setup_project.py

#!/usr/bin/env python3
"""
setup_project.py — Run this FIRST to validate your environment
and initialize the PostgreSQL database.

Usage:
    python setup_project.py
"""
import os
import sys

def check_env():
    print("🔍 Checking environment variables...")
    missing = []
    for var in ("GITHUB_TOKEN", "OPENAI_API_KEY", "DATABASE_URL"):
        if not os.getenv(var):
            missing.append(var)
    if missing:
        print(f"❌ Missing variables in .env: {', '.join(missing)}")
        print("   → Copy .env.example to .env and fill in your values.")
        sys.exit(1)
    print("✅ All required environment variables found.")


def check_db():
    print("\n🔍 Testing database connection...")
    try:
        from sqlalchemy import create_engine, text
        engine = create_engine(os.getenv("DATABASE_URL"), pool_pre_ping=True)
        with engine.connect() as conn:
            conn.execute(text("SELECT 1"))
        print("✅ Database connection successful.")
        return engine
    except Exception as e:
        print(f"❌ Cannot connect to PostgreSQL: {e}")
        print("   → Make sure PostgreSQL is running and DATABASE_URL is correct.")
        sys.exit(1)


def init_tables(engine):
    print("\n📦 Creating database tables...")
    from src.database.models import Base
    Base.metadata.create_all(engine)
    print("✅ Tables created (or already exist).")


def check_github_token():
    print("\n🔍 Testing GitHub API token...")
    import requests
    token = os.getenv("GITHUB_TOKEN")
    r = requests.get(
        "https://api.github.com/rate_limit",
        headers={"Authorization": f"token {token}"}
    )
    if r.status_code == 200:
        core = r.json()["resources"]["core"]
        print(f"✅ GitHub token valid. Rate limit: {core['remaining']}/{core['limit']}")
    else:
        print(f"❌ GitHub token invalid (status {r.status_code}). Check your GITHUB_TOKEN.")
        sys.exit(1)


def check_openai_key():
    print("\n🔍 Testing OpenAI API key...")
    try:
        from openai import OpenAI
        client = OpenAI()
        models = client.models.list()
        print("✅ OpenAI API key valid.")
    except Exception as e:
        print(f"❌ OpenAI key error: {e}")
        sys.exit(1)


def antigravity_reminder():
    print("\n🚀 Easter Egg Reminder:")
    print("   Run this in Python and take a screenshot for your submission:")
    print("   >>> import antigravity")


if __name__ == "__main__":
    from dotenv import load_dotenv
    load_dotenv()

    print("=" * 50)
    print("  GitHub Peru Analytics — Project Setup")
    print("=" * 50)

    check_env()
    engine = check_db()
    init_tables(engine)
    check_github_token()
    check_openai_key()
    antigravity_reminder()

    print("\n🎉 Setup complete! You're ready to start collecting data.")
    print("   Next step: python scripts/extract_data.py")