Added support for Layer Package (lpk) imports (#1)

* Added support for Layer Package (lpk) imports

* Code cleanup. Change "tmp_job" references to "job".

* Add test function for LPK imports

Author: mikegough

databasin/client.py

@@ -149,7 +149,51 @@ def get_temporary_file(self, uuid):
         except HTTPException as e:
             raise_for_authorization(e.response, self.username is not None)
             raise
+    
+    def import_lpk(self, lpk_file):
+        if lpk_file.endswith('.lpk'):
+            f = open(lpk_file, 'rb')
+        else:
+            raise ValueError('File must be an ArcGIS Layer Package with a .lpk extension')
+
+        filename = os.path.basename(lpk_file)
+
+        tmp_file = self.upload_temporary_file(f, filename=filename)
+
+        f.close()
+
+        job_args = {
+            'file': tmp_file.uuid,
+            'url': None,
+            'dataset_type': 'ArcGIS_Native'
+        }
+       
+        job = self.create_job('create_import_job', job_args=job_args, block=True)
+        uri = job.message.split("/")[-2]
+        
+        final_job_args = {
+            'import_id': uri
+        }
+
+        final_job = self.create_job('finalize_import_job', job_args=final_job_args, block=True)
 
+        if final_job.status != 'succeeded':
+            raise DatasetImportError('Import failed: {0}'.format(final_job.message))
+
+        data = json.loads(final_job.message)
+        next_uri = data['next_uri']
+        if '/import/' in next_uri:
+            dataset_import_id = DATASET_IMPORT_ID_RE.search(next_uri).group(1)
+            dataset_import = self.get_import(dataset_import_id)
+            dataset_import.cancel()
+
+            raise DatasetImportError(
+                'Layer Package imports must have all necessary metadata information necessary for one-step import.'
+            )
+
+        dataset_id = next_uri.strip('/').split('/')[-1]
+        return self.get_dataset(dataset_id)
+    
     def import_netcdf_dataset(self, nc_or_zip_file, style=None):
         if nc_or_zip_file.endswith('.zip'):
             f = open(nc_or_zip_file, 'a+b')

tests/test_client.py

@@ -121,6 +121,27 @@ def test_login_no_redirect():
         assert m.call_count == 2
         assert not any(r.url for r in m.request_history if r.url == 'https://databasin.org/redirect/')
 
+def test_import_lpk(import_job_data, dataset_data, tmp_file_data):
+    with requests_mock.mock() as m:
+        m.post('https://databasin.org/uploads/upload-temporary-file/', text=json.dumps({'uuid': 'abcd'}))
+        m.get('https://databasin.org/api/v1/uploads/temporary-files/abcd/', text=json.dumps(tmp_file_data))
+        m.post('https://databasin.org/api/v1/jobs/', headers={'Location': 'https://databasin.org/api/v1/jobs/1234/'})
+        m.get('https://databasin.org/api/v1/jobs/1234/', text=json.dumps(import_job_data))
+        m.get('https://databasin.org/api/v1/datasets/a1b2c3/', text=json.dumps(dataset_data))
+
+        f = six.BytesIO()
+        with mock.patch.object(builtins, 'open', mock.Mock(return_value=f)) as open_mock:
+            c = Client()
+            c._session.cookies['csrftoken'] = 'abcd'
+            dataset = c.import_lpk('test.lpk')
+
+            open_mock.assert_called_once_with('test.lpk', 'rb')
+            assert m.call_count == 7
+            assert dataset.id == 'a1b2c3'
+            request_data = json.loads(m.request_history[2].text)
+            assert request_data['job_name'] == 'create_import_job'
+            assert request_data['job_args']['file'] == 'abcd'
+            assert request_data['job_args']['dataset_type'] == 'ArcGIS_Native'
 
 def test_import_netcdf_dataset_with_zip(import_job_data, dataset_data, tmp_file_data):
     with requests_mock.mock() as m: