import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import io
from PIL import Image   # ⭐ REQUIRED FOR IMAGE FIX


# ============================================================
#  SLOTING OPTIMIZATION — PHASE 1
# ============================================================

def run_slotting_analysis(message, slotting_df):
    reasoning_steps = []

    df = slotting_df.copy()

    velocity_map = {
        "fast": 3,
        "medium": 2,
        "slow": 1
    }

    df["VelocityNorm"] = df["Velocity"].str.lower().map(velocity_map)
    reasoning_steps.append("Mapped velocity categories to numerical weights.")

    df["FreqNorm"] = (df["Frequency"] - df["Frequency"].min()) / (
        df["Frequency"].max() - df["Frequency"].min() + 1e-8
    )
    reasoning_steps.append("Normalized frequency to 0–1 scale.")

    df["Score"] = (0.6 * df["VelocityNorm"]) + (0.4 * df["FreqNorm"])
    reasoning_steps.append("Computed weighted slotting score.")

    df = df.sort_values("Score", ascending=False).reset_index(drop=True)

    df["Aisle"] = np.arange(1, len(df) + 1)
    df["Rack"] = np.linspace(1, 20, len(df)).astype(int)
    reasoning_steps.append("Assigned optimal aisle & rack positions.")

    explanation = (
        "### 📦 Slotting Optimization\n"
        "High-velocity & high-frequency SKUs placed in prime aisles.\n\n"
        "#### 🔍 Reasoning\n" + "\n".join([f"- {r}" for r in reasoning_steps])
    )

    return explanation, df


# ============================================================
#  PICKING ROUTE OPTIMIZATION — PHASE 1
# ============================================================

def run_picking_optimization(message, picking_df):
    reasoning_steps = []

    df = picking_df.copy()

    df["x"] = df["Aisle"]
    df["y"] = df["Rack"]
    reasoning_steps.append("Converted Aisle–Rack to coordinate grid.")

    df["Distance"] = df["x"].abs() + df["y"].abs()
    df = df.sort_values("Distance").reset_index(drop=True)
    reasoning_steps.append("Calculated Manhattan distance & sorted sequence.")

    # --- Generate plot ---
    plt.figure(figsize=(6, 6))
    plt.plot(df["x"], df["y"], marker="o", linestyle="-")
    plt.title("Optimized Picking Route")
    plt.xlabel("Aisle")
    plt.ylabel("Rack")

    buffer = io.BytesIO()
    plt.savefig(buffer, format="png")
    plt.close()
    buffer.seek(0)

    # ⭐ Convert BytesIO → PIL (Gradio requirement)
    image = Image.open(buffer)

    reasoning_steps.append("Generated walking route visualization.")

    explanation = (
        "### 🚚 Picking Route Optimization\n"
        "Manhattan-distance-based walk path generated.\n\n"
        "#### 🔍 Reasoning\n" + "\n".join([f"- {r}" for r in reasoning_steps])
    )

    return explanation, image


# ============================================================
#  DEMAND FORECASTING — MODULE 1
# ============================================================

def run_demand_forecast(message, slotting_df):
    reasoning_steps = []
    df = slotting_df.copy()

    if "Frequency" not in df.columns:
        return "Frequency missing — cannot forecast.", None, None

    demand = df["Frequency"].astype(float)
    reasoning_steps.append("Used SKU picking frequency as demand signal.")

    moving_avg = demand.mean()
    reasoning_steps.append(f"Computed moving average: {moving_avg:.2f}")

    weights = np.linspace(0.1, 1.0, len(demand))
    trend = np.sum(demand * weights) / np.sum(weights)
    reasoning_steps.append(f"Weighted trend adjustment: {trend:.2f}")

    forecast_value = (moving_avg * 0.6) + (trend * 0.4)

    next_7_days = [forecast_value * (1 + 0.05 * i) for i in range(7)]
    forecast_df = pd.DataFrame({
        "Day": [f"Day {i+1}" for i in range(7)],
        "Forecasted_Demand": next_7_days
    })

    reasoning_steps.append("Generated 7-day demand projection.")

    # --- Plot forecast ---
    plt.figure(figsize=(6, 4))
    plt.plot(forecast_df["Day"], forecast_df["Forecasted_Demand"], marker="o")
    plt.title("7-Day Demand Forecast")
    plt.xlabel("Day")
    plt.ylabel("Forecasted Demand")

    buffer = io.BytesIO()
    plt.savefig(buffer, format="png")
    plt.close()
    buffer.seek(0)

    # ⭐ Convert BytesIO → PIL
    image = Image.open(buffer)

    explanation = (
        "### 📈 Demand Forecasting\n"
        "Trend-weighted moving average model applied.\n\n"
        "#### 🔍 Reasoning\n" + "\n".join([f"- {r}" for r in reasoning_steps])
    )

    return explanation, image, forecast_df


# ============================================================
#  REPLENISHMENT ANALYSIS — MODULE 2
# ============================================================

def run_replenishment_analysis(message, slotting_df):
    reasoning_steps = []
    df = slotting_df.copy()

    if "Frequency" not in df.columns:
        return "Frequency missing — cannot run replenishment.", None

    BIN_CAPACITY = 200
    df["CurrentStock"] = BIN_CAPACITY * 0.5
    reasoning_steps.append("Assumed current stock = 50% bin capacity.")

    df["SafetyStock"] = df["Frequency"] * 3
    reasoning_steps.append("Safety stock = 3 days of demand.")

    df["DaysUntilStockout"] = df["CurrentStock"] / df["Frequency"].replace(0, 0.1)
    reasoning_steps.append("Estimated days until stock-out.")

    df["ReplenishmentQty"] = (BIN_CAPACITY - df["CurrentStock"]).clip(lower=0)
    reasoning_steps.append("Calculated replenishment quantity needed.")

    df["Risk"] = df["DaysUntilStockout"].apply(
        lambda x: "🔴 HIGH" if x < 3 else ("🟡 MEDIUM" if x < 7 else "🟢 LOW")
    )
    reasoning_steps.append("Assigned risk level based on depletion rate.")

    explanation = (
        "### 🔄 Replenishment Analysis\n"
        "Replenishment needs evaluated using bin capacity, demand & stock-out timing.\n\n"
        "#### 🔍 Reasoning\n" + "\n".join([f"- {r}" for r in reasoning_steps])
    )

    return explanation, df


# ============================================================
#  INVENTORY REBALANCING — MODULE 3
# ============================================================

def run_rebalancing_analysis(message, slotting_df):
    reasoning = []
    df = slotting_df.copy()

    if "Frequency" not in df.columns:
        return "Frequency missing — cannot rebalance.", None

    if "Aisle" not in df.columns:
        df["Aisle"] = np.arange(1, len(df) + 1)
        reasoning.append("Aisle data missing — assigned aisles automatically.")

    velocity_map = {"fast": 3, "medium": 2, "slow": 1}
    df["VelScore"] = df["Velocity"].str.lower().map(velocity_map)
    df["LoadScore"] = df["VelScore"] * 0.6 + df["Frequency"] * 0.4
    reasoning.append("Calculated SKU load score (velocity + frequency).")

    aisle_load = df.groupby("Aisle")["LoadScore"].sum().reset_index()
    avg_load = aisle_load["LoadScore"].mean()

    aisle_load["Congestion"] = aisle_load["LoadScore"].apply(
        lambda x: "🔴 High" if x > avg_load * 1.25
        else ("🟡 Medium" if x > avg_load * 0.75 else "🟢 Low")
    )

    df = df.merge(aisle_load[["Aisle", "Congestion"]], on="Aisle", how="left")
    reasoning.append("Assigned congestion levels to aisles.")

    high_aisles = aisle_load[aisle_load["Congestion"] == "🔴 High"]["Aisle"].tolist()
    low_aisles = aisle_load[aisle_load["Congestion"] == "🟢 Low"]["Aisle"].tolist()

    move_plan = []

    if high_aisles and low_aisles:
        for aisle in high_aisles:
            congested_skus = df[df["Aisle"] == aisle].sort_values("LoadScore", ascending=False)
            top_to_move = congested_skus.head(2)

            for i, row in top_to_move.iterrows():
                target_aisle = low_aisles[i % len(low_aisles)]
                move_plan.append({
                    "SKU": row["SKU"],
                    "FromAisle": row["Aisle"],
                    "ToAisle": target_aisle,
                    "LoadScore": round(row["LoadScore"], 2),
                    "Reason": "Reduce congestion"
                })

        reasoning.append("Generated SKU redistribution plan.")
    else:
        reasoning.append("No congestion found — no rebalancing needed.")

    move_df = pd.DataFrame(move_plan)

    explanation = (
        "### 🔄 Inventory Rebalancing\n"
        "SKU redistribution plan to reduce aisle congestion.\n\n"
        "#### 🔍 Reasoning\n" + "\n".join([f"- {r}" for r in reasoning])
    )

    return explanation, move_df


# ============================================================
#  WORKFORCE OPTIMIZATION — MODULE 4
# ============================================================

def run_workforce_optimization(message, slotting_df):
    reasoning = []
    df = slotting_df.copy()

    if "Frequency" not in df.columns:
        return "Cannot calculate workforce — missing Frequency column.", None

    df["Workload"] = df["Frequency"] * 1.2
    total_workload = df["Workload"].sum()
    workers_needed = max(1, int(total_workload // 150))

    reasoning.append(f"Total workload: {total_workload:.2f}")
    reasoning.append(f"Workers required (estimated): {workers_needed}")

    result = pd.DataFrame({
        "Metric": ["Total Workload", "Estimated Workers Needed"],
        "Value": [total_workload, workers_needed]
    })

    explanation = (
        "### 👷 Workforce Optimization\n"
        "Estimated staffing requirement based on SKU workload.\n\n"
        "#### 🔍 Reasoning\n" + "\n".join([f"- {r}" for r in reasoning])
    )

    return explanation, result


# ============================================================
#  DOCK SCHEDULING OPTIMIZATION — MODULE 5
# ============================================================

def run_dock_scheduling(message, slotting_df):
    reasoning = []
    df = slotting_df.copy()

    df["Priority"] = df["Frequency"].rank(ascending=False)
    df["AssignedDock"] = df["Priority"].apply(lambda x: int((x - 1) % 3) + 1)

    reasoning.append("Assigned SKUs to 3 docks based on priority rank.")

    explanation = (
        "### 🚛 Dock Scheduling Optimization\n"
        "SKUs allocated to dock doors based on priority.\n\n"
        "#### 🔍 Reasoning\n" + "\n".join([f"- {r}" for r in reasoning])
    )

    return explanation, df[["SKU", "Frequency", "Priority", "AssignedDock"]]