import numpy as np
import pandas as pd
from datetime import datetime, timedelta

VEHICLE_IDS = ["EV-FRT-001", "EV-FRT-002", "EV-FRT-003", "EV-LM-004", "EV-LM-005",
               "EV-BUS-006", "EV-MIN-007", "EV-MIN-008", "EV-CST-009", "EV-CST-010"]

def generate_bms_history(vehicle_id: str, days: int = 365, seed: int = None) -> pd.DataFrame:
    if seed is not None:
        np.random.seed(seed)

    # Vehicle profile influences degradation
    profiles = {
        "EV-FRT-001": {"base_temp": 32, "fast_charge_freq": 0.05, "cycles_per_day": 1.2},
        "EV-FRT-002": {"base_temp": 34, "fast_charge_freq": 0.10, "cycles_per_day": 1.4},
        "EV-FRT-003": {"base_temp": 30, "fast_charge_freq": 0.03, "cycles_per_day": 1.0},
        "EV-LM-004": {"base_temp": 28, "fast_charge_freq": 0.20, "cycles_per_day": 2.0},
        "EV-LM-005": {"base_temp": 29, "fast_charge_freq": 0.15, "cycles_per_day": 1.8},
        "EV-BUS-006": {"base_temp": 35, "fast_charge_freq": 0.02, "cycles_per_day": 1.0},
        "EV-MIN-007": {"base_temp": 38, "fast_charge_freq": 0.08, "cycles_per_day": 1.6},
        "EV-MIN-008": {"base_temp": 40, "fast_charge_freq": 0.12, "cycles_per_day": 1.7},
        "EV-CST-009": {"base_temp": 33, "fast_charge_freq": 0.06, "cycles_per_day": 1.3},
        "EV-CST-010": {"base_temp": 31, "fast_charge_freq": 0.04, "cycles_per_day": 1.1},
    }
    profile = profiles.get(vehicle_id, profiles["EV-FRT-001"])

    records = []
    start = datetime.utcnow() - timedelta(days=days)
    cycle_count = 0

    for i in range(days):
        timestamp = start + timedelta(days=i)
        # Daily reading (one representative point per day)
        temp = profile['base_temp'] + np.random.normal(0, 3)
        if np.random.random() < 0.02:
            temp += np.random.uniform(8, 18)  # thermal event

        cycle_count += profile['cycles_per_day'] + np.random.normal(0, 0.1)
        soc = np.random.uniform(30, 90)

        # Fast charge event
        is_fast_charge = np.random.random() < profile['fast_charge_freq']
        current = np.random.uniform(-80, 120)
        if is_fast_charge:
            current = np.random.uniform(200, 320)

        voltage = 350 + (100 - soc) * 0.8 + np.random.normal(0, 2)

        records.append({
            "timestamp": timestamp,
            "vehicle_id": vehicle_id,
            "soc": round(soc, 1),
            "voltage": round(voltage, 1),
            "current": round(current, 1),
            "temperature": round(temp, 1),
            "cycle_count": int(cycle_count),
            "odometer_km": round(cycle_count * np.random.uniform(80, 120), 1)
        })

    return pd.DataFrame(records)


def get_all_bms_data() -> dict:
    data = {}
    for vid in VEHICLE_IDS:
        data[vid] = generate_bms_history(vid, days=365, seed=hash(vid) % 10000)
    return data