Update main.py

main.py

@@ -10,6 +10,8 @@ import pickle
 import requests
 from datetime import date, datetime
 
+
+TMDB_API_KEY = os.environ.get("TMDB_API_KEY")
 # load the nlp model and tfidf vectorizer from disk
 filename = 'nlp_model.pkl'
 clf = pickle.load(open(filename, 'rb'))
@@ -42,6 +44,58 @@ def home():
     return render_template('home.html',suggestions=suggestions)
 
 
+def get_tmdb_reviews_from_imdb(imdb_id, max_reviews=10):
+    """
+    Given an imdb_id like 'tt1234567', return a dict {review_text: sentiment}
+    using TMDb's API. Returns {} on failure or if no reviews found.
+    """
+    if not TMDB_API_KEY:
+        # Key not available — caller will handle fallback
+        print("TMDb API key not set; skipping TMDb fetch.")
+        return {}
+
+    try:
+        find_url = f"https://api.themoviedb.org/3/find/{imdb_id}"
+        params = {"api_key": TMDB_API_KEY, "external_source": "imdb_id"}
+        r = requests.get(find_url, params=params, timeout=8)
+        r.raise_for_status()
+        found = r.json()
+
+        results = found.get("movie_results") or []
+        if not results:
+            return {}
+
+        tmdb_id = results[0]["id"]
+
+        reviews_url = f"https://api.themoviedb.org/3/movie/{tmdb_id}/reviews"
+        params = {"api_key": TMDB_API_KEY, "language": "en-US", "page": 1}
+        r2 = requests.get(reviews_url, params=params, timeout=8)
+        r2.raise_for_status()
+        reviews_data = r2.json()
+
+        out = {}
+        count = 0
+        for item in reviews_data.get("results", []):
+            if count >= max_reviews:
+                break
+            content = item.get("content")
+            if not content:
+                continue
+            # run your classifier to keep UX consistent
+            movie_review_list = np.array([content])
+            movie_vector = vectorizer.transform(movie_review_list)
+            pred = clf.predict(movie_vector)
+            sentiment = "Positive" if pred else "Negative"
+            out[content] = sentiment
+            count += 1
+
+        return out
+
+    except Exception as e:
+        print(f"TMDb fetch failed: {e}")
+        return {}
+
+
 @app.route("/recommend",methods=["POST"])
 @app.route("/recommend", methods=["POST"])
 def recommend():
@@ -108,7 +162,7 @@ def recommend():
         for i in range(len(cast_places))
     }
 
-    # ---------- SAFE REVIEWS BLOCK ----------
+   
     movie_reviews = {}  # default: no reviews
     try:
         url = f"https://www.imdb.com/title/{imdb_id}/reviews?ref_=tt_ov_rt"
@@ -135,13 +189,16 @@ def recommend():
                 pred = clf.predict(movie_vector)
                 reviews_status.append('Positive' if pred else 'Negative')
 
-        movie_reviews = {
-            reviews_list[i]: reviews_status[i] for i in range(len(reviews_list))
-        }
+        if reviews_list:
+            movie_reviews = {
+                reviews_list[i]: reviews_status[i] for i in range(len(reviews_list))
+            }
     except Exception as e:
         print(f"IMDb reviews fetch failed: {e}")
         movie_reviews = {}
-    # ---------- END SAFE REVIEWS BLOCK ----------
+
+    if not movie_reviews:
+        movie_reviews = get_tmdb_reviews_from_imdb(imdb_id)
 
     # dates
     movie_rel_date = ""