response, 400) @skip_without_tool("create_2") @uses_test_history(require_new=True) def test_deleting_output_keep_running_until_all_deleted(self, history_id): job_state, outputs = self._setup_running_two_output_job( history_id, 120) self._hack_to_skip_test_if_state_ok(job_state) # Delete one of the two outputs and make sure the job is still running. self._raw_update_history_item(history_id, outputs[0]["id"], {"deleted": True}) self._hack_to_skip_test_if_state_ok(job_state) time.sleep(1) self._hack_to_skip_test_if_state_ok(job_state) state = job_state().json()["state"] assert state == "running", state # Delete the second output and make sure the job is cancelled. self._raw_update_history_item(history_id, outputs[1]["id"], {"deleted": True}) final_state = wait_on_state(job_state, assert_ok=False, timeout=15) assert final_state in ["deleting", "deleted"], final_state @skip_without_tool("create_2") @uses_test_history(require_new=True) def test_purging_output_keep_running_until_all_purged(self, history_id): job_state, outputs = self._setup_running_two_output_job( history_id, 120) # Pretty much right away after the job is running, these paths should be populated - # if they are grab them and make sure they are deleted at the end of the job. dataset_1 = self._get_history_item_as_admin(history_id, outputs[0]["id"]) dataset_2 = self._get_history_item_as_admin(history_id, outputs[1]["id"]) if "file_name" in dataset_1: output_dataset_paths = [ dataset_1["file_name"], dataset_2["file_name"] ] # This may or may not exist depending on if the test is local or not. output_dataset_paths_exist = os.path.exists( output_dataset_paths[0]) else: output_dataset_paths = [] output_dataset_paths_exist = False self._hack_to_skip_test_if_state_ok(job_state) current_state = job_state().json()["state"] assert current_state == "running", current_state # Purge one of the two outputs and make sure the job is still running. self._raw_update_history_item(history_id, outputs[0]["id"], {"purged": True}) time.sleep(1) self._hack_to_skip_test_if_state_ok(job_state) current_state = job_state().json()["state"] assert current_state == "running", current_state # Purge the second output and make sure the job is cancelled. self._raw_update_history_item(history_id, outputs[1]["id"], {"purged": True}) final_state = wait_on_state(job_state, assert_ok=False, timeout=15) assert final_state in ["deleting", "deleted"], final_state def paths_deleted(): if not os.path.exists( output_dataset_paths[0]) and not os.path.exists( output_dataset_paths[1]): return True if output_dataset_paths_exist: wait_on(paths_deleted, "path deletion") @skip_without_tool("create_2") @uses_test_history(require_new=True) def test_purging_output_cleaned_after_ok_run(self, history_id): job_state, outputs = self._setup_running_two_output_job(history_id, 10) # Pretty much right away after the job is running, these paths should be populated - # if they are grab them and make sure they are deleted at the end of the job. dataset_1 = self._get_history_item_as_admin(history_id, outputs[0]["id"]) dataset_2 = self._get_history_item_as_admin(history_id, outputs[1]["id"]) if "file_name" in dataset_1: output_dataset_paths = [ dataset_1["file_name"], dataset_2["file_name"] ] # This may or may not exist depending on if the test is local or not. output_dataset_paths_exist = os.path.exists( output_dataset_paths[0]) else: output_dataset_paths = [] output_dataset_paths_exist = False if not output_dataset_paths_exist: # Given this Galaxy configuration - there is nothing more to be tested here. # Consider throwing a skip instead. return # Purge one of the two outputs and wait for the job to complete. self._raw_update_history_item(history_id, outputs[0]["id"], {"purged": True}) wait_on_state(job_state, assert_ok=True) if output_dataset_paths_exist: time.sleep(.5) # Make sure the non-purged dataset is on disk and the purged one is not. assert os.path.exists(output_dataset_paths[1]) assert not os.path.exists(output_dataset_paths[0]) def _hack_to_skip_test_if_state_ok(self, job_state): from nose.plugins.skip import SkipTest if job_state().json()["state"] == "ok": message = "Job state switch from running to ok too quickly - the rest of the test requires the job to be in a running state. Skipping test." raise SkipTest(message) def _setup_running_two_output_job(self, history_id, sleep_time): payload = self.dataset_populator.run_tool_payload( tool_id='create_2', inputs=dict(sleep_time=sleep_time, ), history_id=history_id, ) run_response = self._post("tools", data=payload) run_response.raise_for_status() run_object = run_response.json() outputs = run_object["outputs"] jobs = run_object["jobs"] assert len(outputs) == 2 assert len(jobs) == 1 def job_state(): jobs_response = self._get(f"jobs/{jobs[0]['id']}") return jobs_response # Give job some time to get up and running. time.sleep(2) running_state = wait_on_state(job_state, skip_states=["queued", "new"], assert_ok=False, timeout=15) assert running_state == "running", running_state return job_state, outputs def _raw_update_history_item(self, history_id, item_id, data): update_url = self._api_url( f"histories/{history_id}/contents/{item_id}", use_key=True) update_response = requests.put(update_url, json=data) assert_status_code_is_ok(update_response) return update_response @skip_without_tool("cat_data_and_sleep") @uses_test_history(require_new=True) def test_resume_job(self, history_id): hda1 = self.dataset_populator.new_dataset( history_id, content="samp1\t10.0\nsamp2\t20.0\n") hda2 = self.dataset_populator.new_dataset( history_id, content="samp1\t30.0\nsamp2\t40.0\n") # Submit first job payload = self.dataset_populator.run_tool_payload( tool_id='cat_data_and_sleep', inputs={ 'sleep_time': 15, 'input1': { 'src': 'hda', 'id': hda2['id'] }, 'queries_0|input2': { 'src': 'hda', 'id': hda2['id'] } }, history_id=history_id, ) run_response = self._post("tools", data=payload).json() output = run_response["outputs"][0] # Submit second job that waits on job1 payload = self.dataset_populator.run_tool_payload( tool_id='cat1', inputs={ 'input1': { 'src': 'hda', 'id': hda1['id'] }, 'queries_0|input2': { 'src': 'hda', 'id': output['id'] } }, history_id=history_id, ) run_response = self._post("tools", data=payload).json() job_id = run_response['jobs'][0]['id'] output = run_response["outputs"][0] # Delete second jobs input while second job is waiting for first job delete_response = self._delete( f"histories/{history_id}/contents/{hda1['id']}") self._assert_status_code_is(delete_response, 200) self.dataset_populator.wait_for_history_jobs(history_id, assert_ok=False) dataset_details = self._get( f"histories/{history_id}/contents/{output['id']}").json() assert dataset_details['state'] == 'paused' # Undelete input dataset undelete_response = self._put( f"histories/{history_id}/contents/{hda1['id']}", data={'deleted': False}, json=True) self._assert_status_code_is(undelete_response, 200) resume_response = self._put(f"jobs/{job_id}/resume") self._assert_status_code_is(resume_response, 200) self.dataset_populator.wait_for_history_jobs(history_id, assert_ok=True) dataset_details = self._get( f"histories/{history_id}/contents/{output['id']}").json() assert dataset_details['state'] == 'ok' def _get_history_item_as_admin(self, history_id, item_id): response = self._get( f"histories/{history_id}/contents/{item_id}?view=detailed", admin=True) assert_status_code_is_ok(response) return response.json() @uses_test_history(require_new=True) def test_search(self, history_id): dataset_id = self.__history_with_ok_dataset(history_id) # We first copy the datasets, so that the update time is lower than the job creation time new_history_id = self.dataset_populator.new_history() copy_payload = { "content": dataset_id, "source": "hda", "type": "dataset" } copy_response = self._post(f"histories/{new_history_id}/contents", data=copy_payload, json=True) self._assert_status_code_is(copy_response, 200) inputs = json.dumps({'input1': {'src': 'hda', 'id': dataset_id}}) self._job_search(tool_id='cat1', history_id=history_id, inputs=inputs) # We test that a job can be found even if the dataset has been copied to another history new_dataset_id = copy_response.json()['id'] copied_inputs = json.dumps( {'input1': { 'src': 'hda', 'id': new_dataset_id }}) search_payload = self._search_payload(history_id=history_id, tool_id='cat1', inputs=copied_inputs) self._search(search_payload, expected_search_count=1) # Now we delete the original input HDA that was used -- we should still be able to find the job delete_respone = self._delete( f"histories/{history_id}/contents/{dataset_id}") self._assert_status_code_is(delete_respone, 200) self._search(search_payload, expected_search_count=1) # Now we also delete the copy -- we shouldn't find a job delete_respone = self._delete( f"histories/{new_history_id}/contents/{new_dataset_id}") self._assert_status_code_is(delete_respone, 200) self._search(search_payload, expected_search_count=0) @uses_test_history(require_new=True) def test_search_handle_identifiers(self, history_id): # Test that input name and element identifier of a jobs' output must match for a job to be returned. dataset_id = self.__history_with_ok_dataset(history_id) inputs = json.dumps({'input1': {'src': 'hda', 'id': dataset_id}}) self._job_search(tool_id='identifier_single', history_id=history_id, inputs=inputs) dataset_details = self._get( f"histories/{history_id}/contents/{dataset_id}").json() dataset_details['name'] = 'Renamed Test Dataset' dataset_update_response = self._put( f"histories/{history_id}/contents/{dataset_id}", data=dict(name='Renamed Test Dataset'), json=True) self._assert_status_code_is(dataset_update_response, 200) assert dataset_update_response.json()['name'] == 'Renamed Test Dataset' search_payload = self._search_payload(history_id=history_id, tool_id='identifier_single', inputs=inputs) self._search(search_payload, expected_search_count=0) @uses_test_history(require_new=True) def test_search_delete_outputs(self, history_id): dataset_id = self.__history_with_ok_dataset(history_id) inputs = json.dumps({'input1': {'src': 'hda', 'id': dataset_id}}) tool_response = self._job_search(tool_id='cat1', history_id=history_id, inputs=inputs) output_id = tool_response.json()['outputs'][0]['id'] delete_respone = self._delete( f"histories/{history_id}/contents/{output_id}") self._assert_status_code_is(delete_respone, 200) search_payload = self._search_payload(history_id=history_id, tool_id='cat1', inputs=inputs) self._search(search_payload, expected_search_count=0) @uses_test_history(require_new=True) def test_search_with_hdca_list_input(self, history_id): list_id_a = self.__history_with_ok_collection(collection_type='list', history_id=history_id) list_id_b = self.__history_with_ok_collection(collection_type='list', history_id=history_id) inputs = json.dumps({ 'f1': { 'src': 'hdca', 'id': list_id_a }, 'f2': { 'src': 'hdca', 'id': list_id_b }, }) tool_response = self._job_search(tool_id='multi_data_param', history_id=history_id, inputs=inputs) # We switch the inputs, this should not return a match inputs_switched = json.dumps({ 'f2': { 'src': 'hdca', 'id': list_id_a }, 'f1': { 'src': 'hdca', 'id': list_id_b }, }) search_payload = self._search_payload(history_id=history_id, tool_id='multi_data_param', inputs=inputs_switched) self._search(search_payload, expected_search_count=0) # We delete the ouput (this is a HDA, as multi_data_param reduces collections) # and use the correct input job definition, the job should not be found output_id = tool_response.json()['outputs'][0]['id'] delete_respone = self._delete( f"histories/{history_id}/contents/{output_id}") self._assert_status_code_is(delete_respone, 200) search_payload = self._search_payload(history_id=history_id, tool_id='multi_data_param', inputs=inputs) self._search(search_payload, expected_search_count=0) @uses_test_history(require_new=True) def test_search_delete_hdca_output(self, history_id): list_id_a = self.__history_with_ok_collection(collection_type='list', history_id=history_id) inputs = json.dumps({ 'input1': { 'src': 'hdca', 'id': list_id_a }, }) tool_response = self._job_search(tool_id='collection_creates_list', history_id=history_id, inputs=inputs) output_id = tool_response.json()['outputs'][0]['id'] # We delete a single tool output, no job should be returned delete_respone = self._delete( f"histories/{history_id}/contents/{output_id}") self._assert_status_code_is(delete_respone, 200) search_payload = self._search_payload( history_id=history_id, tool_id='collection_creates_list', inputs=inputs) self._search(search_payload, expected_search_count=0) tool_response = self._job_search(tool_id='collection_creates_list', history_id=history_id, inputs=inputs) output_collection_id = tool_response.json( )['output_collections'][0]['id'] # We delete a collection output, no job should be returned delete_respone = self._delete( f"histories/{history_id}/contents/dataset_collections/{output_collection_id}" ) self._assert_status_code_is(delete_respone, 200) search_payload = self._search_payload( history_id=history_id, tool_id='collection_creates_list', inputs=inputs) self._search(search_payload, expected_search_count=0) @uses_test_history(require_new=True) def test_search_with_hdca_pair_input(self, history_id): list_id_a = self.__history_with_ok_collection(collection_type='pair', history_id=history_id) inputs = json.dumps({ 'f1': { 'src': 'hdca', 'id': list_id_a }, 'f2': { 'src': 'hdca', 'id': list_id_a }, }) self._job_search(tool_id='multi_data_param', history_id=history_id, inputs=inputs) # We test that a job can be found even if the collection has been copied to another history new_history_id = self.dataset_populator.new_history() copy_payload = { "content": list_id_a, "source": "hdca", "type": "dataset_collection" } copy_response = self._post(f"histories/{new_history_id}/contents", data=copy_payload, json=True) self._assert_status_code_is(copy_response, 200) new_list_a = copy_response.json()['id'] copied_inputs = json.dumps({ 'f1': { 'src': 'hdca', 'id': new_list_a }, 'f2': { 'src': 'hdca', 'id': new_list_a }, }) search_payload = self._search_payload(history_id=new_history_id, tool_id='multi_data_param', inputs=copied_inputs) self._search(search_payload, expected_search_count=1) # Now we delete the original input HDCA that was used -- we should still be able to find the job delete_respone = self._delete( f"histories/{history_id}/contents/dataset_collections/{list_id_a}") self._assert_status_code_is(delete_respone, 200) self._search(search_payload, expected_search_count=1) # Now we also delete the copy -- we shouldn't find a job delete_respone = self._delete( f"histories/{history_id}/contents/dataset_collections/{new_list_a}" ) self._assert_status_code_is(delete_respone, 200) self._search(search_payload, expected_search_count=0) @uses_test_history(require_new=True) def test_search_with_hdca_list_pair_input(self, history_id): list_id_a = self.__history_with_ok_collection( collection_type='list:pair', history_id=history_id) inputs = json.dumps({ 'f1': { 'src': 'hdca', 'id': list_id_a }, 'f2': { 'src': 'hdca', 'id': list_id_a }, }) self._job_search(tool_id='multi_data_param', history_id=history_id, inputs=inputs) @uses_test_history(require_new=True) def test_search_with_hdca_list_pair_collection_mapped_over_pair_input( self, history_id): list_id_a = self.__history_with_ok_collection( collection_type='list:pair', history_id=history_id) inputs = json.dumps({ 'f1': { 'batch': True, 'values': [{ 'src': 'hdca', 'id': list_id_a, 'map_over_type': 'paired' }] }, }) self._job_search(tool_id='collection_paired_test', history_id=history_id, inputs=inputs) def _get_simple_rerun_params(self, history_id, private=False): list_id_a = self.__history_with_ok_collection( collection_type='list:pair', history_id=history_id) inputs = { 'f1': { 'batch': True, 'values': [{ 'src': 'hdca', 'id': list_id_a, 'map_over_type': 'paired' }] } } run_response = self._run( history_id=history_id, tool_id="collection_paired_test", inputs=inputs, wait_for_job=True, assert_ok=True, ) rerun_params = self._get( f"jobs/{run_response['jobs'][0]['id']}/build_for_rerun").json() # Since we call rerun on the first (and only) job we should get the expanded input # which is a dataset collection element (and not the list:pair hdca that was used as input to the original # job). assert rerun_params['state_inputs']['f1']['values'][0]['src'] == 'dce' if private: hdca = self.dataset_populator.get_history_collection_details( history_id=history_id, content_id=list_id_a) for element in hdca['elements'][0]['object']['elements']: self.dataset_populator.make_private(history_id, element['object']['id']) return rerun_params @skip_without_tool("collection_paired_test") @uses_test_history(require_new=False) def test_job_build_for_rerun(self, history_id): rerun_params = self._get_simple_rerun_params(history_id) self._run( history_id=history_id, tool_id="collection_paired_test", inputs=rerun_params['state_inputs'], wait_for_job=True, assert_ok=True, ) @skip_without_tool("collection_paired_test") @uses_test_history(require_new=False) def test_dce_submission_security(self, history_id): rerun_params = self._get_simple_rerun_params(history_id, private=True) with self._different_user(): other_history_id = self.dataset_populator.new_history() response = self._run( history_id=other_history_id, tool_id="collection_paired_test", inputs=rerun_params['state_inputs'], wait_for_job=False, assert_ok=False, ) assert response.status_code == 403 @skip_without_tool("identifier_collection") @uses_test_history(require_new=False) def test_job_build_for_rerun_list_list(self, history_id): list_id_a = self.__history_with_ok_collection(collection_type='list', history_id=history_id) list_id_b = self.__history_with_ok_collection(collection_type='list', history_id=history_id) list_list = self.dataset_collection_populator.create_nested_collection( history_id=history_id, collection_type='list:list', name='list list collection', collection=[list_id_a, list_id_b]).json() list_list_id = list_list['id'] first_element = list_list['elements'][0] assert first_element['element_type'] == 'dataset_collection' assert first_element['element_identifier'] == 'test0' assert first_element['model_class'] == 'DatasetCollectionElement' inputs = { 'input1': { 'batch': True, 'values': [{ 'src': 'hdca', 'id': list_list_id, 'map_over_type': 'list' }] } } run_response = self._run( history_id=history_id, tool_id="identifier_collection", inputs=inputs, wait_for_job=True, assert_ok=True, ) assert len(run_response['jobs']) == 2 rerun_params = self._get( f"jobs/{run_response['jobs'][0]['id']}/build_for_rerun").json() # Since we call rerun on the first (and only) job we should get the expanded input # which is a dataset collection element (and not the list:list hdca that was used as input to the original # job). assert rerun_params['state_inputs']['input1']['values'][0][ 'src'] == 'dce' rerun_response = self._run( history_id=history_id, tool_id="identifier_collection", inputs=rerun_params['state_inputs'], wait_for_job=True, assert_ok=True, ) assert len(rerun_response['jobs']) == 1 rerun_content = self.dataset_populator.get_history_dataset_content( history_id=history_id, dataset=rerun_response['outputs'][0]) run_content = self.dataset_populator.get_history_dataset_content( history_id=history_id, dataset=run_response['outputs'][0]) assert rerun_content == run_content def _job_search(self, tool_id, history_id, inputs): search_payload = self._search_payload(history_id=history_id, tool_id=tool_id, inputs=inputs) empty_search_response = self._post("jobs/search", data=search_payload) self._assert_status_code_is(empty_search_response, 200) self.assertEqual(len(empty_search_response.json()), 0) tool_response = self._post("tools", data=search_payload) self.dataset_populator.wait_for_tool_run(history_id, run_response=tool_response) self._search(search_payload, expected_search_count=1) return tool_response def _search_payload(self, history_id, tool_id, inputs, state='ok'): search_payload = dict(tool_id=tool_id, inputs=inputs, history_id=history_id, state=state) return search_payload def _search(self, payload, expected_search_count=1): # in case job and history aren't updated at exactly the same # time give time to wait for _ in range(5): search_count = self._search_count(payload) if search_count == expected_search_count: break time.sleep(1) assert search_count == expected_search_count, "expected to find %d jobs, got %d jobs" % ( expected_search_count, search_count) return search_count def _search_count(self, search_payload): search_response = self._post("jobs/search", data=search_payload) self._assert_status_code_is(search_response, 200) search_json = search_response.json() return len(search_json) def __uploads_with_state(self, *states): jobs_response = self._get("jobs", data=dict(state=states)) self._assert_status_code_is(jobs_response, 200) jobs = jobs_response.json() assert not [j for j in jobs if not j['state'] in states] return [j for j in jobs if j['tool_id'] == 'upload1'] def __history_with_new_dataset(self, history_id): dataset_id = self.dataset_populator.new_dataset(history_id)["id"] return dataset_id def __history_with_ok_dataset(self, history_id): dataset_id = self.dataset_populator.new_dataset(history_id, wait=True)["id"] return dataset_id def __history_with_ok_collection(self, collection_type='list', history_id=None): if not history_id: history_id = self.dataset_populator.new_history() if collection_type == 'list': fetch_response = self.dataset_collection_populator.create_list_in_history( history_id, direct_upload=True).json() elif collection_type == 'pair': fetch_response = self.dataset_collection_populator.create_pair_in_history( history_id, direct_upload=True).json() elif collection_type == 'list:pair': fetch_response = self.dataset_collection_populator.create_list_of_pairs_in_history( history_id).json() self.dataset_collection_populator.wait_for_fetched_collection( fetch_response) return fetch_response["outputs"][0]['id'] def __jobs_index(self, **kwds): jobs_response = self._get("jobs", **kwds) self._assert_status_code_is(jobs_response, 200) jobs = jobs_response.json() assert isinstance(jobs, list) return jobs
class DatasetCollectionApiTestCase(ApiTestCase): def setUp(self): super(DatasetCollectionApiTestCase, self).setUp() self.dataset_populator = DatasetPopulator(self.galaxy_interactor) self.dataset_collection_populator = DatasetCollectionPopulator( self.galaxy_interactor) self.history_id = self.dataset_populator.new_history() def test_create_pair_from_history(self): payload = self.dataset_collection_populator.create_pair_payload( self.history_id, instance_type="history", ) create_response = self._post("dataset_collections", payload) dataset_collection = self._check_create_response(create_response) returned_datasets = dataset_collection["elements"] assert len(returned_datasets) == 2, dataset_collection def test_create_list_from_history(self): element_identifiers = self.dataset_collection_populator.list_identifiers( self.history_id) payload = dict( instance_type="history", history_id=self.history_id, element_identifiers=json.dumps(element_identifiers), collection_type="list", ) create_response = self._post("dataset_collections", payload) dataset_collection = self._check_create_response(create_response) returned_datasets = dataset_collection["elements"] assert len(returned_datasets) == 3, dataset_collection def test_create_list_of_existing_pairs(self): pair_payload = self.dataset_collection_populator.create_pair_payload( self.history_id, instance_type="history", ) pair_create_response = self._post("dataset_collections", pair_payload) dataset_collection = self._check_create_response(pair_create_response) hdca_id = dataset_collection["id"] element_identifiers = [dict(name="test1", src="hdca", id=hdca_id)] payload = dict( instance_type="history", history_id=self.history_id, element_identifiers=json.dumps(element_identifiers), collection_type="list", ) create_response = self._post("dataset_collections", payload) dataset_collection = self._check_create_response(create_response) returned_collections = dataset_collection["elements"] assert len(returned_collections) == 1, dataset_collection def test_create_list_of_new_pairs(self): identifiers = self.dataset_collection_populator.nested_collection_identifiers( self.history_id, "list:paired") payload = dict( collection_type="list:paired", instance_type="history", history_id=self.history_id, name="a nested collection", element_identifiers=json.dumps(identifiers), ) create_response = self._post("dataset_collections", payload) dataset_collection = self._check_create_response(create_response) assert dataset_collection["collection_type"] == "list:paired" assert dataset_collection["name"] == "a nested collection" returned_collections = dataset_collection["elements"] assert len(returned_collections) == 1, dataset_collection pair_1_element = returned_collections[0] self._assert_has_keys(pair_1_element, "element_identifier", "element_index", "object") assert pair_1_element[ "element_identifier"] == "test_level_1", pair_1_element assert pair_1_element["element_index"] == 0, pair_1_element pair_1_object = pair_1_element["object"] self._assert_has_keys(pair_1_object, "collection_type", "elements", "element_count") self.assertEqual(pair_1_object["collection_type"], "paired") self.assertEqual(pair_1_object["populated"], True) pair_elements = pair_1_object["elements"] assert len(pair_elements) == 2 pair_1_element_1 = pair_elements[0] assert pair_1_element_1["element_index"] == 0 def test_list_download(self): fetch_response = self.dataset_collection_populator.create_list_in_history( self.history_id, direct_upload=True).json() dataset_collection = self.dataset_collection_populator.wait_for_fetched_collection( fetch_response) returned_dce = dataset_collection["elements"] assert len(returned_dce) == 3, dataset_collection create_response = self._download_dataset_collection( history_id=self.history_id, hdca_id=dataset_collection['id']) self._assert_status_code_is(create_response, 200) tar_contents = tarfile.open(fileobj=BytesIO(create_response.content)) namelist = tar_contents.getnames() assert len(namelist) == 3, "Expected 3 elements in [%s]" % namelist collection_name = dataset_collection['name'] for element, zip_path in zip(returned_dce, namelist): assert "%s/%s.%s" % (collection_name, element['element_identifier'], element['object']['file_ext']) == zip_path def test_pair_download(self): fetch_response = self.dataset_collection_populator.create_pair_in_history( self.history_id, direct_upload=True).json() dataset_collection = self.dataset_collection_populator.wait_for_fetched_collection( fetch_response) returned_dce = dataset_collection["elements"] assert len(returned_dce) == 2, dataset_collection hdca_id = dataset_collection['id'] create_response = self._download_dataset_collection( history_id=self.history_id, hdca_id=hdca_id) self._assert_status_code_is(create_response, 200) tar_contents = tarfile.open(fileobj=BytesIO(create_response.content)) namelist = tar_contents.getnames() assert len(namelist) == 2, "Expected 2 elements in [%s]" % namelist collection_name = dataset_collection['name'] for element, zip_path in zip(returned_dce, namelist): assert "%s/%s.%s" % (collection_name, element['element_identifier'], element['object']['file_ext']) == zip_path def test_list_pair_download(self): fetch_response = self.dataset_collection_populator.create_list_of_pairs_in_history( self.history_id).json() dataset_collection = self.dataset_collection_populator.wait_for_fetched_collection( fetch_response) returned_dce = dataset_collection["elements"] assert len(returned_dce) == 1, dataset_collection list_collection_name = dataset_collection['name'] pair = returned_dce[0] create_response = self._download_dataset_collection( history_id=self.history_id, hdca_id=dataset_collection['id']) self._assert_status_code_is(create_response, 200) tar_contents = tarfile.open(fileobj=BytesIO(create_response.content)) namelist = tar_contents.getnames() assert len(namelist) == 2, "Expected 2 elements in [%s]" % namelist pair_collection_name = pair['element_identifier'] for element, zip_path in zip(pair['object']['elements'], namelist): assert "%s/%s/%s.%s" % (list_collection_name, pair_collection_name, element['element_identifier'], element['object']['file_ext']) == zip_path def test_list_list_download(self): dataset_collection = self.dataset_collection_populator.create_list_of_list_in_history( self.history_id).json() self.dataset_collection_populator.wait_for_dataset_collection( dataset_collection, assert_ok=True) returned_dce = dataset_collection["elements"] assert len(returned_dce) == 1, dataset_collection create_response = self._download_dataset_collection( history_id=self.history_id, hdca_id=dataset_collection['id']) self._assert_status_code_is(create_response, 200) tar_contents = tarfile.open(fileobj=BytesIO(create_response.content)) namelist = tar_contents.getnames() assert len(namelist) == 3, "Expected 3 elements in [%s]" % namelist def test_list_list_list_download(self): dataset_collection = self.dataset_collection_populator.create_list_of_list_in_history( self.history_id, collection_type='list:list:list').json() self.dataset_collection_populator.wait_for_dataset_collection( dataset_collection, assert_ok=True) returned_dce = dataset_collection["elements"] assert len(returned_dce) == 1, dataset_collection create_response = self._download_dataset_collection( history_id=self.history_id, hdca_id=dataset_collection['id']) self._assert_status_code_is(create_response, 200) tar_contents = tarfile.open(fileobj=BytesIO(create_response.content)) namelist = tar_contents.getnames() assert len(namelist) == 3, "Expected 3 elements in [%s]" % namelist def test_hda_security(self): element_identifiers = self.dataset_collection_populator.pair_identifiers( self.history_id) self.dataset_populator.make_private(self.history_id, element_identifiers[0]["id"]) with self._different_user(): history_id = self.dataset_populator.new_history() payload = dict( instance_type="history", history_id=history_id, element_identifiers=json.dumps(element_identifiers), collection_type="paired", ) create_response = self._post("dataset_collections", payload) self._assert_status_code_is(create_response, 403) def test_enforces_unique_names(self): element_identifiers = self.dataset_collection_populator.list_identifiers( self.history_id) element_identifiers[2]["name"] = element_identifiers[0]["name"] payload = dict( instance_type="history", history_id=self.history_id, element_identifiers=json.dumps(element_identifiers), collection_type="list", ) create_response = self._post("dataset_collections", payload) self._assert_status_code_is(create_response, 400) def test_upload_collection(self): elements = [{ "src": "files", "dbkey": "hg19", "info": "my cool bed", "tags": ["name:data1", "group:condition:treated", "machine:illumina"] }] targets = [{ "destination": { "type": "hdca" }, "elements": elements, "collection_type": "list", "name": "Test upload", "tags": ["name:collection1"] }] payload = { "history_id": self.history_id, "targets": json.dumps(targets), "__files": { "files_0|file_data": open(self.test_data_resolver.get_filename("4.bed")) }, } self.dataset_populator.fetch(payload) hdca = self._assert_one_collection_created_in_history() self.assertEqual(hdca["name"], "Test upload") hdca_tags = hdca["tags"] assert len(hdca_tags) == 1 assert "name:collection1" in hdca_tags assert len(hdca["elements"]) == 1, hdca element0 = hdca["elements"][0] assert element0["element_identifier"] == "4.bed" dataset0 = element0["object"] assert dataset0["file_size"] == 61 dataset_tags = dataset0["tags"] assert len(dataset_tags) == 3, dataset0 def test_upload_nested(self): elements = [{ "name": "samp1", "elements": [{ "src": "files", "dbkey": "hg19", "info": "my cool bed" }] }] targets = [{ "destination": { "type": "hdca" }, "elements": elements, "collection_type": "list:list", "name": "Test upload", }] payload = { "history_id": self.history_id, "targets": json.dumps(targets), "__files": { "files_0|file_data": open(self.test_data_resolver.get_filename("4.bed")) }, } self.dataset_populator.fetch(payload) hdca = self._assert_one_collection_created_in_history() self.assertEqual(hdca["name"], "Test upload") assert len(hdca["elements"]) == 1, hdca element0 = hdca["elements"][0] assert element0["element_identifier"] == "samp1" @skip_if_github_down def test_upload_collection_from_url(self): elements = [{ "src": "url", "url": "https://raw.githubusercontent.com/galaxyproject/galaxy/dev/test-data/4.bed", "info": "my cool bed" }] targets = [{ "destination": { "type": "hdca" }, "elements": elements, "collection_type": "list", }] payload = { "history_id": self.history_id, "targets": json.dumps(targets), "__files": { "files_0|file_data": open(self.test_data_resolver.get_filename("4.bed")) }, } self.dataset_populator.fetch(payload) hdca = self._assert_one_collection_created_in_history() assert len(hdca["elements"]) == 1, hdca element0 = hdca["elements"][0] assert element0["element_identifier"] == "4.bed" assert element0["object"]["file_size"] == 61 def _assert_one_collection_created_in_history(self): contents_response = self._get( "histories/%s/contents/dataset_collections" % self.history_id) self._assert_status_code_is(contents_response, 200) contents = contents_response.json() assert len(contents) == 1 hdca = contents[0] assert hdca["history_content_type"] == "dataset_collection" hdca_id = hdca["id"] collection_response = self._get( "histories/%s/contents/dataset_collections/%s" % (self.history_id, hdca_id)) self._assert_status_code_is(collection_response, 200) return collection_response.json() def _check_create_response(self, create_response): self._assert_status_code_is(create_response, 200) dataset_collection = create_response.json() self._assert_has_keys(dataset_collection, "elements", "url", "name", "collection_type", "element_count") return dataset_collection def _download_dataset_collection(self, history_id, hdca_id): return self._get( "histories/%s/contents/dataset_collections/%s/download" % (history_id, hdca_id))