Improve consistency of client

Author: rliebz

dataverse/dataset.py

@@ -26,10 +26,10 @@ def __init__(self, entry=SWORD_BOOTSTRAP, dataverse=None, edit_uri=None,
         self.edit_uri = edit_uri
         self.edit_media_uri = edit_media_uri
         self.statement_uri = statement_uri
+        self.is_deleted = False
 
         self._entry = etree.XML(entry) if isinstance(entry, str) else entry
         self._statement = None
-        self._state = None
         self._json = {}
         self._id = None
 
@@ -159,16 +159,15 @@ def get_statement(self, refresh=False):
         return self._statement
 
     def get_state(self, refresh=False):
-        if not refresh and self._state or self._state == 'DEACCESSIONED':
-            return self._state
+        if self.is_deleted:
+            return 'DEACCESSIONED'
 
-        self._state = get_element(
+        return get_element(
             self.get_statement(refresh),
             tag='category',
             attribute='term',
             attribute_value='latestVersionState'
         ).text
-        return self._state
 
     def get_json(self, version='latest', refresh=False):
         if not refresh and self._json.get(version):
@@ -193,15 +192,28 @@ def get_json(self, version='latest', refresh=False):
         self._json[version] = resp.json()['data']
         return self._json[version]
 
-    def get_file(self, file_name, version='latest', refresh=True):
+    def publish(self):
+        resp = requests.post(
+            self.edit_uri,
+            headers={'In-Progress': 'false', 'Content-Length': 0},
+            auth=self.connection.auth,
+        )
+
+        if resp.status_code != 200:
+            raise OperationFailedError('The Dataverse could not be published.')
+
+        receipt = resp.content
+        self._refresh(receipt=receipt, published=True)
+
+    def get_file(self, file_name, version='latest', refresh=False):
         files = self.get_files(version, refresh)
         return next((f for f in files if f.name == file_name), None)
 
-    def get_file_by_id(self, file_id, version='latest', refresh=True):
+    def get_file_by_id(self, file_id, version='latest', refresh=False):
         files = self.get_files(version, refresh)
         return next((f for f in files if f.id == file_id), None)
 
-    def get_files(self, version='latest', refresh=True):
+    def get_files(self, version='latest', refresh=False):
         try:
             files_json = self.get_json(version, refresh)['files']
             return [DataverseFile.from_json(self, file_json)
@@ -250,20 +262,7 @@ def upload_file(self, filename, content, zip=True):
             auth=self.connection.auth,
         )
 
-        self._refresh()
-
-    def publish(self):
-        resp = requests.post(
-            self.edit_uri,
-            headers={'In-Progress': 'false', 'Content-Length': 0},
-            auth=self.connection.auth,
-        )
-
-        if resp.status_code != 200:
-            raise OperationFailedError('The Dataverse could not be published.')
-
-        receipt = resp.content
-        self._refresh(receipt=receipt)
+        self.get_json(refresh=True)
 
     def delete_file(self, dataverse_file):
         resp = requests.delete(
@@ -274,6 +273,8 @@ def delete_file(self, dataverse_file):
         if resp.status_code != 204:
             raise OperationFailedError('The file could not be deleted.')
 
+        self.get_json(refresh=True)
+
     def delete_all_files(self):
         for f in self.get_files():
             self.delete_file(f)
@@ -290,7 +291,7 @@ def delete_all_files(self):
     #     self._refresh(deposit_receipt=depositReceipt)
 
     # if we perform a server operation, we should refresh the dataset object
-    def _refresh(self, receipt=None):
+    def _refresh(self, receipt=None, published=False):
         if receipt:
             self.edit_uri = get_element(
                 receipt,
@@ -312,4 +313,6 @@ def _refresh(self, receipt=None):
             ).get('href')
         self.get_statement(refresh=True)
         self.get_entry(refresh=True)
-        self.get_state(refresh=True)
+
+        update_version = 'latest-published' if published else 'latest'
+        self.get_json(update_version, refresh=True)

dataverse/dataverse.py

@@ -87,13 +87,17 @@ def publish(self):
         if resp.status_code != 200:
             raise OperationFailedError('The Dataverse could not be published.')
 
-    def add_dataset(self, dataset):
-        if get_element(dataset._entry, 'title', 'dcterms') is None:
-            raise InsufficientMetadataError('This dataset must have a title.')
-        if get_element(dataset._entry, 'description', 'dcterms') is None:
-            raise InsufficientMetadataError('This dataset must have a description.')
-        if get_element(dataset._entry, 'creator', 'dcterms') is None:
-            raise InsufficientMetadataError('This dataset must have an author.')
+    def create_dataset(self, title, description, creator, **kwargs):
+        dataset = Dataset(
+            title=title,
+            description=description,
+            creator=creator,
+            **kwargs
+        )
+
+        self._add_dataset(dataset)
+
+    def _add_dataset(self, dataset):
 
         resp = requests.post(
             self.collection.get('href'),
@@ -110,7 +114,7 @@ def add_dataset(self, dataset):
         self.get_collection_info(refresh=True)
 
     def delete_dataset(self, dataset):
-        if dataset._state == 'DELETED' or dataset._state == 'DEACCESSIONED':
+        if dataset.get_state() == 'DELETED' or dataset.get_state() == 'DEACCESSIONED':
             return
 
         resp = requests.delete(
@@ -124,7 +128,7 @@ def delete_dataset(self, dataset):
                 'https://github.com/IQSS/dataverse/issues/778'
             )
 
-        dataset._state = 'DEACCESSIONED'
+        dataset.is_deleted = True
         self.get_collection_info(refresh=True)
 
     def get_datasets(self, refresh=False):

dataverse/test/test_dataverse.py

@@ -220,19 +220,28 @@ def teardown_class(cls):
     def setup_method(self, method):
 
         # create a dataset for each test
-        s = Dataset(**PICS_OF_CATS_DATASET)
-        self.dataverse.add_dataset(s)
-        self.dataset = self.dataverse.get_dataset_by_doi(s.doi)
+        dataset = Dataset(**PICS_OF_CATS_DATASET)
+        self.dataverse._add_dataset(dataset)
+        self.dataset = self.dataverse.get_dataset_by_doi(dataset.doi)
 
     def teardown_method(self, method):
         try:
             self.dataverse.delete_dataset(self.dataset)
         finally:
             return
 
-    def test_create_dataset_from_xml(self):
+    def test_create_dataset(self):
+        title = str(uuid.uuid1())   # must be unique
+        self.dataverse.create_dataset(title, 'Descripty', 'foo@test.com')
+        dataset = self.dataverse.get_dataset_by_title(title)
+        try:
+            assert dataset.title == title
+        finally:
+            self.dataverse.delete_dataset(dataset)
+
+    def test_add_dataset_from_xml(self):
         new_dataset = Dataset.from_xml_file(ATOM_DATASET)
-        self.dataverse.add_dataset(new_dataset)
+        self.dataverse._add_dataset(new_dataset)
         retrieved_dataset = self.dataverse.get_dataset_by_title('Roasting at Home')
         assert retrieved_dataset
         self.dataverse.delete_dataset(retrieved_dataset)
@@ -279,7 +288,7 @@ def test_delete_a_file(self):
 
     def test_delete_a_dataset(self):
         xmlDataset = Dataset.from_xml_file(ATOM_DATASET)
-        self.dataverse.add_dataset(xmlDataset)
+        self.dataverse._add_dataset(xmlDataset)
         atomDataset = self.dataverse.get_dataset_by_title('Roasting at Home')
         num_datasets = len(self.dataverse.get_datasets())