src/main.py

import os
import sys
from dotenv import load_dotenv

load_dotenv(override=True)

# pyrefly: ignore [missing-import]
from src.crew import build_crew


def validate_cv_path(cv_path: str) -> str:
    """Validate the CV file path exists and is a supported format."""
    if not cv_path:
        print("❌ CV path cannot be empty.")
        sys.exit(1)

    if not os.path.isfile(cv_path):
        print(f"❌ File not found: {cv_path}")
        sys.exit(1)

    if not cv_path.lower().endswith((".pdf", ".docx")):
        print(f"❌ Unsupported file format. Only .pdf and .docx are supported.")
        sys.exit(1)

    return cv_path


def validate_not_empty(value: str, field_name: str) -> str:
    """Ensure a required input is not empty."""
    if not value:
        print(f"❌ {field_name} cannot be empty.")
        sys.exit(1)
    return value


def validate_env():
    missing = [k for k in ("GROQ_API_KEY", "SERPER_API_KEY") if not os.getenv(k)]
    if missing:
        for key in missing:
            print(f"❌ Missing required environment variable: {key}")
        print("   Set them in your .env file (see .env.example).")
        sys.exit(1)


def run():
    validate_env()

    print("\n🚀 Job Application Intelligence Agent\n")
    print("=" * 40)

    cv_path = input("Enter path to your CV (PDF/DOCX): ").strip()
    cv_path = validate_cv_path(cv_path)

    jd_source = input("Enter LinkedIn job URL or paste JD text: ").strip()
    jd_source = validate_not_empty(jd_source, "Job description source")

    company_name = input("Enter company name: ").strip()
    company_name = validate_not_empty(company_name, "Company name")

    city = input("Enter company city: ").strip()
    city = validate_not_empty(city, "City")

    crew = build_crew(
        cv_path=cv_path,
        jd_source=jd_source,
        company_name=company_name,
        city=city
    )

    print("\n⏳ Starting analysis...\n")

    try:
        result = crew.kickoff()
        print("\n✅ Done! Here's your full report:\n")

        # tasks order: cv(0), jd(1), company(2), gap(3), optimization(4), outreach(5)
        sections = [
            ("📊 SKILL GAP ANALYSIS", 3),
            ("✏️  CV OPTIMIZATION SUGGESTIONS", 4),
            ("📨 OUTREACH MESSAGES", 5),
        ]
        outputs = result.tasks_output or []
        for title, idx in sections:
            if idx < len(outputs):
                print("\n" + "=" * 60)
                print(f"  {title}")
                print("=" * 60)
                print(outputs[idx].raw)

        return result
    except Exception as e:
        error_msg = str(e)

        if "rate_limit" in error_msg.lower() or "ratelimit" in error_msg.lower():
            print("\n" + "=" * 50)
            print("⛔ RATE LIMIT ERROR")
            print("=" * 50)
            print("The Groq API token limit was exceeded.")
            print("Options:")
            print("  1. Wait 15-20 minutes and try again")
            print("  2. Upgrade to Groq Dev Tier at https://console.groq.com/settings/billing")
            print("  3. Paste the JD text directly instead of a URL (uses fewer tokens)")
            print("=" * 50 + "\n")
        else:
            print("\n" + "=" * 50)
            print("❌ PIPELINE ERROR")
            print("=" * 50)
            print(f"Something went wrong: {error_msg[:300]}")
            print("=" * 50 + "\n")

        return None


if __name__ == "__main__":
    run()