"""Insider transactions — recent buys/sells with summary and detail table."""
import pandas as pd
import plotly.graph_objects as go
import streamlit as st
from datetime import datetime, timedelta

from services.data_service import get_insider_transactions
from utils.formatters import fmt_large


def _classify(text: str) -> str:
    """Return 'Buy', 'Sell', or 'Other' from the transaction text."""
    t = str(text or "").lower()
    if any(k in t for k in ("sale", "sold", "disposition")):
        return "Sell"
    if any(k in t for k in ("purchase", "bought", "acquisition", "grant", "award", "exercise")):
        return "Buy"
    return "Other"


def render_insiders(ticker: str):
    with st.spinner("Loading insider transactions…"):
        df = get_insider_transactions(ticker)

    if df.empty:
        st.info("No insider transaction data available for this ticker.")
        return

    # Normalise columns — yfinance returns: Shares, URL, Text, Insider, Position,
    # Transaction, Start Date, Ownership, Value
    df = df.copy()
    df["Direction"] = df["Text"].apply(_classify)

    # Parse dates
    def _to_dt(val):
        try:
            return pd.to_datetime(val)
        except Exception:
            return pd.NaT

    df["_date"] = df["Start Date"].apply(_to_dt)

    # ── Summary: last 6 months ────────────────────────────────────────────
    cutoff = pd.Timestamp(datetime.now() - timedelta(days=180))
    recent = df[df["_date"] >= cutoff]

    buys = recent[recent["Direction"] == "Buy"]
    sells = recent[recent["Direction"] == "Sell"]

    def _total_value(subset):
        try:
            return subset["Value"].dropna().astype(float).sum()
        except Exception:
            return 0.0

    buy_val = _total_value(buys)
    sell_val = _total_value(sells)

    st.markdown("**Insider Activity — Last 6 Months**")
    c1, c2, c3, c4 = st.columns(4)
    c1.metric("Buy Transactions", len(buys))
    c2.metric("Total Bought", fmt_large(buy_val) if buy_val else "—")
    c3.metric("Sell Transactions", len(sells))
    c4.metric("Total Sold", fmt_large(sell_val) if sell_val else "—")

    # Monthly bar chart
    if not recent.empty:
        monthly: dict[str, dict] = {}
        for _, row in recent.iterrows():
            if pd.isna(row["_date"]):
                continue
            key = row["_date"].strftime("%Y-%m")
            monthly.setdefault(key, {"Buy": 0.0, "Sell": 0.0})
            try:
                val = float(row["Value"]) if pd.notna(row["Value"]) else 0.0
            except (TypeError, ValueError):
                val = 0.0
            if row["Direction"] in ("Buy", "Sell"):
                monthly[key][row["Direction"]] += val

        months = sorted(monthly.keys())
        if months:
            fig = go.Figure()
            fig.add_trace(go.Bar(
                x=months, y=[monthly[m]["Buy"] / 1e6 for m in months],
                name="Buys", marker_color="#2ecc71",
            ))
            fig.add_trace(go.Bar(
                x=months, y=[-monthly[m]["Sell"] / 1e6 for m in months],
                name="Sells", marker_color="#e74c3c",
            ))
            fig.update_layout(
                title="Monthly Insider Net Activity ($M)",
                barmode="relative",
                yaxis_title="Value ($M)",
                plot_bgcolor="rgba(0,0,0,0)",
                paper_bgcolor="rgba(0,0,0,0)",
                margin=dict(l=0, r=0, t=40, b=0),
                height=280,
                legend=dict(orientation="h", yanchor="bottom", y=1.02, xanchor="right", x=1),
            )
            st.plotly_chart(fig, use_container_width=True)

    st.divider()

    # ── Transaction table ─────────────────────────────────────────────────
    st.markdown("**All Transactions**")

    filter_col, _ = st.columns([1, 3])
    with filter_col:
        direction_filter = st.selectbox(
            "Filter", options=["All", "Buy", "Sell"], index=0, key="insider_filter"
        )

    filtered = df if direction_filter == "All" else df[df["Direction"] == direction_filter]

    if filtered.empty:
        st.info("No transactions match the current filter.")
        return

    display = pd.DataFrame({
        "Date": filtered["Start Date"].astype(str).str[:10],
        "Insider": filtered["Insider"],
        "Position": filtered["Position"],
        "Type": filtered["Direction"],
        "Shares": filtered["Shares"].apply(
            lambda v: f"{int(v):,}" if pd.notna(v) else "—"
        ),
        "Value": filtered["Value"].apply(
            lambda v: fmt_large(float(v)) if pd.notna(v) and float(v) > 0 else "—"
        ),
    }).reset_index(drop=True)

    def _color_type(row):
        if row["Type"] == "Buy":
            return [""] * 3 + ["background-color: rgba(46,204,113,0.15)"] + [""] * 2
        if row["Type"] == "Sell":
            return [""] * 3 + ["background-color: rgba(231,76,60,0.15)"] + [""] * 2
        return [""] * len(row)

    st.dataframe(
        display.style.apply(_color_type, axis=1),
        use_container_width=True,
        hide_index=True,
    )