News API add ability to retrieve images (#1076) (#1082)

* News API add ability to retrieve images

* Updated following review

* Set superdesk core version

(cherry picked from commit d74c6ec)

features/news_api_item.feature

@@ -20,7 +20,7 @@ Feature: News API Item
       "headline": "Headline of the story"
     }]
     """
-    When we get "v1/news/item/#items._id#"
+    When we get "/news/item/#items._id#"
     Then we get OK response
 
   Scenario: Attempt to Retrieve an item with unknown format
@@ -32,7 +32,7 @@ Feature: News API Item
       "headline": "Headline of the story"
     }]
     """
-    When we get "v1/news/item/#items._id#?format=bogus"
+    When we get "/news/item/#items._id#?format=bogus"
     Then we get response code 404
 
   Scenario: Retrieve an item that does not exist
@@ -44,7 +44,7 @@ Feature: News API Item
       "headline": "Headline of the story"
     }]
     """
-    When we get "v1/news/item/999"
+    When we get "/news/item/999"
     Then we get response code 404
 
   Scenario: Retrieve a version of an item
@@ -57,7 +57,7 @@ Feature: News API Item
       "version" : "5"
     }]
     """
-    When we get "v1/news/item/111?version=5"
+    When we get "/news/item/111?version=5"
     Then we get OK response
 
   Scenario: Retrieve an item in ninjs
@@ -69,7 +69,7 @@ Feature: News API Item
       "headline": "Headline of the story"
     }]
     """
-    When we get "v1/news/item/#items._id#?format=NINJSFormatter"
+    When we get "/news/item/#items._id#?format=NINJSFormatter"
     Then we get existing resource
     """
      {"guid": "111",
@@ -86,7 +86,7 @@ Feature: News API Item
         "versioncreated": "2018-11-01T03:01:40.000Z"
       }]
       """
-    When we get "v1/news/item/#items._id#?format=NINJSFormatter"
+    When we get "/news/item/#items._id#?format=NINJSFormatter"
     Then we get response code 404
 
   Scenario: Retrieve an item in text format
@@ -99,7 +99,46 @@ Feature: News API Item
       "body_html": "<p>test&nbsp;test</p>"
     }]
     """
-    When we get "v1/news/item/#items._id#?format=TextFormatter"
+    When we get "/news/item/#items._id#?format=TextFormatter"
     Then we get OK response
     Then we get "test test" in text response
 
+  Scenario: Retrieve an item with associations
+    Given "items"
+    """
+    [{
+      "_id": "111",
+      "pubstatus": "usable",
+      "headline": "Headline of the story",
+      "body_html": "<p>test&nbsp;test</p>",
+      "associations": {
+        "featuremedia": {
+          "renditions": {
+            "16-9": {
+              "href": "/assets/1234567"
+            },
+            "_newsroom_thumbnail": {
+              "href": "/assets/987654"
+            }
+          }
+        }
+      }
+    }]
+    """
+    When we get "/news/item/#items._id#?format=NINJSFormatter2"
+    Then we get existing resource
+    """
+    {
+      "guid": "111",
+      "headline": "Headline of the story",
+      "associations": {
+        "featuremedia": {
+          "renditions": {
+            "16-9": {
+              "href": "/assets/1234567"
+            }
+          }
+        }
+      }
+    }
+    """

newsroom/news_api/news/assets/assets.py

@@ -0,0 +1,41 @@
+import superdesk
+import flask
+from newsroom.news_api.api_tokens import CompanyTokenAuth
+from flask import abort
+from newsroom.upload import ASSETS_RESOURCE
+from flask_babel import gettext
+import bson.errors
+from werkzeug.wsgi import wrap_file
+from newsroom.news_api.utils import post_api_audit
+
+blueprint = superdesk.Blueprint('assets', __name__)
+
+
+def init_app(app):
+    superdesk.blueprint(blueprint, app)
+
+
+@blueprint.route('/assets/<path:asset_id>', methods=['GET'])
+def get_item(asset_id):
+    if CompanyTokenAuth().check_auth(flask.request.headers.get('Authorization'), None, None, 'GET'):
+        try:
+            media_file = flask.current_app.media.get(asset_id, ASSETS_RESOURCE)
+        except bson.errors.InvalidId:
+            media_file = None
+        if not media_file:
+            flask.abort(404)
+
+        data = wrap_file(flask.request.environ, media_file, buffer_size=1024 * 256)
+        response = flask.current_app.response_class(
+            data,
+            mimetype=media_file.content_type,
+            direct_passthrough=True)
+        response.content_length = media_file.length
+        response.last_modified = media_file.upload_date
+        response.set_etag(media_file.md5)
+        response.make_conditional(flask.request)
+        response.headers['Content-Disposition'] = 'inline'
+        post_api_audit({'_items': [{'_id': asset_id}]})
+        return response
+    else:
+        abort(401, gettext('Invalid token'))

newsroom/news_api/news/item/item.py

@@ -1,5 +1,4 @@
 import superdesk
-from newsroom.news_api.settings import URL_PREFIX
 import flask
 from superdesk import get_resource_service
 from flask import current_app as app, abort
@@ -10,7 +9,11 @@
 blueprint = superdesk.Blueprint('news/item', __name__)
 
 
-@blueprint.route('/{}/news/item/<path:item_id>'.format(URL_PREFIX), methods=['GET'])
+def init_app(app):
+    superdesk.blueprint(blueprint, app)
+
+
+@blueprint.route('/news/item/<path:item_id>', methods=['GET'])
 def get_item(item_id):
     if CompanyTokenAuth().check_auth(flask.request.headers.get('Authorization'), None, None, 'GET'):
         _format = flask.request.args.get('format', 'NINJSFormatter')

newsroom/news_api/news/search_service.py

@@ -16,7 +16,7 @@
 from content_api.errors import BadParameterValueError, UnexpectedParameterError
 
 from newsroom.news_api.settings import ELASTIC_DATETIME_FORMAT
-from newsroom.news_api.utils import post_api_audit
+from newsroom.news_api.utils import post_api_audit, remove_internal_renditions
 from newsroom.search import BaseSearchService, query_string
 from newsroom.products.products import get_products_by_company
 
@@ -40,7 +40,7 @@ class NewsAPINewsService(BaseSearchService):
     # set of fields that can be specified in the include_fields parameter
     allowed_include_fields = {'type', 'urgency', 'priority', 'language', 'description_html', 'located', 'keywords',
                               'source', 'subject', 'place', 'wordcount', 'charcount', 'body_html', 'readtime',
-                              'profile', 'service', 'genre'}
+                              'profile', 'service', 'genre', 'associations'}
 
     default_fields = {
         '_id', 'uri', 'embargoed', 'pubstatus', 'ednote', 'signal', 'copyrightnotice', 'copyrightholder',
@@ -49,7 +49,7 @@ class NewsAPINewsService(BaseSearchService):
 
     # set of fields that will be removed from all responses, we are not currently supporting associations and
     # the products embedded in the items are the superdesk products
-    mandatory_exclude_fields = {'associations', '_current_version', 'products'}
+    mandatory_exclude_fields = {'_current_version', 'products'}
 
     section = 'news_api'
     limit_days_setting = 'news_api_time_limit_days'
@@ -63,10 +63,14 @@ def get(self, req, lookup):
         exclude_fields = self.mandatory_exclude_fields.union(
             set(orig_request_params.get('exclude_fields').split(','))) if orig_request_params.get(
             'exclude_fields') else self.mandatory_exclude_fields
+
         for doc in resp.docs:
             for field in exclude_fields:
                 doc.pop(field, None)
 
+            if 'associations' in orig_request_params.get('include_fields', ''):
+                remove_internal_renditions(doc)
+
         return resp
 
     def prefill_search_query(self, search, req=None, lookup=None):

newsroom/news_api/settings.py

@@ -14,19 +14,17 @@
     'newsroom.news_api.products',
     'newsroom.news_api.formatters',
     'newsroom.news_api.news',
-    'newsroom.news_api.news.item',
+    'newsroom.news_api.news.item.item',
     'newsroom.news_api.news.search',
     'newsroom.news_api.news.feed',
     'newsroom.products',
     'newsroom.news_api.api_audit',
+    'newsroom.news_api.news.assets.assets',
+    'newsroom.upload',
 ]
 
 INSTALLED_APPS = []
 
-BLUEPRINTS = [
-    'newsroom.news_api.news.item.item'
-]
-
 #: mongo db name, only used when mongo_uri is not set
 MONGO_DBNAME = env('MONGO_DBNAME', 'newsroom')
 

newsroom/news_api/utils.py

@@ -33,3 +33,17 @@ def format_report_results(search_result, unique_endpoints, companies):
                 unique_endpoints.append(endpoint_bucket['key'])
 
     return results
+
+
+def remove_internal_renditions(item):
+    clean_renditions = dict()
+
+    # associations featuremedia will contain the internal newsroom renditions, we need to remove these.
+    if ((item.get('associations') or {}).get('featuremedia') or {}).get('renditions'):
+        for key, rendition in\
+                item['associations']['featuremedia']['renditions'].items():
+            if not key.startswith('_newsroom'):
+                clean_renditions[key] = rendition
+        item['associations']['featuremedia']['renditions'] = clean_renditions
+
+    return item

newsroom/upload.py

@@ -64,4 +64,5 @@ def init_app(app):
     app.config['DOMAIN'].setdefault('upload', {
         'authentication': None,
         'mongo_prefix': newsroom.MONGO_PREFIX,
+        'internal_resource': True
     })

newsroom/wire/formatters/__init__.py

@@ -23,3 +23,4 @@ def __init__(cls, name, bases, attrs):
 from .json import JsonFormatter  # noqa
 from .ninjs import NINJSFormatter  # noqa
 from .picture import PictureFormatter  # noqa
+from .ninjs2 import NINJSFormatter2  # noqa

newsroom/wire/formatters/ninjs2.py

@@ -0,0 +1,14 @@
+from .ninjs import NINJSFormatter
+from newsroom.news_api.utils import remove_internal_renditions
+
+
+class NINJSFormatter2(NINJSFormatter):
+    """
+    Overload the NINJSFormatter and add the associations as a field to copy
+    """
+
+    def __init__(self):
+        self.direct_copy_properties += ('associations',)
+
+    def _transform_to_ninjs(self, item):
+        return remove_internal_renditions(super()._transform_to_ninjs(item))

requirements.txt

@@ -5,4 +5,4 @@ xhtml2pdf
 werkzeug>=0.9.4,<=0.11.15
 -e .
 git+git://github.com/superdesk/[email protected]#egg=superdesk-planning
-git+git://github.com/superdesk/superdesk-core.git@master#egg=superdesk-core
+git+git://github.com/superdesk/superdesk-core.git@v1.33.7#egg=superdesk-core

tests/news_api/test_api_assets.py

@@ -0,0 +1,29 @@
+import os
+from superdesk.storage.desk_media_storage import SuperdeskGridFSMediaStorage
+from tests.news_api.test_api_audit import audit_check
+
+
+def get_fixture_path(fixture):
+    return os.path.join(os.path.dirname(__file__), '../fixtures', fixture)
+
+
+def setup_image(app):
+    with open(get_fixture_path('picture.jpg'), 'rb') as f:
+        res = SuperdeskGridFSMediaStorage(app=app).put(f, 'picture.jpg', content_type='image/jpg')
+    return res
+
+
+def test_get_asset(client, app):
+    app.data.insert('companies', [{"_id": "company_123", "name": "Test Company", "is_enabled": True}])
+    app.data.insert('news_api_tokens', [{"company": "company_123", "enabled": True}])
+    token = app.data.find_one('news_api_tokens', req=None, company='company_123')
+
+    id = setup_image(app)
+    response = client.get('api/v1/assets/{}'.format(id), headers={'Authorization': token.get('token')})
+    assert response.status_code == 200
+    audit_check(str(id))
+
+
+def test_authorization_get_asset(client, app):
+    response = client.get('api/v1/assets/{}'.format(id), headers={'Authorization': 'xxxxxxxx'})
+    assert response.status_code == 401