Sort Apps on app listing page based on findings (#417)

* remove topics from prompt data

* Sort retrievals data

* Ruff Fixes

* UT fix.

pebblo/app/models/models.py

@@ -26,8 +26,12 @@ class AiDataModel(BaseModel):
     data: Optional[Union[list, str]]
     entityCount: int
     entities: dict
-    topicCount: int
-    topics: dict
+    topicCount: Optional[int] = None
+    topics: Optional[dict] = None
+
+    def dict(self, **kwargs):
+        kwargs["exclude_none"] = True
+        return super().dict(**kwargs)
 
 
 class AiDocs(BaseModel):
@@ -100,7 +104,11 @@ class RetrievalContext(BaseModel):
 
 class AiClassificationData(BaseModel):
     entities: dict
-    topics: dict
+    topics: Optional[dict] = None
+
+    def dict(self, **kwargs):
+        kwargs["exclude_none"] = True
+        return super().dict(**kwargs)
 
 
 class RetrievalData(BaseModel):

pebblo/app/service/local_ui_service.py

@@ -134,6 +134,12 @@ def prepare_retrieval_response(self, app_dir, app_json):
             logger.warning(f"Skipping app '{app_dir}' due to missing or invalid file")
             return
 
+        # Sort retrievals data
+        retrievals = self.sort_retrievals_data(
+            app_metadata_content.get("retrievals", [])
+        )
+        app_metadata_content["retrievals"] = retrievals
+
         # fetch total retrievals
         for retrieval in app_metadata_content.get("retrievals", []):
             retrieval_data = {"name": app_json.get("name")}
@@ -336,6 +342,8 @@ def get_latest_load_id(load_ids, app_dir):
     def get_retrieval_app_details(self, app_content):
         retrieval_data = app_content.get("retrievals", [])
 
+        retrieval_data = self.sort_retrievals_data(retrieval_data)
+
         active_users = self.get_active_users(retrieval_data)
         documents = self.get_all_documents(retrieval_data)
         vector_dbs = self.get_all_vector_dbs(retrieval_data)
@@ -474,6 +482,17 @@ def sort_retrievals(retrieval_data: list, search_key: str) -> dict:
 
         return sorted_resp
 
+    @staticmethod
+    def _calculate_total_count(item: dict):
+        prompt_count = item.get("prompt", {}).get("entityCount") or 0
+        response_count = item.get("prompt", {}).get("entityCount") or 0
+        return prompt_count + response_count
+
+    def sort_retrievals_data(self, retrieval):
+        # Sort the list based on the total count in descending order
+        sorted_data = sorted(retrieval, key=self._calculate_total_count, reverse=True)
+        return sorted_data
+
     def get_all_documents(self, retrieval_data: list) -> dict:
         """
         This function returns documents per app with its metadata in following format:

pebblo/app/service/prompt_service.py

@@ -33,7 +33,7 @@ def __init__(self, data: dict):
         self.entity_classifier_obj = EntityClassifier()
         self.topic_classifier_obj = TopicClassifier()
 
-    def _fetch_classified_data(self, input_data):
+    def _fetch_classified_data(self, input_data, input_type=""):
         """
         Retrieve input data and return its corresponding model object with classification.
         """
@@ -47,15 +47,15 @@ def _fetch_classified_data(self, input_data):
         ) = self.entity_classifier_obj.presidio_entity_classifier_and_anonymizer(
             input_data
         )
-        topics, topic_count = self.topic_classifier_obj.predict(input_data)
-
-        data = {
-            "data": input_data,
-            "entityCount": entity_count,
-            "entities": entities,
-            "topicCount": topic_count,
-            "topics": topics,
-        }
+
+        data = {"data": input_data, "entityCount": entity_count, "entities": entities}
+
+        # Topic classification is performed only for the response.
+        if input_type == "response":
+            topics, topic_count = self.topic_classifier_obj.predict(input_data)
+            data["topicCount"] = topic_count
+            data["topics"] = topics
+
         logger.debug(f"AI_APPS [{self.application_name}]:Classified Details: {data}")
         return data
 
@@ -154,7 +154,7 @@ def process_request(self):
 
             # getting response data
             response_data = self._fetch_classified_data(
-                self.data.get("response", {}).get("data")
+                self.data.get("response", {}).get("data"), input_type="response"
             )
 
             # getting retrieval context data

tests/app/test_prompt_api.py

@@ -52,7 +52,6 @@ def test_app_prompt_success(mock_write_json_to_file):
     assert response.json()["message"] == "AiApp prompt request completed successfully"
     assert response.json()["retrieval_data"]["prompt"] == {
         "entities": {},
-        "topics": {},
     }
     assert response.json()["retrieval_data"]["response"] == {
         "entities": {"us-ssn": 1},