From f3c1babfc50a8b6f685425145ffa70d0d7b22804 Mon Sep 17 00:00:00 2001
From: Trent Smith <1429913+Bento007@users.noreply.github.com>
Date: Wed, 18 Oct 2017 09:59:32 -0700
Subject: [PATCH] Multiple result views supported for search (#564)

Added format to post search. Returns all meta data for bundles in results when format=raw. https://www.elastic.co/guide/en/elasticsearch/reference/current/search-request-body.html
- added description to format in dss-api.yml. Added smoketest case to use format raw.
- updated test search to validate new raw format
---
 dss-api.yml          |  6 ++++++
 dss/api/search.py    | 22 ++++++++++++++++------
 tests/smoketest.py   |  2 +-
 tests/test_search.py | 14 ++++++++++++++
 4 files changed, 37 insertions(+), 7 deletions(-)

diff --git a/dss-api.yml b/dss-api.yml
index 19910fcd84..307accf73d 100644
--- a/dss-api.yml
+++ b/dss-api.yml
@@ -48,6 +48,12 @@ paths:
                 type: object
             required:
               - es_query
+        - name: format
+          in: query
+          description: Specifies the output format. `raw` will return the json meta data in the results.
+          required: false
+          type: string
+          enum: [raw]
         - name: replica
           in: query
           description: Replica to search.
diff --git a/dss/api/search.py b/dss/api/search.py
index df6a059126..084464b6c0 100644
--- a/dss/api/search.py
+++ b/dss/api/search.py
@@ -22,7 +22,11 @@ def check(cls, n):
 
 
 @dss_handler
-def post(json_request_body: dict, replica: str, per_page: int, _scroll_id: typing.Optional[str] = None) -> dict:
+def post(json_request_body: dict,
+         replica: str,
+         per_page: int,
+         _scroll_id: typing.Optional[str] = None,
+         format: typing.Optional[str] = None) -> dict:
     es_query = json_request_body['es_query']
     get_logger().debug("Received posted query. Replica: %s Query: %s Per_page: %i Timeout: %s Scroll_id: %s",
                        replica, json.dumps(es_query, indent=4), per_page, _scroll_id)
@@ -63,14 +67,20 @@ def post(json_request_body: dict, replica: str, per_page: int, _scroll_id: typin
 
         # TODO: (tsmith12) allow users to retrieve previous search results
         _scroll_id = page['_scroll_id']
-        result_list = [{
-            'bundle_id': hit['_id'],
-            'bundle_url': _build_bundle_url(hit, replica),
-            'search_score': hit['_score']
-        } for hit in page['hits']['hits']]
+
+        result_list = []  # type: typing.List[dict]
+        for hit in page['hits']['hits']:
+            result = {'bundle_id': hit['_id'],
+                      'bundle_url': _build_bundle_url(hit, replica),
+                      'search_score': hit['_score']
+                      }
+            if format == 'raw':
+                result['meta_data'] = hit['_source']
+            result_list.append(result)
 
         # TODO: (tsmith12) if page returns 0 hits, then all results have been found. delete search id
         request_body = jsonify({'es_query': es_query, 'results': result_list, 'total_hits': page['hits']['total']})
+
         if len(result_list) < per_page:
             response = make_response(request_body, requests.codes.ok)
         else:
diff --git a/tests/smoketest.py b/tests/smoketest.py
index ac411562e5..4f91dda0d0 100755
--- a/tests/smoketest.py
+++ b/tests/smoketest.py
@@ -69,7 +69,7 @@ def run(command, runner=check_call, **kwargs):
 run("hca dss download --replica gcp $(jq -r .bundle_uuid upload.json)")
 
 for replica in "aws", "gcp":
-    run(f"hca dss post-search --es-query='{{}}' --replica {replica}")
+    run(f"hca dss post-search --es-query='{{}}' --format raw --replica {replica}")
 
 search_route = "https://${API_HOST}/v1/search"
 for replica in "aws", "gcp":
diff --git a/tests/test_search.py b/tests/test_search.py
index 7798e6197c..1a3a5834f8 100755
--- a/tests/test_search.py
+++ b/tests/test_search.py
@@ -244,6 +244,20 @@ def test_page_has_N_results_when_per_page_is_N(self):
                 next_url = self.get_next_url(search_obj.response.headers)
                 self.verify_next_url(next_url, per_page)
 
+    def test_format_is_raw(self):
+        bundles = self.populate_search_index(self.index_document, 1)
+        self.check_count(smartseq2_paired_ends_query, 1)
+        url = self.build_url(url_params={'format': 'raw'})
+        search_obj = self.assertPostResponse(
+            path=url,
+            json_request_body=dict(es_query=smartseq2_paired_ends_query),
+            expected_code=requests.codes.ok)
+        next_url = self.get_next_url(search_obj.response.headers)
+        self.assertIsNone(next_url)
+        self.verify_search_result(search_obj.json, smartseq2_paired_ends_query, 1, 1)
+        self.verify_bundles(search_obj.json['results'], bundles)
+        self.assertEqual(search_obj.json['results'][0]['meta_data'], self.index_document)
+
     def test_error_returned_when_per_page_is_out_of_range(self):
         expected_error = ExpectedErrorFields(code="illegal_arguments",
                                              status=requests.codes.bad_request,