From d7a587c1336a0e25102245add8a596f20fe88f4d Mon Sep 17 00:00:00 2001
From: Jeff Hull <jeff@solvebio.com>
Date: Fri, 24 Apr 2020 11:06:27 -0400
Subject: [PATCH 1/7] get or create for an Object (also closes #326)

---
 solvebio/resource/dataset.py |  83 +-----------------------
 solvebio/resource/object.py  | 120 ++++++++++++++++++++++++++++++++---
 solvebio/resource/vault.py   |  11 ++--
 3 files changed, 120 insertions(+), 94 deletions(-)

diff --git a/solvebio/resource/dataset.py b/solvebio/resource/dataset.py
index b3636cbf..e2af79e7 100644
--- a/solvebio/resource/dataset.py
+++ b/solvebio/resource/dataset.py
@@ -3,7 +3,6 @@
 
 from ..client import client
 from ..query import Query
-from ..errors import NotFoundError
 
 from .solveobject import convert_to_solve_object
 from .apiresource import CreateableAPIResource
@@ -67,86 +66,10 @@ def get_by_full_path(cls, full_path, **kwargs):
 
     @classmethod
     def get_or_create_by_full_path(cls, full_path, **kwargs):
-        from solvebio import Vault
         from solvebio import Object
-
-        _client = kwargs.pop('client', None) or cls._client or client
-        create_vault = kwargs.pop('create_vault', False)
-        create_folders = kwargs.pop('create_folders', True)
-
-        try:
-            return Dataset.get_by_full_path(full_path, assert_type='dataset',
-                                            client=_client)
-        except NotFoundError:
-            pass
-
-        # Dataset not found, create it step-by-step
-        full_path, parts = Object.validate_full_path(full_path, client=_client)
-
-        if create_vault:
-            vault = Vault.get_or_create_by_full_path(
-                '{0}:{1}'.format(parts['domain'], parts['vault']),
-                client=_client)
-        else:
-            vaults = Vault.all(account_domain=parts['domain'],
-                               name=parts['vault'],
-                               client=_client)
-            if len(vaults.solve_objects()) == 0:
-                raise Exception(
-                    'Vault does not exist with name {0}:{1}'.format(
-                        parts['domain'], parts['vault'])
-                )
-            vault = vaults.solve_objects()[0]
-
-        # Create the folders to hold the dataset if they do not already exist.
-        object_path = parts['path']
-        curr_path = os.path.dirname(object_path)
-        folders_to_create = []
-        new_folders = []
-        id_map = {'/': None}
-
-        while curr_path != '/':
-            try:
-                obj = Object.get_by_path(curr_path,
-                                         vault_id=vault.id,
-                                         assert_type='folder',
-                                         client=_client)
-                id_map[curr_path] = obj.id
-                break
-            except NotFoundError:
-                if not create_folders:
-                    raise Exception('Folder {} does not exist.  Pass '
-                                    'create_folders=True to auto-create '
-                                    'missing folders')
-
-                folders_to_create.append(curr_path)
-                curr_path = '/'.join(curr_path.split('/')[:-1])
-                if curr_path == '':
-                    break
-
-        for folder in reversed(folders_to_create):
-            new_folder = Object.create(
-                object_type='folder',
-                vault_id=vault.id,
-                filename=os.path.basename(folder),
-                parent_object_id=id_map[os.path.dirname(folder)],
-                client=_client
-            )
-            new_folders.append(new_folder)
-            id_map[folder] = new_folder.id
-
-        if os.path.dirname(object_path) == '/':
-            parent_folder_id = None
-        elif new_folders:
-            parent_folder_id = new_folders[-1].id
-        else:
-            parent_folder_id = id_map[os.path.dirname(object_path)]
-
-        return Dataset.create(name=os.path.basename(object_path),
-                              vault_id=vault.id,
-                              vault_parent_object_id=parent_folder_id,
-                              client=_client,
-                              **kwargs)
+        # Assert this is a dataset
+        kwargs['assert_type'] = 'dataset'
+        return Object.get_or_create_by_full_path(full_path, **kwargs)
 
     def saved_queries(self, **params):
         from solvebio import SavedQuery
diff --git a/solvebio/resource/object.py b/solvebio/resource/object.py
index 2030fde8..b565233f 100644
--- a/solvebio/resource/object.py
+++ b/solvebio/resource/object.py
@@ -145,6 +145,18 @@ def validate_full_path(cls, full_path, **kwargs):
 
         return path_dict['full_path'], path_dict
 
+    @classmethod
+    def get_by_path(cls, path, **params):
+        assert_type = params.pop('assert_type', None)
+        params.update({'path': path})
+        obj = cls._retrieve_helper('object', 'path', path, **params)
+        if obj and assert_type and obj['object_type'] != assert_type:
+            raise SolveError(
+                "Expected a {} but found a {} at {}"
+                .format(assert_type, obj['object_type'], path))
+
+        return obj
+
     @classmethod
     def get_by_full_path(cls, full_path, **params):
         # Don't pop client from params since **params is used below
@@ -162,16 +174,106 @@ def get_by_full_path(cls, full_path, **params):
         return obj
 
     @classmethod
-    def get_by_path(cls, path, **params):
-        assert_type = params.pop('assert_type', None)
-        params.update({'path': path})
-        obj = cls._retrieve_helper('object', 'path', path, **params)
-        if obj and assert_type and obj['object_type'] != assert_type:
-            raise SolveError(
-                "Expected a {} but found a {} at {}"
-                .format(assert_type, obj['object_type'], path))
+    def get_or_create_by_full_path(cls, full_path, **kwargs):
+        from solvebio import Vault
+        from solvebio import Object
 
-        return obj
+        _client = kwargs.pop('client', None) or cls._client or client
+        create_vault = kwargs.pop('create_vault', False)
+        create_folders = kwargs.pop('create_folders', True)
+
+        # Check for object type assertion, if not explicitly added, see
+        # if user has passed object_type, as their intent was to get/create
+        # an object of that type.
+        assert_type = kwargs.pop('assert_type',
+                                 kwargs.get('object_type', None))
+
+        try:
+            return cls.get_by_full_path(
+                full_path, assert_type=assert_type, client=_client)
+        except NotFoundError:
+            pass
+
+        # Object type required when creating Object
+        object_type = kwargs.get('object_type')
+        if not object_type:
+            raise Exception("'object_type' is required when creating a new "
+                            "Object. Pass one of: file, folder, dataset")
+
+        # TODO should we require file contents?
+        # Technically a user could then use this object to the call
+        # upload_file()
+        # if object_type == 'file' and not kwargs.get('content'):
+        #     raise Exception('')
+
+        # Object not found, create it step-by-step
+        full_path, parts = Object.validate_full_path(full_path, client=_client)
+
+        if create_vault:
+            vault = Vault.get_or_create_by_full_path(
+                '{0}:{1}'.format(parts['domain'], parts['vault']),
+                client=_client)
+        else:
+            vaults = Vault.all(account_domain=parts['domain'],
+                               name=parts['vault'],
+                               client=_client)
+            if len(vaults.solve_objects()) == 0:
+                raise Exception(
+                    'Vault with name {0}:{1} does not exist. Pass '
+                    'create_vault=True to auto-create'.format(
+                        parts['domain'], parts['vault'])
+                )
+            vault = vaults.solve_objects()[0]
+
+        # Create the folders to hold the object if they do not already exist.
+        object_path = parts['path']
+        curr_path = os.path.dirname(object_path)
+        folders_to_create = []
+        new_folders = []
+        id_map = {'/': None}
+
+        while curr_path != '/':
+            try:
+                obj = Object.get_by_path(curr_path,
+                                         vault_id=vault.id,
+                                         assert_type='folder',
+                                         client=_client)
+                id_map[curr_path] = obj.id
+                break
+            except NotFoundError:
+                if not create_folders:
+                    raise Exception('Folder {} does not exist.  Pass '
+                                    'create_folders=True to auto-create '
+                                    'missing folders')
+
+                folders_to_create.append(curr_path)
+                curr_path = '/'.join(curr_path.split('/')[:-1])
+                if curr_path == '':
+                    break
+
+        for folder in reversed(folders_to_create):
+            new_folder = Object.create(
+                object_type='folder',
+                vault_id=vault.id,
+                filename=os.path.basename(folder),
+                parent_object_id=id_map[os.path.dirname(folder)],
+                client=_client
+            )
+            new_folders.append(new_folder)
+            id_map[folder] = new_folder.id
+
+        if os.path.dirname(object_path) == '/':
+            parent_folder_id = None
+        elif new_folders:
+            parent_folder_id = new_folders[-1].id
+        else:
+            parent_folder_id = id_map[os.path.dirname(object_path)]
+
+        return Object.create(filename=os.path.basename(object_path),
+                             vault_id=vault.id,
+                             parent_object_id=parent_folder_id,
+                             client=_client,
+                             **kwargs)
 
     @classmethod
     def upload_file(cls, local_path, remote_path, vault_full_path, **kwargs):
diff --git a/solvebio/resource/vault.py b/solvebio/resource/vault.py
index ae0b7613..a834efb2 100644
--- a/solvebio/resource/vault.py
+++ b/solvebio/resource/vault.py
@@ -138,19 +138,20 @@ def _get_parent_folder(self, path):
         )
 
     def create_dataset(self, name, **params):
-        from solvebio import Dataset
+        from solvebio import Object
 
         params['vault_id'] = self.id
+        params['object_type'] = 'dataset'
         path = params.pop('path', None)
 
         if path == '/' or path is None:
-            params['vault_parent_object_id'] = None
+            params['parent_object_id'] = None
         else:
             parent_object = self._get_parent_folder(path)
-            params['vault_parent_object_id'] = parent_object.id
+            params['parent_object_id'] = parent_object.id
 
-        params['name'] = name
-        return Dataset.create(**params)
+        params['filename'] = name
+        return Object.create(**params)
 
     def create_folder(self, filename, **params):
         from solvebio import Object

From a42c8aa8387ea9e0e3387be7381692d3607c326d Mon Sep 17 00:00:00 2001
From: Jeff Hull <jeff@solvebio.com>
Date: Fri, 24 Apr 2020 12:58:13 -0400
Subject: [PATCH 2/7] add Object getattr for #287

---
 solvebio/cli/data.py            | 23 +++++++++++++----------
 solvebio/resource/dataset.py    |  5 +++++
 solvebio/resource/object.py     | 30 ++++++++++++++++++++----------
 solvebio/test/client_mocks.py   | 10 +++++++++-
 solvebio/test/test_shortcuts.py | 33 ++++++++++++++++-----------------
 5 files changed, 63 insertions(+), 38 deletions(-)

diff --git a/solvebio/cli/data.py b/solvebio/cli/data.py
index 9694acc4..88e1d3dd 100644
--- a/solvebio/cli/data.py
+++ b/solvebio/cli/data.py
@@ -14,6 +14,9 @@
 
 from solvebio import Vault
 from solvebio import Object
+from solvebio import Dataset
+from solvebio import DatasetImport
+from solvebio import DatasetTemplate
 from solvebio.utils.files import check_gzip_path
 from solvebio.errors import SolveError
 from solvebio.errors import NotFoundError
@@ -157,11 +160,11 @@ def _create_template_from_file(template_file, dry_run=False):
             sys.exit(1)
 
     if dry_run:
-        template = solvebio.DatasetTemplate(**template_json)
+        template = DatasetTemplate(**template_json)
         print("A new dataset template will be created from: {0}"
               .format(template_file))
     else:
-        template = solvebio.DatasetTemplate.create(**template_json)
+        template = DatasetTemplate.create(**template_json)
         print("A new dataset template was created with id: {0}"
               .format(template.id))
 
@@ -189,7 +192,7 @@ def create_dataset(args, template=None):
 
     try:
         # Fail if a dataset already exists.
-        solvebio.Dataset.get_by_full_path(full_path)
+        Object.get_by_full_path(full_path, assert_type='dataset')
         print('A dataset already exists at path: {0}'.format(full_path))
         sys.exit(1)
     except NotFoundError:
@@ -202,8 +205,8 @@ def create_dataset(args, template=None):
         pass
     elif args.template_id:
         try:
-            template = solvebio.DatasetTemplate.retrieve(args.template_id)
-        except solvebio.SolveError as e:
+            template = DatasetTemplate.retrieve(args.template_id)
+        except SolveError as e:
             if e.status_code != 404:
                 raise e
             print("No template with ID {0} found!".format(args.template_id))
@@ -256,7 +259,7 @@ def create_dataset(args, template=None):
             print("Metadata: {}".format(metadata))
         return
 
-    return solvebio.Dataset.get_or_create_by_full_path(
+    return Dataset.get_or_create_by_full_path(
         full_path,
         capacity=args.capacity,
         fields=fields,
@@ -394,8 +397,8 @@ def import_file(args):
 
     if args.template_id:
         try:
-            template = solvebio.DatasetTemplate.retrieve(args.template_id)
-        except solvebio.SolveError as e:
+            template = DatasetTemplate.retrieve(args.template_id)
+        except SolveError as e:
             if e.status_code != 404:
                 raise e
             print("No template with ID {0} found!".format(args.template_id))
@@ -410,7 +413,7 @@ def import_file(args):
         dataset = create_dataset(args, template=template)
     else:
         try:
-            dataset = solvebio.Dataset.get_by_full_path(full_path)
+            dataset = Object.get_by_full_path(full_path, assert_type='dataset')
         except solvebio.errors.NotFoundError:
             print("Dataset not found: {0}".format(full_path))
             print("Tip: use the --create-dataset flag "
@@ -442,7 +445,7 @@ def import_file(args):
             kwargs.update(template.import_params)
 
         # Create the import
-        import_ = solvebio.DatasetImport.create(
+        import_ = DatasetImport.create(
             dataset_id=dataset.id,
             commit_mode=args.commit_mode,
             **kwargs
diff --git a/solvebio/resource/dataset.py b/solvebio/resource/dataset.py
index e2af79e7..0d43e01b 100644
--- a/solvebio/resource/dataset.py
+++ b/solvebio/resource/dataset.py
@@ -38,6 +38,10 @@ class Dataset(CreateableAPIResource,
     def make_full_path(cls, vault_name, path, name, **kwargs):
         from solvebio import SolveError
 
+        print('[Deprecated] Warning this method has been deprecated'
+              'and will be removed in a future released.'
+              'Use self.vault_object.full_path')
+
         _client = kwargs.pop('client', None) or cls._client or client
 
         try:
@@ -69,6 +73,7 @@ def get_or_create_by_full_path(cls, full_path, **kwargs):
         from solvebio import Object
         # Assert this is a dataset
         kwargs['assert_type'] = 'dataset'
+        kwargs['object_type'] = 'dataset'
         return Object.get_or_create_by_full_path(full_path, **kwargs)
 
     def saved_queries(self, **params):
diff --git a/solvebio/resource/object.py b/solvebio/resource/object.py
index b565233f..749121b3 100644
--- a/solvebio/resource/object.py
+++ b/solvebio/resource/object.py
@@ -418,17 +418,27 @@ def objects(self, **params):
     def ls(self, **params):
         return self.objects(**params)
 
-    def query(self, query=None, **params):
-        """Shortcut to query the underlying Dataset object"""
-        from solvebio import Dataset
+    def __getattr__(self, name):
+        """Shortcut to access attributes of the underlying Dataset resource"""
+        from solvebio.resource import Dataset
+
+        valid_attrs = [
+            # query data
+            'query', 'lookup',
+            # transform data
+            'import_file', 'export', 'migrate',
+            # dataset meta
+            'fields', 'template', 'imports', 'commits',
+            # helpers
+            'activity', 'saved_queries'
+        ]
+        if name in valid_attrs and self.is_dataset:
+            return getattr(Dataset(self.dataset_id, client=self._client), name)
 
-        if not self.is_dataset:
-            raise SolveError(
-                "The query method can only be used by a dataset. Found a {}"
-                .format(self.object_type))
-
-        return Dataset(self.dataset_id, client=self._client).query(
-            query=query, **params)
+        try:
+            return self[name]
+        except KeyError as err:
+            raise AttributeError(*err.args)
 
     @property
     def parent(self):
diff --git a/solvebio/test/client_mocks.py b/solvebio/test/client_mocks.py
index 6476c869..f47fdf3c 100644
--- a/solvebio/test/client_mocks.py
+++ b/solvebio/test/client_mocks.py
@@ -41,7 +41,7 @@ def solve_objects(self):
         return ExtendedList([self.create()])
 
     def update_paths(self):
-        """Sets vault_name, path and full_path"""
+        """Sets vault_name, path and full_path based on mock inputs"""
         if not self.object['vault_name']:
             self.object['vault_name'] = 'mock_vault'
 
@@ -55,6 +55,11 @@ def update_paths(self):
             self.object['full_path'] = 'solvebio:{0}:{1}'.format(
                 self.object['vault_name'], self.object['path'])
 
+    def update_dataset(self):
+        """Sets dataset_id if object is a dataset"""
+        if self.object['object_type'] == 'dataset':
+            self.object['dataset_id'] = self.object['id']
+
 
 class FakeMigrationResponse(Fake201Response):
 
@@ -122,6 +127,7 @@ def __init__(self, data):
         }
         self.object.update(data)
         self.update_paths()
+        self.update_dataset()
 
 
 class FakeDatasetResponse(Fake201Response):
@@ -220,6 +226,8 @@ def fake_object_retrieve(*args, **kwargs):
 
 
 def fake_dataset_create(*args, **kwargs):
+    # Dataset.create() is deprecated
+    print("WARNING: Your code likely wants to be using FakeObjectCreate.")
     return FakeDatasetResponse(kwargs).create()
 
 
diff --git a/solvebio/test/test_shortcuts.py b/solvebio/test/test_shortcuts.py
index 38ac5b61..b83c01b1 100644
--- a/solvebio/test/test_shortcuts.py
+++ b/solvebio/test/test_shortcuts.py
@@ -20,7 +20,6 @@
 from solvebio.test.client_mocks import fake_object_all
 from solvebio.test.client_mocks import fake_object_create
 from solvebio.test.client_mocks import fake_object_retrieve
-from solvebio.test.client_mocks import fake_dataset_create
 from solvebio.test.client_mocks import fake_dataset_tmpl_create
 from solvebio.test.client_mocks import fake_dataset_tmpl_retrieve
 from solvebio.test.client_mocks import fake_dataset_import_create
@@ -43,23 +42,23 @@ def setUp(self):
 
     @mock.patch('solvebio.resource.Vault.all')
     @mock.patch('solvebio.resource.Object.all')
-    @mock.patch('solvebio.resource.Dataset.create')
+    @mock.patch('solvebio.resource.Object.create')
     def test_create_dataset(self, DatasetCreate, ObjectAll, VaultAll):
-        DatasetCreate.side_effect = fake_dataset_create
+        DatasetCreate.side_effect = fake_object_create
         ObjectAll.side_effect = fake_object_all
         VaultAll.side_effect = fake_vault_all
         args = ['create-dataset', 'solvebio:test_vault:/test-dataset',
                 '--capacity', 'small']
         ds = main.main(args)
-        self.assertEqual(ds.name, 'test-dataset')
+        self.assertEqual(ds.filename, 'test-dataset')
         self.assertEqual(ds.path, '/test-dataset')
 
     @mock.patch('solvebio.resource.Vault.all')
     @mock.patch('solvebio.resource.Object.all')
-    @mock.patch('solvebio.resource.Dataset.create')
+    @mock.patch('solvebio.resource.Object.create')
     def test_create_dataset_by_filename(self, DatasetCreate, ObjectAll,
                                         VaultAll):
-        DatasetCreate.side_effect = fake_dataset_create
+        DatasetCreate.side_effect = fake_object_create
         ObjectAll.side_effect = fake_object_all
         VaultAll.side_effect = fake_vault_all
         args = [
@@ -71,7 +70,7 @@ def test_create_dataset_by_filename(self, DatasetCreate, ObjectAll,
             '--metadata', 'TEST2=tag2',
         ]
         ds = main.main(args)
-        self.assertEqual(ds.name, 'test-dataset-filename')
+        self.assertEqual(ds.filename, 'test-dataset-filename')
         self.assertEqual(ds.path, '/test-dataset-filename')
         self.assertEqual(ds.capacity, 'small')
         self.assertEqual(ds.tags, ['tag_test'])
@@ -90,13 +89,13 @@ def _validate_tmpl_fields(self, fields):
 
     @mock.patch('solvebio.resource.Vault.all')
     @mock.patch('solvebio.resource.Object.all')
-    @mock.patch('solvebio.resource.Dataset.create')
+    @mock.patch('solvebio.resource.Object.create')
     @mock.patch('solvebio.resource.DatasetTemplate.create')
     def test_create_dataset_upload_template(self, TmplCreate,
                                             DatasetCreate, ObjectAll,
                                             VaultAll):
         TmplCreate.side_effect = fake_dataset_tmpl_create
-        DatasetCreate.side_effect = fake_dataset_create
+        DatasetCreate.side_effect = fake_object_create
         ObjectAll.side_effect = fake_object_all
         VaultAll.side_effect = fake_vault_all
 
@@ -112,14 +111,14 @@ def test_create_dataset_upload_template(self, TmplCreate,
 
     @mock.patch('solvebio.resource.Vault.all')
     @mock.patch('solvebio.resource.Object.all')
-    @mock.patch('solvebio.resource.Dataset.create')
+    @mock.patch('solvebio.resource.Object.create')
     @mock.patch('solvebio.resource.DatasetTemplate.retrieve')
     @mock.patch('solvebio.resource.DatasetTemplate.create')
     def test_create_dataset_template_id(self, TmplCreate, TmplRetrieve,
                                         DatasetCreate, ObjectAll, VaultAll):
         VaultAll.side_effect = fake_vault_all
         ObjectAll.side_effect = fake_object_all
-        DatasetCreate.side_effect = fake_dataset_create
+        DatasetCreate.side_effect = fake_object_create
         TmplRetrieve.side_effect = fake_dataset_tmpl_retrieve
         TmplCreate.side_effect = fake_dataset_tmpl_create
 
@@ -145,15 +144,15 @@ def test_create_dataset_template_id(self, TmplCreate, TmplRetrieve,
     @mock.patch('solvebio.resource.Vault.all')
     @mock.patch('solvebio.resource.Object.all')
     @mock.patch('solvebio.resource.Object.upload_file')
-    @mock.patch('solvebio.resource.Dataset.create')
+    @mock.patch('solvebio.resource.Object.create')
     @mock.patch('solvebio.resource.DatasetImport.create')
-    def _test_import_file(self, args, DatasetImportCreate, DatasetCreate,
-                          ObjectCreate, ObjectAll, VaultAll, VaultLookup,
+    def _test_import_file(self, args, DatasetImportCreate, ObjectCreate,
+                          UploadFile, ObjectAll, VaultAll, VaultLookup,
                           UploadPath, TmplCreate, TmplRetrieve):
         DatasetImportCreate.side_effect = fake_dataset_import_create
-        DatasetCreate.side_effect = fake_dataset_create
-        ObjectAll.side_effect = fake_object_all
         ObjectCreate.side_effect = fake_object_create
+        UploadFile.side_effect = fake_object_create
+        ObjectAll.side_effect = fake_object_all
         VaultAll.side_effect = fake_vault_all
         UploadPath.side_effect = upload_path
         VaultLookup.side_effect = fake_vault_create
@@ -199,7 +198,7 @@ def test_import_tilde(self):
         ]:
             args = ['import', '--create-dataset', dataset_path, file_]
             imports, ds = self._test_import_file(args)
-            self.assertEqual(ds.name, 'test-dataset')
+            self.assertEqual(ds.filename, 'test-dataset')
             # should be a manifest with a single file
             self.assertEqual(len(imports[0].manifest['files']), 1)
 

From 5d873718e8e1d7636d794e6438d3864c46d7cbb6 Mon Sep 17 00:00:00 2001
From: Jeff Hull <jeff@solvebio.com>
Date: Fri, 24 Apr 2020 13:42:29 -0400
Subject: [PATCH 3/7] handle other cases

---
 solvebio/resource/dataset.py | 20 ++++++++------------
 solvebio/resource/object.py  | 22 ++++++++++++++++++----
 solvebio/test/test_object.py | 35 +++++++++++++++++++++++++++++++++++
 3 files changed, 61 insertions(+), 16 deletions(-)

diff --git a/solvebio/resource/dataset.py b/solvebio/resource/dataset.py
index 0d43e01b..e5476659 100644
--- a/solvebio/resource/dataset.py
+++ b/solvebio/resource/dataset.py
@@ -94,9 +94,8 @@ def saved_queries(self, **params):
 
     def fields(self, name=None, **params):
         if 'fields_url' not in self:
-            raise Exception(
-                'Please use Dataset.retrieve({ID}) before looking '
-                'up fields')
+            # Dataset object may not have been retrieved. Grab it.
+            self.refresh()
 
         if name:
             params.update({
@@ -118,18 +117,16 @@ def fields(self, name=None, **params):
 
     def template(self, **params):
         if 'template_url' not in self:
-            raise Exception(
-                'Please use Dataset.retrieve({ID}) before retrieving '
-                'a template')
+            # Dataset object may not have been retrieved. Grab it.
+            self.refresh()
 
         response = self._client.get(self.template_url, params)
         return convert_to_solve_object(response, client=self._client)
 
     def commits(self, **params):
         if 'commits_url' not in self:
-            raise Exception(
-                'Please use Dataset.retrieve({ID}) before looking '
-                'up commits')
+            # Dataset object may not have been retrieved. Grab it.
+            self.refresh()
 
         response = self._client.get(self.commits_url, params)
         results = convert_to_solve_object(response, client=self._client)
@@ -142,9 +139,8 @@ def commits(self, **params):
 
     def imports(self, **params):
         if 'imports_url' not in self:
-            raise Exception(
-                'Please use Dataset.retrieve({ID}) before looking '
-                'up imports')
+            # Dataset object may not have been retrieved. Grab it.
+            self.refresh()
 
         response = self._client.get(self.imports_url, params)
         results = convert_to_solve_object(response, client=self._client)
diff --git a/solvebio/resource/object.py b/solvebio/resource/object.py
index 1e9c65fe..3b290d19 100644
--- a/solvebio/resource/object.py
+++ b/solvebio/resource/object.py
@@ -422,9 +422,10 @@ def __getattr__(self, name):
         """Shortcut to access attributes of the underlying Dataset resource"""
         from solvebio.resource import Dataset
 
-        valid_attrs = [
+        # Valid override attributes that let Object act like a Dataset
+        valid_dataset_attrs = [
             # query data
-            'query', 'lookup',
+            'query', 'lookup', 'beacon',
             # transform data
             'import_file', 'export', 'migrate',
             # dataset meta
@@ -432,14 +433,27 @@ def __getattr__(self, name):
             # helpers
             'activity', 'saved_queries'
         ]
-        if name in valid_attrs and self.is_dataset:
-            return getattr(Dataset(self.dataset_id, client=self._client), name)
 
         try:
             return self[name]
         except KeyError as err:
+            if name in valid_dataset_attrs and self.is_dataset:
+                return getattr(
+                    Dataset(self.dataset_id, client=self._client), name)
+
             raise AttributeError(*err.args)
 
+    @property
+    def dataset(self):
+        """ Returns the dataset object """
+        from solvebio import Dataset
+        if not self.is_dataset:
+            raise SolveError(
+                "Only dataset objects have a Dataset resource. This is a {}"
+                .format(self.object_type))
+
+        return Dataset.retrieve(self['dataset_id'], client=self._client)
+
     @property
     def parent(self):
         """ Returns the parent object """
diff --git a/solvebio/test/test_object.py b/solvebio/test/test_object.py
index da0fbcb1..8deed432 100644
--- a/solvebio/test/test_object.py
+++ b/solvebio/test/test_object.py
@@ -3,6 +3,7 @@
 
 from .helper import SolveBioTestCase
 from solvebio.test.client_mocks import fake_object_create
+from solvebio.test.client_mocks import fake_dataset_create
 
 
 class ObjectTests(SolveBioTestCase):
@@ -131,3 +132,37 @@ def test_object_has_tag(self, ObjectMock):
         obj = self.client.Object.create(name='blah_untagged')
         self.assertEqual(obj.tags, [])
         self.assertFalse(obj.has_tag("foo"))
+
+    @mock.patch('solvebio.resource.Dataset.create')
+    @mock.patch('solvebio.resource.Object.create')
+    def test_object_dataset_getattr(self, ObjectCreate, DatasetCreate):
+        ObjectCreate.side_effect = fake_object_create
+        DatasetCreate.side_effect = fake_dataset_create
+
+        valid_attrs = [
+            'query', 'lookup', 'beacon',
+            'import_file', 'export', 'migrate',
+            'fields', 'template', 'imports', 'commits',
+            'activity', 'saved_queries'
+        ]
+
+        ds = self.client.Dataset.create(name='foo')
+        ds_obj = self.client.Object.create(name='foo_dataset',
+                                           object_type='dataset')
+        file_ = self.client.Object.create(name='foo_file',
+                                          object_type='file')
+        folder_ = self.client.Object.create(name='foo_folder',
+                                            object_type='folder')
+        for attr in valid_attrs:
+            self.assertTrue(getattr(ds, attr))
+            self.assertTrue(getattr(ds_obj, attr))
+            with self.assertRaises(AttributeError):
+                getattr(file_, attr)
+            with self.assertRaises(AttributeError):
+                getattr(folder_, attr)
+
+        # Test that any old attr doesnt work
+        fake_attr = 'foobar'
+        for obj in [file_, folder_, ds, ds_obj]:
+            with self.assertRaises(AttributeError):
+                getattr(obj, fake_attr)

From 63554cc0b87396e5a99986cb72c58880c0884718 Mon Sep 17 00:00:00 2001
From: Jeff Hull <jeff@solvebio.com>
Date: Fri, 24 Apr 2020 14:39:36 -0400
Subject: [PATCH 4/7] update

---
 solvebio/resource/dataset.py | 4 ----
 solvebio/resource/object.py  | 2 +-
 2 files changed, 1 insertion(+), 5 deletions(-)

diff --git a/solvebio/resource/dataset.py b/solvebio/resource/dataset.py
index e5476659..27ec7c1f 100644
--- a/solvebio/resource/dataset.py
+++ b/solvebio/resource/dataset.py
@@ -38,10 +38,6 @@ class Dataset(CreateableAPIResource,
     def make_full_path(cls, vault_name, path, name, **kwargs):
         from solvebio import SolveError
 
-        print('[Deprecated] Warning this method has been deprecated'
-              'and will be removed in a future released.'
-              'Use self.vault_object.full_path')
-
         _client = kwargs.pop('client', None) or cls._client or client
 
         try:
diff --git a/solvebio/resource/object.py b/solvebio/resource/object.py
index 3b290d19..2a6e770d 100644
--- a/solvebio/resource/object.py
+++ b/solvebio/resource/object.py
@@ -452,7 +452,7 @@ def dataset(self):
                 "Only dataset objects have a Dataset resource. This is a {}"
                 .format(self.object_type))
 
-        return Dataset.retrieve(self['dataset_id'], client=self._client)
+        return Dataset.retrieve(self.dataset_id, client=self._client)
 
     @property
     def parent(self):

From 1b9ff5e17ba513cdfbd18cf8fd3fb588ed3717e1 Mon Sep 17 00:00:00 2001
From: Jeff Hull <jeff@solvebio.com>
Date: Fri, 24 Apr 2020 14:46:31 -0400
Subject: [PATCH 5/7] update examples

---
 examples/import_data.py | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/examples/import_data.py b/examples/import_data.py
index a1a983d5..0d848895 100644
--- a/examples/import_data.py
+++ b/examples/import_data.py
@@ -27,7 +27,7 @@
 
 # Prints updates as the data is processed
 # and indexed into SolveBio
-imp.follow()
+dataset.activity(follow=True)
 
 #
 # You now have data!
@@ -49,5 +49,4 @@
     dataset_id=dataset.id,
     data_records=new_records
 )
-
-imp.follow()
+dataset.activity(follow=True)

From c15262e94a95a26d958b2ecd9fae3794209bd410 Mon Sep 17 00:00:00 2001
From: Jeff Hull <jeff@solvebio.com>
Date: Fri, 24 Apr 2020 14:46:44 -0400
Subject: [PATCH 6/7] add a note

---
 solvebio/errors.py | 18 ++++++++++++++++++
 1 file changed, 18 insertions(+)

diff --git a/solvebio/errors.py b/solvebio/errors.py
index 5f9f8fc9..f61493ee 100644
--- a/solvebio/errors.py
+++ b/solvebio/errors.py
@@ -59,5 +59,23 @@ def __init__(self, message=None, response=None):
                 if self.json_body:
                     self.message += ' %s' % self.json_body
 
+            # TODO
+            # NOTE there are other keys that exist in some Errors that
+            # are not detail or non_field_errors. For instance 'manifest'
+            # is a key if uploading using a manifest with invalid file format.
+            # It includes a very useful error message that gets lost.
+            # Is there harm in just handling any error key here?
+            # (handle keys with list values and those without)
+            # Implementation below
+            #
+            # for k, v in self.json_body:
+            #     if isinstance(v, list):
+            #         self.message += ' %s Errors: %s' % \
+            #             (k, ', '.join(self.json_body[k]))
+            #     else:
+            #         self.message += ' %s Errors: %s' % \
+            #             (k, self.json_body[k])
+            #     del self.json_body[k]
+
     def __str__(self):
         return self.message

From fa89f682c69a2c5566029802f3e6b29f29ecf941 Mon Sep 17 00:00:00 2001
From: Jeff Hull <jeff@solvebio.com>
Date: Fri, 24 Apr 2020 15:15:02 -0400
Subject: [PATCH 7/7] use Object lookup

---
 examples/import_data.py           |  2 +-
 solvebio/test/test_beacon.py      |  4 ++--
 solvebio/test/test_dataset.py     | 21 ++++-----------------
 solvebio/test/test_exports.py     |  2 +-
 solvebio/test/test_lookup.py      |  2 +-
 solvebio/test/test_query.py       |  2 +-
 solvebio/test/test_query_batch.py |  2 +-
 7 files changed, 11 insertions(+), 24 deletions(-)

diff --git a/examples/import_data.py b/examples/import_data.py
index 0d848895..74f0ae88 100644
--- a/examples/import_data.py
+++ b/examples/import_data.py
@@ -11,7 +11,7 @@
 dataset_name = 'SampleDataset'
 
 # Create a dataset
-dataset = solvebio.Dataset.get_or_create_by_full_path(
+dataset = solvebio.Object.get_or_create_by_full_path(
     '{0}:/{1}/{2}'.format(vault.name, path, dataset_name),
 )
 
diff --git a/solvebio/test/test_beacon.py b/solvebio/test/test_beacon.py
index 184894a8..fe2e37cd 100644
--- a/solvebio/test/test_beacon.py
+++ b/solvebio/test/test_beacon.py
@@ -12,7 +12,7 @@ def test_beacon_request(self):
         Check that current Clinvar/Variants returns correct
         fields for beacon
         """
-        dataset = self.client.Dataset.get_by_full_path(
+        dataset = self.client.Object.get_by_full_path(
             self.TEST_DATASET_FULL_PATH)
         beacon = dataset.beacon(chromosome='6',
                                      coordinate=51612854,  # staging
@@ -26,7 +26,7 @@ def test_beacon_request(self):
         # Check that Clinvar/Variants version 3.7.0-2015-12-06
         # returns true for specific case
 
-        dataset = self.client.Dataset.get_by_full_path(
+        dataset = self.client.Object.get_by_full_path(
             self.TEST_DATASET_FULL_PATH)
         beacontwo = dataset.beacon(chromosome='13',
                                    coordinate=113803460,
diff --git a/solvebio/test/test_dataset.py b/solvebio/test/test_dataset.py
index e4d13542..ab16a288 100644
--- a/solvebio/test/test_dataset.py
+++ b/solvebio/test/test_dataset.py
@@ -14,7 +14,8 @@ def test_dataset_retrieval(self):
         self.assertTrue('id' in dataset,
                         'Should be able to get id in dataset')
 
-        check_fields = ['class_name', 'created_at',
+        check_fields = ['class_name',
+                        'created_at',
                         'data_url',
                         'vault_id',
                         'vault_object_id',
@@ -49,7 +50,7 @@ def test_dataset_tree_traversal_shortcuts(self):
         )
 
     def test_dataset_fields(self):
-        dataset = self.client.Dataset.get_by_full_path(
+        dataset = self.client.Object.get_by_full_path(
             self.TEST_DATASET_FULL_PATH)
         fields = dataset.fields()
         dataset_field = fields.data[0]
@@ -67,22 +68,8 @@ def test_dataset_fields(self):
         self.assertSetEqual(set(dataset_field.keys()), check_fields)
 
     def test_dataset_facets(self):
-        dataset = self.client.Dataset.get_by_full_path(
+        dataset = self.client.Object.get_by_full_path(
             self.TEST_DATASET_FULL_PATH)
         field = dataset.fields('status')
         facets = field.facets()
         self.assertTrue(len(facets['facets']) >= 0)
-
-    """
-    # TODO support a Genomic test dataset (grab clinvar one from API build)
-    def test_dataset_beacon(self):
-        obj = Object.get_by_full_path(self.TEST_DATASET_FULL_PATH)
-        resp = Dataset.retrieve(obj['dataset_id']).beacon(chromosome="6",
-                                                          coordinate=123,
-                                                          allele='G')
-        self.assertTrue('total' in resp and resp['total'] == 0)
-        self.assertTrue('exist' in resp and resp['exist'] == False)
-        self.assertTrue('query' in resp and resp['query']['chromosome'] == '6')
-        self.assertTrue('query' in resp and resp['query']['allele'] == 'G')
-        self.assertTrue('query' in resp and resp['query']['coordinate'] == 123)
-    """
diff --git a/solvebio/test/test_exports.py b/solvebio/test/test_exports.py
index a05bdd34..98b862de 100644
--- a/solvebio/test/test_exports.py
+++ b/solvebio/test/test_exports.py
@@ -14,7 +14,7 @@ class ExportsTests(SolveBioTestCase):
     def setUp(self):
         super(ExportsTests, self).setUp()
         filters = Filter(rgd_id='RGD:2645')
-        self.dataset = self.client.Dataset.get_by_full_path(
+        self.dataset = self.client.Object.get_by_full_path(
             'solvebio:public:/HGNC/3.0.0-2016-11-10/HGNC')
         self.query = self.dataset.query(filters=filters, fields=['rgd_id'],
                                         genome_build='GRCh37', limit=10)
diff --git a/solvebio/test/test_lookup.py b/solvebio/test/test_lookup.py
index fa409d40..a7e43941 100644
--- a/solvebio/test/test_lookup.py
+++ b/solvebio/test/test_lookup.py
@@ -7,7 +7,7 @@ class LookupTests(SolveBioTestCase):
 
     def setUp(self):
         super(LookupTests, self).setUp()
-        self.dataset = self.client.Dataset.get_by_full_path(
+        self.dataset = self.client.Object.get_by_full_path(
             self.TEST_DATASET_FULL_PATH)
 
     def test_lookup_error(self):
diff --git a/solvebio/test/test_query.py b/solvebio/test/test_query.py
index 96d368b6..7c1ca86c 100644
--- a/solvebio/test/test_query.py
+++ b/solvebio/test/test_query.py
@@ -11,7 +11,7 @@ class BaseQueryTest(SolveBioTestCase):
     """Test Paging Queries"""
     def setUp(self):
         super(BaseQueryTest, self).setUp()
-        self.dataset = self.client.Dataset.get_by_full_path(
+        self.dataset = self.client.Object.get_by_full_path(
             self.TEST_DATASET_FULL_PATH)
 
     def test_basic(self):
diff --git a/solvebio/test/test_query_batch.py b/solvebio/test/test_query_batch.py
index 95ea5e1a..fee86727 100644
--- a/solvebio/test/test_query_batch.py
+++ b/solvebio/test/test_query_batch.py
@@ -6,7 +6,7 @@
 class BatchQueryTest(SolveBioTestCase):
     def setUp(self):
         super(BatchQueryTest, self).setUp()
-        self.dataset = self.client.Dataset.get_by_full_path(
+        self.dataset = self.client.Object.get_by_full_path(
             self.TEST_DATASET_FULL_PATH)
 
     def test_invalid_batch_query(self):