From 0514fa063f1f640aff1122764016ed601b9058d2 Mon Sep 17 00:00:00 2001
From: kilian <kilian.scheidecker@orange.fr>
Date: Thu, 20 Nov 2025 18:47:34 +0100
Subject: [PATCH] suggested fix to avoid UnboundLocalError

---
 backend/src/landmarks/cluster_manager.py | 93 +++++++++++++-----------
 1 file changed, 49 insertions(+), 44 deletions(-)

diff --git a/backend/src/landmarks/cluster_manager.py b/backend/src/landmarks/cluster_manager.py
index 85f9af5..92a26a7 100644
--- a/backend/src/landmarks/cluster_manager.py
+++ b/backend/src/landmarks/cluster_manager.py
@@ -102,52 +102,57 @@ class ClusterManager:
             selector = sel,
             out = out
         )
-        except Exception as e:
-            self.logger.warning(f"Error fetching clusters: {e}")
-
-        if result is None :
-            self.logger.debug(f"Found no {cluster_type} clusters, overpass query returned no datapoints.")
-            self.valid = False
-
-        else :
-            points = []
-            for elem in result:
-                osm_type = elem.get('type')
-
-                # Get coordinates and append them to the points list
-                _, coords = get_base_info(elem, osm_type)
-                if coords is not None :
-                    points.append(coords)
-
-            if points :
-                self.all_points = np.array(points)
-
-                # Apply DBSCAN to find clusters. Choose different settings for different cities.
-                if self.cluster_type == 'shopping' and len(self.all_points) > 200 :
-                    dbscan = DBSCAN(eps=0.00118, min_samples=15, algorithm='kd_tree')  # for large cities
-                elif self.cluster_type == 'sightseeing' :
-                    dbscan = DBSCAN(eps=0.0025, min_samples=15, algorithm='kd_tree')  # for historic neighborhoods
-                else :
-                    dbscan = DBSCAN(eps=0.00075, min_samples=10, algorithm='kd_tree')  # for small cities
-
-                labels = dbscan.fit_predict(self.all_points)
-
-                # Check that there are is least 1 cluster
-                if len(set(labels)) > 1 :
-                    self.logger.info(f"Found {len(set(labels))} different {cluster_type} clusters.")
-                    # Separate clustered points and noise points
-                    self.cluster_points = self.all_points[labels != -1]
-                    self.cluster_labels = labels[labels != -1]
-                    self.filter_clusters()      # ValueError here sometimes. I dont know why. # Filter the clusters to keep only the largest ones.
-                    self.valid = True
-
-                else :
-                    self.logger.info(f"Found 0 {cluster_type} clusters.")
-                    self.valid = False
+            if result is None :
+                self.logger.debug(f"Found no {cluster_type} clusters, overpass query returned no datapoints.")
+                self.valid = False
 
             else :
-                self.logger.debug(f"Detected 0 {cluster_type} clusters.")
-                self.valid = False
+                points = []
+                for elem in result:
+                    osm_type = elem.get('type')
+
+                    # Get coordinates and append them to the points list
+                    _, coords = get_base_info(elem, osm_type)
+                    if coords is not None :
+                        points.append(coords)
+
+                if points :
+                    self.all_points = np.array(points)
+
+                    # Apply DBSCAN to find clusters. Choose different settings for different cities.
+                    if self.cluster_type == 'shopping' and len(self.all_points) > 200 :
+                        dbscan = DBSCAN(eps=0.00118, min_samples=15, algorithm='kd_tree')  # for large cities
+                    elif self.cluster_type == 'sightseeing' :
+                        dbscan = DBSCAN(eps=0.0025, min_samples=15, algorithm='kd_tree')  # for historic neighborhoods
+                    else :
+                        dbscan = DBSCAN(eps=0.00075, min_samples=10, algorithm='kd_tree')  # for small cities
+
+                    labels = dbscan.fit_predict(self.all_points)
+
+                    # Check that there are is least 1 cluster
+                    if len(set(labels)) > 1 :
+                        self.logger.info(f"Found {len(set(labels))} different {cluster_type} clusters.")
+                        # Separate clustered points and noise points
+                        self.cluster_points = self.all_points[labels != -1]
+                        self.cluster_labels = labels[labels != -1]
+                        self.filter_clusters()      # ValueError here sometimes. I dont know why. # Filter the clusters to keep only the largest ones.
+                        self.valid = True
+
+                    else :
+                        self.logger.info(f"Found 0 {cluster_type} clusters.")
+                        self.valid = False
+
+                else :
+                    self.logger.debug(f"Detected 0 {cluster_type} clusters.")
+                    self.valid = False
+
+        except UnboundLocalError as ule: 
+            self.logger.warning(f"Error fetching clusters (most likely due to overpass): {e}")
+            self.valid = False
+        except Exception as e:
+            self.logger.warning(f"Error fetching clusters: {e}")
+            raise Exception from e
+
 
 
     def generate_clusters(self) -> list[Landmark]: