app.py

from fastapi import FastAPI, HTTPException,Query
from supabase import create_client, Client
from typing import List, Dict
from statistics import mean
from collections import Counter
import os
from datetime import datetime
import json

app = FastAPI()

# Initialize Supabase client
url: str = os.getenv('SUPABASE_URL')
key: str = os.getenv('SUPABASE_KEY')
supabase: Client = create_client(url, key)



@app.get('/get_travel_data')
async def get_user_travel_data(hushh_id:str):
    resp = supabase.table("receipt_radar_structured_data_duplicate").select("metadata,message_id,logo,company").eq("brand_category","Travel").eq('user_id',hushh_id).execute()
    print(resp.data)
    data = {}
    arrival_location_frequency = {}
    for msg in resp.data:
        print(msg)
        date = msg.get('Date')
        if msg.get('metadata') is not None and msg.get('metadata') != 'null':
            ds = json.loads(msg.get('metadata'))
            arrival_date = ds.get('arrival_date') or  date
            travel_type = ds.get('travel_type',None)
            if ds.get('travel_type') == 'null' or travel_type is None:
                continue
            
            departure_destination = ds.get('departure_destination') if ds.get('departure_destination') is not None else None
            arrival_destination = ds.get('arrival_destination') if ds.get('arrival_destination') is not None else None
            arrival_city = ds.get('arrival_city') if ds.get('arrival_city') is not None else None
            if 'travel_history' not in data:
                data['travel_history'] = []
            data['travel_history'].append({"message_id":msg.get('message_id'),"domain":msg.get('company'),"logo":msg.get('logo'),'arrival_date':arrival_date,'travel_type':travel_type,'departure_destination':departure_destination,'arrival_destination':arrival_destination})
            print('Data')
            print(data)
                    # Increment arrival_location_frequency count for the current arrival_destination
            if arrival_city:
                if arrival_city in arrival_location_frequency:
                    arrival_location_frequency[arrival_city] += 1
                else:
                    arrival_location_frequency[arrival_city] = 1
            sorted_arrival_location_frequency = dict(
            sorted(arrival_location_frequency.items(), key=lambda item: item[1], reverse=True)
            )
            data['arrival_location_frequency'] = sorted_arrival_location_frequency
            
    return data





@app.get("/travel-analytics")
async def get_travel_analytics(user_id: str = Query(..., description="User's hush ID")):
    try:
        # Fetch data from Supabase
        response = supabase.table("receipt_radar_structured_data_duplicate_duplicate").select("metadata, total_cost, brand_category").eq('user_id',user_id).execute()
        
        # Extract metadata from the response
        metadata_list: List[Dict] = [json.loads(row['metadata']) for row in response.data if row['metadata']]
        total_costs = [float(row['total_cost']) for row in response.data if row.get('total_cost')]
        print(metadata_list)
        print(total_costs)
        # Initialize variables for analytics
        total_trips = len(metadata_list)
        trip_durations = []
        domestic_trips = 0
        international_trips = 0
        destination_types = Counter()
        booking_lead_times = []
        accommodation_spending = []
        transport_spending = []
        activities_spending = []
        domestic_departure_countries = Counter()
        international_departure_countries = Counter()
        domestic_arrival_countries = Counter()
        international_arrival_countries = Counter()
        all_dates = []
        
        # Process each metadata entry
        for metadata in metadata_list:
            # Trip duration
            if 'check_in_date' in metadata and 'check_out_date' in metadata:
                print("inside 1st")
                check_in_date = metadata['check_in_date']
                check_out_date = metadata['check_out_date']
                duration = (check_out_date - check_in_date).days
                trip_durations.append(duration)
                all_dates.append(check_in_date)
                all_dates.append(check_out_date)
            
            # Domestic vs International
            if 'departure_country' in metadata and 'arrival_country' in metadata:
                print("inside 2nd")
                departure_country = metadata['departure_country']
                arrival_country = metadata['arrival_country']
                
                if departure_country == arrival_country:
                    print("inside 2nd")
                    domestic_trips += 1
                    domestic_departure_countries[departure_country] += 1
                    domestic_arrival_countries[arrival_country] += 1
                else:
                    international_trips += 1
                    international_departure_countries[departure_country] += 1
                    international_arrival_countries[arrival_country] += 1
            
            # Destination type (simplified)
            if 'destination_type' in metadata:
                print("inside 3rd")
                destination_types[metadata['destination_type']] += 1
            
            # Booking lead time
            if 'date_of_purchase' in metadata and 'departure_date' in metadata:
                print("inside 4th")
                lead_time = (datetime.strptime(metadata['departure_date'],"%d-%m-%Y") - datetime.strptime(metadata['date_of_purchase'],"%d-%m-%Y")).days
                booking_lead_times.append(lead_time)
            
            # Spending
            if 'accommodation_cost' in metadata:
                accommodation_spending.append(metadata['accommodation_cost'])
            if 'transport_cost' in metadata:
                transport_spending.append(metadata['transport_cost'])
            if 'activities_cost' in metadata:
                activities_spending.append(metadata['activities_cost'])
        
        # Calculate the number of years covered by the data
        if all_dates:
            min_date = min(all_dates)
            max_date = max(all_dates)
            date_range_years = (max_date - min_date).days / 365.25
        else:
            date_range_years = 1  # Default to 1 year if no dates are available
        
        # Calculate analytics
        analytics = {
            "travel_frequency": {
                "trips_per_year": total_trips / date_range_years,
                "average_trip_duration": mean(trip_durations) if trip_durations else None,
                "domestic_vs_international": f"{domestic_trips}:{international_trips}",
                "domestic_departure_countries": dict(domestic_departure_countries),
                "international_departure_countries": dict(international_departure_countries),
                "domestic_arrival_countries": dict(domestic_arrival_countries),
                "international_arrival_countries": dict(international_arrival_countries),
            },
            # "destination_preferences": {
            #     "popular_types": dict(destination_types.most_common(5))
            # },
            "booking_patterns": {
                "average_lead_time": mean(booking_lead_times) if booking_lead_times else None
            },
            "travel_expenditure": {
                "average_accommodation_cost": mean(accommodation_spending) if accommodation_spending else None,
                "average_transport_cost": mean(transport_spending) if transport_spending else None,
                "average_activities_cost": mean(activities_spending) if activities_spending else None
            }
        }
        
        return analytics
    
    except Exception as e:
        raise HTTPException(status_code=500, detail=str(e))