cleaned with black

app.py

@@ -19,6 +19,7 @@ index = pinecone.Index("movies")
 
 app = Flask(__name__, template_folder="./templates")
 
+
 def title2trakt_id(title: str, df=df):
     # Matches Exact Title, Otherwise Returns None
     records = df[df["title"].str.lower() == title.lower()]
@@ -29,27 +30,36 @@ def title2trakt_id(title: str, df=df):
     else:
         return 2, records.trakt_id.tolist()
 
+
 def get_vector_value(trakt_id: int):
     fetch_response = index.fetch(ids=[str(trakt_id)])
     return fetch_response["vectors"][str(trakt_id)]["values"]
 
-def query_vectors(vector: list, top_k: int = 20, include_values: bool = False, include_metada: bool = True):
+
+def query_vectors(
+    vector: list,
+    top_k: int = 20,
+    include_values: bool = False,
+    include_metada: bool = True,
+):
     query_response = index.query(
         queries=[
             (vector),
         ],
         top_k=top_k,
         include_values=include_values,
-      include_metadata=include_metada
+        include_metadata=include_metada,
     )
     return query_response
 
+
 def query2ids(query_response):
     trakt_ids = []
     for match in query_response["results"][0]["matches"]:
         trakt_ids.append(int(match["id"]))
     return trakt_ids
 
+
 def get_deets_by_trakt_id(df, trakt_id: int):
     df = df[df["trakt_id"] == trakt_id]
     return {
@@ -58,9 +68,10 @@ def get_deets_by_trakt_id(df, trakt_id: int):
         "runtime": int(df.runtime.values[0]),
         "year": int(df.year.values[0]),
         "trakt_id": trakt_id,
-      "tagline": df.tagline.values[0]
+        "tagline": df.tagline.values[0],
     }
 
+
 @app.route("/similar")
 def get_similar_titles():
     trakt_id = request.args.get("trakt_id")
@@ -111,11 +122,14 @@ def get_similar_titles():
         if cur_res >= max_res:
             break
         deets = get_deets_by_trakt_id(df, trakt_id)
-        if ((deets["year"]>=min_year) and (deets["year"]<=max_year)) and ((deets["runtime"]>=minRuntime) and (deets["runtime"]<=maxRuntime)):
+        if ((deets["year"] >= min_year) and (deets["year"] <= max_year)) and (
+            (deets["runtime"] >= minRuntime) and (deets["runtime"] <= maxRuntime)
+        ):
             results.append(deets)
             cur_res += 1
     return render_template("show_results.html", deets=results)
 
+
 @app.route("/", methods=("GET", "POST"))
 def find_similar_title():
     if request.method == "GET":
@@ -125,7 +139,9 @@ def find_similar_title():
         code, values = title2trakt_id(to_search_title)
         print(f"Code {code} for {to_search_title}")
         if code == 0:
-            search_results = process.extract(to_search_title, movie_titles, scorer=fuzz.token_sort_ratio)
+            search_results = process.extract(
+                to_search_title, movie_titles, scorer=fuzz.token_sort_ratio
+            )
             to_search_titles = []
             to_search_ids = []
             results = []

database.py

@@ -25,15 +25,17 @@ movies_table = Table(
     Column("votes", Integer),
     Column("comment_count", Integer),
     Column("tagline", String),
-    Column("embeddings", PickleType)
+    Column("embeddings", PickleType),
-
 )
 
+
 def init_db_stuff(database_url: str):
     engine = create_engine(database_url)
     meta.create_all(engine)
     Session = sessionmaker(bind=engine)
     return engine, Session
+
+
 """
     movie = {
         "title": movie["movie"]["title"],

db2pc.py

@@ -20,7 +20,15 @@ model = SentenceTransformer("paraphrase-multilingual-MiniLM-L12-v2")
 batch_size = 32
 
 df = pd.read_sql("Select * from movies", engine)
-df["combined_text"] = df["title"] + ": " + df["overview"].fillna('') + " -  " + df["tagline"].fillna('') + " Genres:-  " + df["genres"].fillna('')
+df["combined_text"] = (
+    df["title"]
+    + ": "
+    + df["overview"].fillna("")
+    + " -  "
+    + df["tagline"].fillna("")
+    + " Genres:-  "
+    + df["genres"].fillna("")
+)
 
 print(f'Length of Combined Text: {len(df["combined_text"].tolist())}')
 
@@ -30,8 +38,5 @@ for x in tqdm(range(0,len(df),batch_size)):
     sentences = df["combined_text"][x : x + batch_size].tolist()
     embeddings = model.encode(sentences)
     for idx, value in enumerate(trakt_ids):
-		to_send.append(
+        to_send.append((str(value), embeddings[idx].tolist()))
-			(
-				str(value), embeddings[idx].tolist()
-			))
     index.upsert(to_send)

main.py

@@ -25,7 +25,7 @@ database_url = "sqlite:///jlm.db"
 headers = {
     "Content-Type": "application/json",
     "trakt-api-version": "2",
-	"trakt-api-key": trakt_id
+    "trakt-api-key": trakt_id,
 }
 
 params = {
@@ -34,7 +34,7 @@ params = {
     "page": page,
     "extended": extended,
     "limit": limit,
-	"languages": languages
+    "languages": languages,
 }
 
 
@@ -53,12 +53,11 @@ def create_movie_dict(movie: dict):
         "rating": int(m["rating"]),
         "votes": int(m["votes"]),
         "comment_count": int(m["comment_count"]),
-		"tagline": m["tagline"]
+        "tagline": m["tagline"],
     }
     return movie_dict
 
 
-
 params["limit"] = 1
 res = requests.get(f"{api_base}/search/movie", headers=headers, params=params)
 total_items = res.headers["x-pagination-item-count"]
@@ -111,11 +110,20 @@ for page in tqdm(range(1,max_requests+10)):
         for movie in movies:
             with conn.begin() as trans:
                 stmt = insert(movies_table).values(
-					trakt_id=movie["trakt_id"], title=movie["title"], genres=" ".join(movie["genres"]),
+                    trakt_id=movie["trakt_id"],
-					language=movie["language"], year=movie["year"], released=movie["released"],
+                    title=movie["title"],
-					runtime=movie["runtime"], country=movie["country"], overview=movie["overview"],
+                    genres=" ".join(movie["genres"]),
-					rating=movie["rating"], votes=movie["votes"], comment_count=movie["comment_count"],
+                    language=movie["language"],
-					tagline=movie["tagline"])
+                    year=movie["year"],
+                    released=movie["released"],
+                    runtime=movie["runtime"],
+                    country=movie["country"],
+                    overview=movie["overview"],
+                    rating=movie["rating"],
+                    votes=movie["votes"],
+                    comment_count=movie["comment_count"],
+                    tagline=movie["tagline"],
+                )
                 try:
                     result = conn.execute(stmt)
                     trans.commit()

pc2rec.py

@@ -10,7 +10,9 @@ from sqlalchemy import func
 movie_name = "Forrest Gump"
 
 with engine.connect() as conn:
-	movie_deets = select(movies_table).filter(func.lower(movies_table.columns.title)==func.lower(movie_name))
+    movie_deets = select(movies_table).filter(
+        func.lower(movies_table.columns.title) == func.lower(movie_name)
+    )
     result = conn.execute(movie_deets)
     for row in result:
         print(row)