Update src/utils.py

src/utils.py

@@ -25,6 +25,27 @@ stop_words.update(additional_stop_words)
 # clear_output()
 
 
+# load operation data
+path1 = "data/brand_belong_category_dict.json"
+path2 = "data/product_upper_category_dict.json"
+path3 = "data/offered_brands.pkl"
+path4 = "data/offer_retailer.csv"
+
+with open(path1, 'r') as f:
+    brand_belong_category_dict = json.load(f)
+
+with open(path2, 'rb') as f:
+    category_dict = json.load(f)
+
+with open(path3, 'rb') as f:
+    offered_brands = pickle.load(f)
+
+df_offers_brand_retailer = pd.read_csv(path4)
+
+example_search = "Simply Spiked Lemonade 12 pack at Walmart"
+
+# helper functions
+
 def single_text_cleaner(text: str, remove_stopwords: bool=False, upper_case: bool = False, remove_punctuation: bool=True) -> str:
     """Clean one single text input. By default it will convert text to lower case"""
     if upper_case:
@@ -251,7 +272,9 @@ def offer_finder_by_entity(search_input: str, entities: Tuple, offers_data: pd.D
         return None
 
 
-def search_offers(search_input: str, offers: pd.DataFrame, category_dict: Dict, brand_belong_category_dict: Dict, score: str = "jaccard", score_threshold: float = 0.0):
+def search_offers(search_input: str=example_search, offers: pd.DataFrame=df_offers_brand_retailer, 
+                  category_dict: Dict=category_dict, brand_belong_category_dict: Dict=brand_belong_category_dict, 
+                  score: str="jaccard", score_threshold: float = 0.0):
     """Main function. Takes in a serach_input and decide whether it can find entities or not. Then excecute the appropriate functions
     Inputs:
     - search_input: a string that a user enters
@@ -282,4 +305,4 @@ def search_offers(search_input: str, offers: pd.DataFrame, category_dict: Dict,
         return search_results
 
 if __name__ == "__main__":
-    print("This is a utils file. You should run it from gradio_deploy.py")
+    search_offers()