"""Loan eligibility without heavy ML deps (safe for HF Docker)."""


def calculate_emi(principal: float, rate_percent: float, tenure_years: int) -> float:
    if rate_percent <= 0:
        months = tenure_years * 12
        return principal / months if months else principal
    monthly_rate = rate_percent / 100 / 12
    months = tenure_years * 12
    if monthly_rate == 0:
        return principal / months
    return principal * monthly_rate * ((1 + monthly_rate) ** months) / (((1 + monthly_rate) ** months) - 1)


def analyze_loan(
    salary: float,
    credit_score: int,
    existing_loans: int,
    employment_years: float,
    age: int,
    loan_amount: float,
) -> dict:
    issues: list[str] = []
    score = 50.0

    if credit_score >= 750:
        score += 20
    elif credit_score >= 700:
        score += 12
    elif credit_score >= 650:
        score += 5
    else:
        score -= 10
        issues.append("Credit score below 650 — consider improving before applying")

    if employment_years >= 3:
        score += 10
    elif employment_years < 1:
        score -= 8
        issues.append("Less than 1 year employment — higher scrutiny expected")

    if age < 21 or age > 65:
        issues.append("Age outside typical lending range (21–65)")

    if existing_loans > 3:
        score -= 12
        issues.append("Multiple existing loans may reduce approval odds")

    emi = calculate_emi(loan_amount, 12, 10)
    monthly_salary = salary / 12
    emi_ratio = (emi / monthly_salary * 100) if monthly_salary > 0 else 100

    if emi_ratio > 50:
        score -= 15
        issues.append(f"EMI is {emi_ratio:.0f}% of monthly income — reduce loan amount")
    elif emi_ratio < 35:
        score += 8

    if loan_amount > salary * 5:
        score -= 10
        issues.append("Loan amount exceeds 5× annual salary")

    score = max(0, min(100, score))
    approval_probability = score

    if approval_probability >= 70:
        status = "APPROVED"
        risk = "Low"
    elif approval_probability >= 45:
        status = "UNDER REVIEW"
        risk = "Medium"
    else:
        status = "LIKELY REJECTED"
        risk = "High"

    recommendations = []
    if approval_probability >= 70:
        recommendations.append("Strong profile — you are likely to qualify at competitive rates")
    elif approval_probability < 45:
        recommendations.append("Consider a smaller loan or improving credit score before applying")
    if emi_ratio < 35:
        recommendations.append(f"Healthy EMI ratio ({emi_ratio:.0f}% of monthly income)")
    elif emi_ratio > 40:
        recommendations.append("Try a longer tenure or lower amount to reduce monthly EMI")

    comparison = []
    for rate in (9, 10, 11, 12):
        for tenure in (5, 7, 10):
            e = calculate_emi(loan_amount, rate, tenure)
            total = e * 12 * tenure
            comparison.append({
                "rate": f"{rate}%",
                "tenure": f"{tenure} years",
                "emi": round(e, 2),
                "total_amount": round(total, 2),
                "interest": round(total - loan_amount, 2),
            })

    return {
        "approval_probability": round(approval_probability, 1),
        "approval_status": status,
        "risk_level": risk,
        "loan_score": round(score, 1),
        "emi": round(emi, 2),
        "monthly_emi": round(emi, 2),
        "issues": issues,
        "recommendations": recommendations,
        "comparison": comparison[:12],
    }