collect_all_data.py

import os
import subprocess
import sys
import time

def run_script(script_name):
    """Run a Python script and return its exit code."""
    print(f"\n{'='*50}")
    print(f"Running {script_name}...")
    print(f"{'='*50}\n")
    
    result = subprocess.run([sys.executable, script_name])
    return result.returncode

def main():
    print("\n=== Schoolify Comprehensive Data Collection ===\n")
    print("This script will run all data collection steps in sequence to build")
    print("a complete database of Victorian schools and suburbs.\n")
    
    # Define the scripts to run in order
    scripts = [
        "download_postcodes.py",
        "fetch_vic_govt_schools.py",
        "comprehensive_scraper.py",
        "update_database.py"
    ]
    
    # Run each script in sequence
    for script in scripts:
        if not os.path.exists(script):
            print(f"Error: Script {script} not found!")
            return
        
        exit_code = run_script(script)
        
        if exit_code != 0:
            print(f"\nError: Script {script} failed with exit code {exit_code}")
            print("Stopping the data collection process.")
            return
        
        # Add a short pause between scripts
        time.sleep(1)
    
    # Run the count script to show the results
    print("\n=== Data Collection Complete ===\n")
    print("Running count_database_entries.py to show the results...\n")
    run_script("count_database_entries.py")
    
    print("\n=== All Done! ===\n")
    print("Your Schoolify database now contains comprehensive data for Victorian schools and suburbs.")
    print("You can view the data in:")
    print("- data/school_data_final.json (Schools)")
    print("- js/geocode-db.js (Suburbs)")

if __name__ == "__main__":
    main()