Removing unnecessary time sleep to prevent heroku timeout (#11)

* Feature/tweet acquisition (#1)

* API initial config

* Tweet acquisition API tracking and writing csv file correctly. TODO: Set stop-point to tweet streaming

* Timezone convertion WIP

* Added limit to stop streaming tweets.

* Tweet acquisition done

Co-authored-by: Fabiana Garcia <fabiana.garcia@fit-tecnologia.org.br>

* Sentimental analysis (#2)

* Saved best model

* Preprocessing

* Applying model to live search data. Return positive/negative ratio. Throw exception when no data is found.

* LIME Text explainer

* Mongodb connection

* Sending tweets to crowdsourcing database

Co-authored-by: Fabiana Garcia <fabiana.garcia@fit-tecnologia.org.br>

* Trending topics API (#3)

Co-authored-by: Fabiana Garcia <fabiana.garcia@fit-tecnologia.org.br>

* Procfile (#4)

Co-authored-by: Fabiana Garcia <fabiana.garcia@fit-tecnologia.org.br>

* Fixing cors

* Hotfix/async functions (#6)

* Async functions to prevent 404 error

* Fixing file reading error

* Removed csv writer methods. Everything is being done on MongoDB cloud.

Co-authored-by: Fabiana Garcia <fabiana.garcia@fit-tecnologia.org.br>

* Hotfix - Trending topics

* Fixing cors (#7)

Co-authored-by: Fabiana Garcia <fabiana.garcia@fit-tecnologia.org.br>

Co-authored-by: Fabiana Garcia <fabiana.garcia@fit-tecnologia.org.br>

Author: fabianamasini

api_routes.py

@@ -43,4 +43,4 @@ def get(self):
         try:
             return SearchResultParser().trending_topics_parser(23424768), 200
         except:
-            return 'INTERNAL SERVER ERROR', 500
+            return 'INTERNAL SERVER ERROR', 500

repositories/twitter/tweet_acquisition_repository.py

@@ -15,7 +15,7 @@ def stream_tweets(self, keyword):
         auth.set_access_token(ACCESS_TOKEN, ACCESS_TOKEN_SECRET)
         stream = Stream(auth, stream_listener)
         stream.filter(track=[keyword], languages=["pt"], is_async=True)
-        time.sleep(15)
+        time.sleep(10)
         stream.disconnect()
 
     def get_trending_topics(self, woeid):

services/search_service.py

@@ -11,7 +11,6 @@ class SearchService():
 
     def search(self, keyword):
         tweet_acquisition_repository.stream_tweets(keyword)
-        time.sleep(5)
         try:
             df = MongoDbRepository().read_mongo_into_dataframe(keyword)
             MongoDbRepository().crowdsourcing_mongo_import(df)