fakedata.py

from faker import Faker
import random
import csv

fake = Faker()

# Number of records to generate
num_records = 1000

with open('complex_sample_data.csv', mode='w', newline='') as file:
    writer = csv.writer(file)
    
    # Write the header
    writer.writerow(['user_id', 'name', 'phone_numbers', 'settings'])
    
    for _ in range(num_records):
        # Generating random user_id
        user_id = random.randint(1, 10000)
        
        # Generating a STRUCT for name
        name = {
            'first_name': fake.first_name(),
            'last_name': fake.last_name()
        }
        
        # Generating an ARRAY of phone numbers
        phone_numbers = [fake.phone_number() for _ in range(random.randint(1, 3))]
        
        # Generating a MAP for settings
        settings = {fake.word(): fake.word() for _ in range(random.randint(1, 4))}
        
        # Convert the complex structures to JSON-like strings for CSV format
        name_str = f'{{first_name: {name["first_name"]}, last_name: {name["last_name"]}}}'
        phone_numbers_str = f'[{", ".join([f"\'{phone}\'" for phone in phone_numbers])}]'
        settings_str = f'{{{", ".join([f"\'{k}\': \'{v}\'" for k, v in settings.items()])}}}'
         

        # Write the row
        writer.writerow([user_id, name_str, phone_numbers_str, settings_str])

print("Complex sample data generated and saved to 'complex_sample_data.csv'.")