import geopandas as gpd
import numpy as np

gdf=gpd.read_file('C:/PATH.geojson')

df = gdf[['number', 'street','geometry']]

df=df[['address','geometry']]

from faker import Faker
import numpy as np

fake = Faker()

df2=df.sample(10000)

class CustomProvider:
    def __init__(self, faker):
        self.faker = faker

    def gmail_email(self):
        username = self.faker.user_name()
        return f"{username}@gmail.com"

    def first_name_last_name_email(self):
        first_name = self.faker.first_name()
        last_name = self.faker.last_name()
        email = self.gmail_email()
        return f"{first_name}{last_name}{np.random.randint(0,99)}@gmail.com"

fake = Faker()
fake.add_provider(CustomProvider)
fake_email = fake.gmail_email()
fake_first_name_last_name_email = fake.first_name_last_name_email()

## I am just keeping it as reference  , but I ended up following another path

df2['email']=df2.apply(lambda x: fake.gmail_email(), axis=1)

df2=df2.drop(['email'],axis=1)

df3[['first_name', 'last_name', 'email']] = df2.apply(lambda x: pd.Series([fake.first_name(), fake.last_name(), f"{fake.first_name()}{fake.last_name()}@gmail.com"]), axis=1)

df2['first_name']=df2.apply(lambda x: fake.first_name(), axis=1)
df2['last_name']= df2.apply(lambda x: fake.last_name(), axis=1)

df2['email']=df2['first_name']+df2['last_name']+str(np.random.randint(0,99))+'@gmail.com'

df2['email']=df2.apply(lambda row: row['first_name']+row['last_name']+str(np.random.randint(0,99))+'@gmail.com', axis=1)

df2['email']=df2.apply(lambda row:row['email'].lower(), axis=1)

df3=df2

df3=df3[['first_name','last_name','email','address','geometry']]

def generate_random_time(_):
    start_time = pd.to_timedelta('10:00:00')
    end_time = pd.to_timedelta('22:00:00')
    random_seconds = np.random.randint(start_time.total_seconds(), end_time.total_seconds())
    random_timedelta = pd.to_timedelta(random_seconds, unit='s')
    return random_timedelta

df4=df3 # It is good to create a copy 

import random
df4['date']=df4.apply(lambda row: generate_random_time(row), axis=1)

df4=df4.drop(['date'],axis=1)

df4['date']=df4['date'].to_string()

df4['date'] = df4['date'].astype(str).str.replace('0 days ', '')

df4=df4.rename(columns={'date':'order_time'})

df4['order_date']=df4.apply(lambda row: generate_random_datetime2(row), axis=1)

df4['order_price']=df4.apply(lambda row: round(random.uniform(5.00, 99.99), 2), axis=1)

import requests


api_key = "{YOUR_KEY}"
address = "6739 Dawson st ,Vancouver, BC"
url = f"https://dev.virtualearth.net/REST/v1/Locations?q={address}&key={api_key}"

  response = requests.get(url)

# df4['zip_code'] = df4.apply(lambda row: requests.get(f"https://dev.virtualearth.net/REST/v1/Locations?q={row['address']}&key={api_key}").json()["resourceSets"][0]["resources"][0]["address"]["postalCode"] if requests.get(f"https://dev.virtualearth.net/REST/v1/Locations?q={row['address']}&key={api_key}").status_code == 200 else np.nan, axis=1)

df8=df4.sample(1000) # df5 tiene los zip codes ,ahora meterle el df7,ahora meterle el df8

def get_zip_code(row):
    url = f'http://dev.virtualearth.net/REST/v1/Locations/CA/BC/Vancouver/{row["address"]}?includeNeighborhood=true&key={api_key}'

    response = requests.get(url)
    if response.status_code == 200:
        data = response.json()
        resource_sets = data.get("resourceSets", [])
        if resource_sets:
            resources = resource_sets[0].get("resources", [])
            if resources:
                address = resources[0].get("address", {})
                postal_code = address.get("postalCode")
                if postal_code:
                    return postal_code

    return np.nan

df8['zip_code'] = df8.apply(get_zip_code, axis=1)

df_final=pd.concat([df5,df7,df8])

df_final=df_final.drop_duplicates()

df_final.to_csv('C:/Users/OWNER/Downloads/proyecto_base_datos/vancouver_2700_no_idx.csv', index=False)

df_final.to_csv('C:/Users/OWNER/Downloads/proyecto_base_datos/vancouver_2700_idx.csv', index=True)