Update app.py

app.py

@@ -1,4 +1,4 @@
-from smolagents import CodeAgent,DuckDuckGoSearchTool, HfApiModel,load_tool,tool
+from smolagents import CodeAgent, DuckDuckGoSearchTool, HfApiModel, load_tool, tool
 import datetime
 import requests
 import pytz
@@ -10,6 +10,7 @@ from typing import Optional, Dict, Any, List
 import requests
 from pydantic import BaseModel, Field, field_validator
 from typing import Optional, Union
+import functools
 
 class PDOKLocationSearchInput(BaseModel):
     postal_code: Optional[str] = Field(None, description="Postal code in the format '1234 AA'.")
@@ -22,7 +23,7 @@ class PDOKLocationSearchInput(BaseModel):
         if v is not None and (len(v) != 7 or not v[0:4].isdigit() or v[4] != " " or not v[5:7].isalpha()):
             raise ValueError("Invalid postal code format. It must be '1234 AA'.")
         return v
-           
+
     def construct_query(self) -> str:
         """Constructs the query string based on provided inputs."""
         if self.postal_code and self.house_number:
@@ -32,27 +33,164 @@ class PDOKLocationSearchInput(BaseModel):
         else:
             return ""
 
-class PDOKLocationInfo(BaseModel):  # This is not used in the tool's return anymore
-    type: str
-    weergavenaam: str
-    straatnaam: Optional[str] = None
-    huisnummer: Optional[int] = None
-    huisletter: Optional[str] = None
-    huisnummertoevoeging: Optional[str] = None
-    postcode: Optional[str] = None
-    woonplaatsnaam: Optional[str] = None
-    gemeentenaam: Optional[str] = None
-    provincienaam: Optional[str] = None
-    centroide_ll: Optional[str] = None
-    centroide_rd: Optional[str] = None
-
-    class Config:
-        extra = "allow"  # This will allow extra fields from the API response
+# Create a requests session for reuse
+session = requests.Session()
+
+@functools.lru_cache(maxsize=128)
+def _fetch_cbs_data(postal_code: str, house_number: str, year: str):
+    """Helper function to fetch CBS data for a specific year (with caching)."""
+    url = f"https://service.pdok.nl/cbs/wijkenbuurten/2021/wfs?request=GetFeature&service=WFS&version=1.1.0&typeName=cbs_buurten_2021&outputFormat=json&srsName=EPSG:4326&CQL_FILTER=(postcode='{postal_code}')AND(huisnummer={house_number})"
+    try:
+        response = session.get(url)
+        response.raise_for_status()  # Raise HTTPError for bad responses (4xx or 5xx)
+        data = response.json()
+        # Extract relevant data - adapt field names as needed
+        if data["features"]:
+            return data["features"][0]["properties"]
+        else:
+            return None  # Or raise a custom exception if no features are found
+
+    except requests.exceptions.RequestException as e:
+        print(f"Error fetching CBS data for {year}: {e}")  # Log the error
+        return None
+    except (ValueError, KeyError) as e:
+         print(f"Error parsing CBS data for {year}: {e}")
+         return None
+
+@tool
+def get_cbs_data(postal_code: str, house_number: str) -> str:
+    """
+    Retrieves CBS (Statistics Netherlands) demographic and facility data for a given postal code and house number.
+    Fetches data for 2021.
+
+    Args:
+        postal_code: The postal code (e.g., "1234AB").
+        house_number: The house number (e.g., "12").
+
+    Returns:
+        A JSON string containing the CBS data, or an error message if the data cannot be retrieved.
+    """
+    # Format inputs
+    formatted_postal_code = postal_code.replace(" ", "").upper()
+    formatted_house_number = house_number.strip()
+
+    data_2021 = _fetch_cbs_data(formatted_postal_code, formatted_house_number, "2021")
+
+    if data_2021:
+        #  Create the dictionary to match the desired output format, including *all* fields, handling potential missing data.
+        cbs_data = {
+            "p2021_amount_of_inhabitants": data_2021.get("aantal_inwoners"),
+            "p2021_amount_of_men": data_2021.get("aantal_mannen"),
+            "p2021_amount_of_women": data_2021.get("aantal_vrouwen"),
+            "p2021_amount_of_inhabitants_0_to_15_year": data_2021.get("aantal_inwoners_0_tot_15_jaar"),
+            "p2021_amount_of_inhabitants_15_to_25_year": data_2021.get("aantal_inwoners_15_tot_25_jaar"),
+            "p2021_amount_of_inhabitants_25_to_45_year": data_2021.get("aantal_inwoners_25_tot_45_jaar"),
+            "p2021_amount_of_inhabitants_45_to_65_year": data_2021.get("aantal_inwoners_45_tot_65_jaar"),
+            "p2021_amount_of_inhabitants_65_years_and_older": data_2021.get("aantal_inwoners_65_jaar_en_ouder"),
+            "p2021_nearest_supermarket_distance": data_2021.get("dichtstbijzijnde_grote_supermarkt_afstand_in_km"),
+            "p2021_supermarkets_within_1km": data_2021.get("grote_supermarkt_aantal_binnen_1_km"),
+            "p2021_supermarkets_within_3km": data_2021.get("grote_supermarkt_aantal_binnen_3_km"),
+            "p2021_supermarkets_within_5km": data_2021.get("grote_supermarkt_aantal_binnen_5_km"),
+            "p2021_nearest_daily_grocery_store_distance": data_2021.get("dichtstbijzijnde_winkels_ov_dagel_levensm_afst_in_km"),
+            "p2021_daily_grocery_stores_within_1km": data_2021.get("winkels_ov_dagel_levensm_aantal_binnen_1_km"),
+            "p2021_daily_grocery_stores_within_3km": data_2021.get("winkels_ov_dagel_levensm_aantal_binnen_3_km"),
+            "p2021_daily_grocery_stores_within_5km": data_2021.get("winkels_ov_dagel_levensm_aantal_binnen_5_km"),
+            "p2021_nearest_department_store_distance": data_2021.get("dichtstbijzijnde_warenhuis_afstand_in_km"),
+            "p2021_department_stores_within_5km": data_2021.get("warenhuis_aantal_binnen_5_km"),
+            "p2021_department_stores_within_10km": data_2021.get("warenhuis_aantal_binnen_10_km"),
+            "p2021_department_stores_within_20km": data_2021.get("warenhuis_aantal_binnen_20_km"),
+            "p2021_nearest_pub_distance": data_2021.get("dichtstbijzijnde_cafe_afstand_in_km"),
+            "p2021_pubs_within_1km": data_2021.get("cafe_aantal_binnen_1_km"),
+            "p2021_pubs_within_3km": data_2021.get("cafe_aantal_binnen_3_km"),
+            "p2021_pubs_within_5km": data_2021.get("cafe_aantal_binnen_5_km"),
+            "p2021_nearest_cafeteria_distance": data_2021.get("dichtstbijzijnde_cafetaria_afstand_in_km"),
+            "p2021_cafeterias_within_1km": data_2021.get("cafetaria_aantal_binnen_1_km"),
+            "p2021_cafeterias_within_3km": data_2021.get("cafetaria_aantal_binnen_3_km"),
+            "p2021_cafeterias_within_5km": data_2021.get("cafetaria_aantal_binnen_5_km"),
+            "p2021_nearest_hotel_distance": data_2021.get("dichtstbijzijnde_hotel_afstand_in_km"),
+            "p2021_hotels_within_5km": data_2021.get("hotel_aantal_binnen_5_km"),
+            "p2021_hotels_within_10km": data_2021.get("hotel_aantal_binnen_10_km"),
+            "p2021_hotels_within_20km": data_2021.get("hotel_aantal_binnen_20_km"),
+            "p2021_nearest_restaurant_distance": data_2021.get("dichtstbijzijnde_restaurant_afstand_in_km"),
+            "p2021_restaurants_within_1km": data_2021.get("restaurant_aantal_binnen_1_km"),
+            "p2021_restaurants_within_3km": data_2021.get("restaurant_aantal_binnen_3_km"),
+            "p2021_restaurants_within_5km": data_2021.get("restaurant_aantal_binnen_5_km"),
+            "p2021_nearest_after_school_care_distance": data_2021.get("dichtstbijzijnde_buitenschoolse_opvang_afstand_in_km"),
+            "p2021_after_school_care_within_1km": data_2021.get("buitenschoolse_opvang_aantal_binnen_1_km"),
+            "p2021_after_school_care_within_3km": data_2021.get("buitenschoolse_opvang_aantal_binnen_3_km"),
+            "p2021_after_school_care_within_5km": data_2021.get("buitenschoolse_opvang_aantal_binnen_5_km"),
+            "p2021_nearest_day_care_distance": data_2021.get("dichtstbijzijnde_kinderdagverblijf_afstand_in_km"),
+            "p2021_day_care_within_1km": data_2021.get("kinderdagverblijf_aantal_binnen_1_km"),
+            "p2021_day_care_within_3km": data_2021.get("kinderdagverblijf_aantal_binnen_3_km"),
+            "p2021_day_care_within_5km": data_2021.get("kinderdagverblijf_aantal_binnen_5_km"),
+            "p2021_nearest_fire_station_distance": data_2021.get("dichtstbijzijnde_brandweerkazerne_afstand_in_km"),
+            "p2021_nearest_highway_access_distance": data_2021.get("dichtstbijzijnde_oprit_hoofdverkeersweg_afstand_in_km"),
+            "p2021_nearest_transfer_station_distance": data_2021.get("dichtstbijzijnde_overstapstation_afstand_in_km"),
+            "p2021_nearest_train_station_distance": data_2021.get("dichtstbijzijnde_treinstation_afstand_in_km"),
+            "p2021_nearest_theme_park_distance": data_2021.get("dichtstbijzijnde_attractiepark_afstand_in_km"),
+            "p2021_theme_parks_within_10km": data_2021.get("attractiepark_aantal_binnen_10_km"),
+            "p2021_theme_parks_within_20km": data_2021.get("attractiepark_aantal_binnen_20_km"),
+            "p2021_theme_parks_within_50km": data_2021.get("attractiepark_aantal_binnen_50_km"),
+            "p2021_nearest_cinema_distance": data_2021.get("dichtstbijzijnde_bioscoop_afstand_in_km"),
+            "p2021_cinemas_within_5km": data_2021.get("bioscoop_aantal_binnen_5_km"),
+            "p2021_cinemas_within_10km": data_2021.get("bioscoop_aantal_binnen_10_km"),
+            "p2021_cinemas_within_20km": data_2021.get("bioscoop_aantal_binnen_20_km"),
+            "p2021_nearest_museum_distance": data_2021.get("dichtstbijzijnde_museum_afstand_in_km"),
+            "p2021_museums_within_5km": data_2021.get("museum_aantal_binnen_5_km"),
+            "p2021_museums_within_10km": data_2021.get("museum_aantal_binnen_10_km"),
+            "p2021_museums_within_20km": data_2021.get("museum_aantal_binnen_20_km"),
+            "p2021_nearest_theater_distance": data_2021.get("dichtstbijzijnde_theater_afstand_in_km"),
+            "p2021_theaters_within_5km": data_2021.get("theater_aantal_binnen_5_km"),
+            "p2021_theaters_within_10km": data_2021.get("theater_aantal_binnen_10_km"),
+            "p2021_theaters_within_20km": data_2021.get("theater_aantal_binnen_20_km"),
+"p2021_nearest_library_distance": data_2021.get("dichtstbijzijnde_bibliotheek_afstand_in_km"),
+            "p2021_nearest_ice_rink_distance": data_2021.get("dichtstbijzijnde_kunstijsbaan_afstand_in_km"),
+            "p2021_nearest_music_venue_distance": data_2021.get("dichtstbijzijnde_poppodium_afstand_in_km"),
+            "p2021_nearest_sauna_distance": data_2021.get("dichtstbijzijnde_sauna_afstand_in_km"),
+            "p2021_nearest_tanning_salon_distance": data_2021.get("dichtstbijzijnde_zonnebank_afstand_in_km"),
+            "p2021_nearest_swimming_pool_distance": data_2021.get("dichtstbijzijnde_zwembad_afstand_in_km"),
+            "p2021_nearest_primary_school_distance": data_2021.get("dichtstbijzijnde_basisonderwijs_afstand_in_km"),
+            "p2021_primary_schools_within_1km": data_2021.get("basisonderwijs_aantal_binnen_1_km"),
+            "p2021_primary_schools_within_3km": data_2021.get("basisonderwijs_aantal_binnen_3_km"),
+            "p2021_primary_schools_within_5km": data_2021.get("basisonderwijs_aantal_binnen_5_km"),
+            "p2021_nearest_havovwo_distance": data_2021.get("dichtstbijzijnde_havo_vwo_afstand_in_km"),
+            "p2021_havovwo_within_3km": data_2021.get("havo_vwo_aantal_binnen_3_km"),
+            "p2021_havovwo_within_5km": data_2021.get("havo_vwo_aantal_binnen_5_km"),
+            "p2021_havovwo_within_10km": data_2021.get("havo_vwo_aantal_binnen_10_km"),
+            "p2021_nearest_vmbo_distance": data_2021.get("dichtstbijzijnde_vmbo_afstand_in_km"),
+            "p2021_vmbo_within_3km": data_2021.get("vmbo_aantal_binnen_3_km"),
+            "p2021_vmbo_within_5km": data_2021.get("vmbo_aantal_binnen_5_km"),
+            "p2021_vmbo_within_10km": data_2021.get("vmbo_aantal_binnen_10_km"),
+            "p2021_nearest_secondary_school_distance": data_2021.get("dichtstbijzijnde_voortgezet_onderwijs_afstand_in_km"),
+            "p2021_secondary_schools_within_3km": data_2021.get("voortgezet_onderwijs_aantal_binnen_3_km"),
+            "p2021_secondary_schools_within_5km": data_2021.get("voortgezet_onderwijs_aantal_binnen_5_km"),
+            "p2021_secondary_schools_within_10km": data_2021.get("voortgezet_onderwijs_aantal_binnen_10_km"),
+            "p2021_nearest_gp_distance": data_2021.get("dichtstbijzijnde_huisartsenpraktijk_afstand_in_km"),
+            "p2021_gps_within_1km": data_2021.get("huisartsenpraktijk_aantal_binnen_1_km"),
+            "p2021_gps_within_3km": data_2021.get("huisartsenpraktijk_aantal_binnen_3_km"),
+            "p2021_gps_within_5km": data_2021.get("huisartsenpraktijk_aantal_binnen_5_km"),
+            "p2021_nearest_hospital_excl_outpatient_distance": data_2021.get("dichtstbijzijnde_ziekenh_excl_buitenpoli_afst_in_km"),
+            "p2021_hospitals_excl_outpatient_within_5km": data_2021.get("ziekenhuis_excl_buitenpoli_aantal_binnen_5_km"),
+            "p2021_hospitals_excl_outpatient_within_10km": data_2021.get("ziekenhuis_excl_buitenpoli_aantal_binnen_10_km"),
+            "p2021_hospitals_excl_outpatient_within_20km": data_2021.get("ziekenhuis_excl_buitenpoli_aantal_binnen_20_km"),
+            "p2021_nearest_hospital_incl_outpatient_distance": data_2021.get("dichtstbijzijnde_ziekenh_incl_buitenpoli_afst_in_km"),
+            "p2021_hospitals_incl_outpatient_within_5km": data_2021.get("ziekenhuis_incl_buitenpoli_aantal_binnen_5_km"),
+            "p2021_hospitals_incl_outpatient_within_10km": data_2021.get("ziekenhuis_incl_buitenpoli_aantal_binnen_10_km"),
+            "p2021_hospitals_incl_outpatient_within_20km": data_2021.get("ziekenhuis_incl_buitenpoli_aantal_binnen_20_km"),
+            "p2021_nearest_pharmacy_distance": data_2021.get("dichtstbijzijnde_apotheek_afstand_in_km"),
+            "p2021_nearest_gp_post_distance": data_2021.get("dichtstbijzijnde_huisartsenpost_afstand_in_km"),
+        }
+        return json.dumps(cbs_data)  # Return as JSON string
+
+    else:
+        return json.dumps({"error": "Could not retrieve CBS data."})
+
 
+@functools.lru_cache(maxsize=128)  # Add caching
 @tool
 def pdok_location_info(postal_code: Optional[str] = None, house_number: Optional[str] = None, street_name: Optional[str] = None, city: Optional[str] = None) -> str:
-    """Provides information about a Dutch address or postal code, including a Google Maps link.
-    
+    """Provides information about a Dutch address or postal code, including a Google Maps link and CBS data.
+
     Args:
         postal_code: Postal code in the format '1234 AA'.
         house_number: House number of the address.
@@ -60,19 +198,19 @@ def pdok_location_info(postal_code: Optional[str] = None, house_number: Optional
         city: Name of the city.
 
     Returns:
-        str: JSON string containing the location information, including a Google Maps link, or an error message.
+        str: JSON string containing the location information, including a Google Maps link and CBS data, or an error message.
     """
-    
+
     debug_info = []
     debug_info.append(f"Input values:\n"
                       f"  Postal code: {postal_code}\n"
                       f"  House number: {house_number}\n"
                       f"  Street name: {street_name}\n"
                       f"  City: {city}")
-    
+
     base_url = "https://api.pdok.nl/bzk/locatieserver/search/v3_1/free"
     headers = {"accept": "application/json"}
-    
+
     input_data = PDOKLocationSearchInput(postal_code=postal_code, house_number=house_number, street_name=street_name, city=city)
     query_string = input_data.construct_query()
 
@@ -84,35 +222,36 @@ def pdok_location_info(postal_code: Optional[str] = None, house_number: Optional
     params = {
         "q": query_string,
         "fl": "*",
-        "fq": "type:(gemeente OR woonplaats OR postcode OR adres)",
+        "fq": "type:(gemeente OR woonplaats OR weg OR postcode OR adres)",
         "df": "tekst",
+        "bq": "type:provincie^1.5",
         "bq": "type:gemeente^1.5",
         "bq": "type:woonplaats^1.5",
+        "bq": "type:weg^1.5",
         "bq": "type:postcode^0.5",
         "bq": "type:adres^1",
         "start": 0,
-        "rows": 1,
+        "rows": 1,  # Reduced to 1
         "sort": "score desc,sortering asc,weergavenaam asc",
         "wt": "json",
     }
-    
+
     try:
-        response = requests.get(base_url, params=params, headers=headers)
+        response = session.get(base_url, params=params, headers=headers)  # Use the session
         debug_info.append(f"Response status code: {response.status_code}")
-        
+
         response.raise_for_status()
         data = response.json()
-        
+
         docs = data.get("response", {}).get("docs", [])
         debug_info.append(f"Number of docs found: {len(docs)}")
-        
+
         if not docs:
             return f"Error: No results found\nDebug info:\n" + "\n".join(debug_info)
 
         first_result = docs[0]
         debug_info.append(f"First result data: {first_result}")
-        
-        # Create a dictionary with only the fields we need, *including* the Google Maps link
+
         location_info = {
             "straatnaam": first_result.get("straatnaam", ""),
             "huisnummer": first_result.get("huisnummer", ""),
@@ -126,24 +265,38 @@ def pdok_location_info(postal_code: Optional[str] = None, house_number: Optional
             "centroide_rd": first_result.get("centroide_rd", "")
         }
 
-        # Add Google Maps link *to the dictionary that will be returned*
         centroide_ll = location_info.get('centroide_ll')
         if centroide_ll:
             lon, lat = centroide_ll.replace("POINT(", "").replace(")", "").split()
             location_info["google_maps_url"] = f"https://www.google.com/maps/search/?api=1&query={lat},{lon}"
-        
-        return json.dumps(location_info)  # Return the dictionary as a JSON string
-        
+
+        # --- Get CBS Data ---
+        if location_info["postcode"] and location_info["huisnummer"]:
+            cbs_data_str = get_cbs_data(location_info["postcode"], str(location_info["huisnummer"]))
+            try:
+                cbs_data = json.loads(cbs_data_str)
+                # Check for error in CBS data
+                if "error" in cbs_data:
+                    location_info["cbs_data_error"] = cbs_data["error"] # Add error message.
+                else:
+                    location_info.update(cbs_data)  # Merge CBS data
+            except json.JSONDecodeError:
+                location_info["cbs_data_error"] = "Failed to parse CBS data."
+        else:
+             location_info["cbs_data_error"] = "Postcode or house number missing, cannot fetch CBS data."
+
+
+        return json.dumps(location_info)
+
     except requests.exceptions.RequestException as e:
         return f"Error during API request: {e}\nDebug info:\n" + "\n".join(debug_info)
     except (ValueError, KeyError) as e:
         return f"Error processing API response: {e}\nDebug info:\n" + "\n".join(debug_info)
-
-# ---  Rest of your code (agent definition, etc.) remains the same ---
+    except Exception as e:
+        return f"An unexpected error occurred: {e}\nDebug info:\n" + "\n".join(debug_info)
 
 @tool
-def my_custom_tool(arg1:str, arg2:int)-> str: #it's import to specify the return type
-    #Keep this format for the description / args / args description but feel free to modify the tool
+def my_custom_tool(arg1:str, arg2:int)-> str:
     """A tool that does nothing yet 
     Args:
         arg1: the first argument
@@ -158,39 +311,31 @@ def get_current_time_in_timezone(timezone: str) -> str:
         timezone: A string representing a valid timezone (e.g., 'America/New_York').
     """
     try:
-        # Create timezone object
         tz = pytz.timezone(timezone)
-        # Get current time in that timezone
         local_time = datetime.datetime.now(tz).strftime("%Y-%m-%d %H:%M:%S")
         return f"The current local time in {timezone} is: {local_time}"
     except Exception as e:
         return f"Error fetching time for timezone '{timezone}': {str(e)}"
 
-
 final_answer = FinalAnswerTool()
 
-# If the agent does not answer, the model is overloaded, please use another model or the following Hugging Face Endpoint that also contains qwen2.5 coder:
-# model_id='https://pflgm2locj2t89co.us-east-1.aws.endpoints.huggingface.cloud' 
-
 model = HfApiModel(
-max_tokens=2096,
-temperature=0.5,
-model_id='Qwen/Qwen2.5-Coder-32B-Instruct',# it is possible that this model may be overloaded
-custom_role_conversions=None,
+    max_tokens=2096,
+    temperature=0.5,
+    model_id='Qwen/Qwen2.5-Coder-32B-Instruct',
+    custom_role_conversions=None,
 )
 
-
-# Import tool from Hub
 image_generation_tool = load_tool("agents-course/text-to-image", trust_remote_code=True)
 
 with open("prompts.yaml", 'r') as stream:
     prompt_templates = yaml.safe_load(stream)
-    
+
 agent = CodeAgent(
     model=model,
-    tools=[final_answer, get_current_time_in_timezone, pdok_location_info], ## add your tools here (don't remove final answer)
-    max_steps=3,
-    verbosity_level=1,
+    tools=[final_answer, get_current_time_in_timezone, pdok_location_info, get_cbs_data],  # Include get_cbs_data
+    max_steps=5,  # Tune this value! Start low.
+    verbosity_level=0,  # Set to 0 for production
     grammar=None,
     planning_interval=None,
     name=None,
@@ -198,5 +343,4 @@ agent = CodeAgent(
     prompt_templates=prompt_templates
 )
 
-
 GradioUI(agent).launch()