retour api fix

Author: rayaberova7

app/main.py

@@ -5,7 +5,7 @@
 import gc
 import os
 from contextlib import asynccontextmanager
-from typing import Dict
+from typing import Dict, List
 
 import geopandas as gpd
 import mlflow
@@ -17,7 +17,6 @@
 from pyproj import Transformer
 from shapely.geometry import box
 from shapely.geometry import Point
-import requests
 
 from app.logger_config import configure_logger
 from app.utils import (
@@ -94,53 +93,49 @@ def show_welcome_page():
 
 @app.get("/find_image", tags=["Find Image"])
 async def find_image(
-    lon_gps: float,
-    lat_gps: float,
-    nuts_id: str,
+    gps_point: List[float] = Query(..., description="[longitude, latitude] in WGS84"),
+    nuts_id: str = Query(...),
     year: int = Query(2021, ge=2018, le=2024)
 ) -> str:
     """
     Find image path for a given NUTS3 and year.
 
     Args:
-        lon_gps (float): longitude of the gps point.
-        lat_gps (float): longitude of the gps point.
+        gps_point (List[float]): [longitude, latitude] of the GPS point in WGS84.
         nuts_id (str): The ID of the NUTS.
         year (int): The year of the satellite images.
     Returns:
-        str: Image filepath if the image is finded, otherwise None.
-
+        str: Image filepath if found, otherwise empty string.
     """
+    lon_gps, lat_gps = gps_point[0], gps_point[1]
     logger.info(f"Find the image filepath for this gps point: {lon_gps}, {lat_gps}")
     gc.collect()
 
     url = f"https://minio.lab.sspcloud.fr/projet-formation/diffusion/funathon/2026/project3/data/images/{nuts_id}/{year}/filename2bbox.parquet"
 
-    response = requests.head(url)
-
-    if response.status_code == 200:
+    try:
         df = pd.read_parquet(url)
-    else:
-        print(f"❌ No data for NUTS3='{nuts_id}' and year={year} (HTTP {response.status_code})")
-        return None
+    except Exception:
+        print(f"❌ No data for NUTS3='{nuts_id}' and year={year}")
+        return ""
 
-    # Création de la géométrie
     df["geometry"] = df.apply(
         lambda row: box(row["bbox"][0], row["bbox"][1], row["bbox"][2], row["bbox"][3]),
         axis=1
     )
 
-    # Conversion en GeoDataFrame
     gdf = gpd.GeoDataFrame(df, geometry="geometry", crs="EPSG:3035")
 
-    # Convertir le point GPS (EPSG:4326) en EPSG:3035
     transformer = Transformer.from_crs("EPSG:4326", "EPSG:3035", always_xy=True)
     x, y = transformer.transform(lon_gps, lat_gps)
 
     point = Point(x, y)
     result = gdf[gdf.contains(point)]
 
-    return result["filename"].values
+    if result.empty:
+        return ""
+
+    return str(result["filename"].values[0])
 
 
 @app.get("/predict_image", tags=["Predict Image"])