#!/usr/bin/env python3
"""Build server-side continuous futures series (v2) with several roll/adjust choices and compare them
to each other and to the vendor continuous series (v1).

  continuous_compare.py --root CL --start 2018-01-01 --rolls volume,calendar --adjusts back_adjusted,ratio \
                        [--depth 1] [--timeframe 1day] [--vendor contin_adj_ratio] [--out cl.csv] [--plot cl.png]
"""
from __future__ import annotations

import argparse
import sys
from pathlib import Path

import numpy as np
import pandas as pd

sys.path.insert(0, str(Path(__file__).parent))
import hfmd  # noqa: E402


def describe_rolls(name: str, df: pd.DataFrame, meta: dict) -> None:
    rolls = meta.get("roll_dates") or []
    if df.empty:
        print(f"\n[{name}] no data")
        return
    years = max((df["datetime"].iloc[-1] - df["datetime"].iloc[0]).days / 365.25, 1e-9)
    roll_txt = f"{len(rolls)} rolls ({len(rolls) / years:.1f}/yr)" if rolls or not name.startswith("vendor") else "roll dates not exposed (vendor construction)"
    print(f"\n[{name}] {len(df):,} bars {df['datetime'].iloc[0].date()} → {df['datetime'].iloc[-1].date()} · {roll_txt}")
    if rolls:
        shown = rolls if len(rolls) <= 12 else rolls[:6] + ["…"] + rolls[-6:]
        print("  roll dates:", ", ".join(str(r) if isinstance(r, str) else str(r.get("date", r)) for r in shown))
    if "close" in df and len(df) > 1:
        print(f"  last close {df['close'].iloc[-1]:.4f} · first close {df['close'].iloc[0]:.4f}"
              + ("  (adjusted levels are NOT tradable prices)" if "adjust=none" not in name else ""))


def main() -> int:
    ap = argparse.ArgumentParser(description=__doc__, formatter_class=argparse.RawDescriptionHelpFormatter)
    ap.add_argument("--root", required=True)
    ap.add_argument("--start", default="2015-01-01")
    ap.add_argument("--end")
    ap.add_argument("--timeframe", default="1day")
    ap.add_argument("--depth", type=int, default=1)
    ap.add_argument("--rolls", default="volume", help="comma list: volume, open_interest, calendar")
    ap.add_argument("--adjusts", default="back_adjusted", help="comma list: none, back_adjusted, ratio")
    ap.add_argument("--vendor", help="also fetch the v1 vendor continuous with this adjustment: contin_UNadj, contin_adj_ratio, contin_adj_absolute")
    ap.add_argument("--out", help="CSV with all series' closes aligned on datetime")
    ap.add_argument("--plot")
    a = ap.parse_args()
    root = a.root.upper()

    series: dict[str, pd.Series] = {}
    for roll in a.rolls.split(","):
        for adj in a.adjusts.split(","):
            name = f"v2 roll={roll} adjust={adj} depth={a.depth}"
            try:
                df, meta = hfmd.continuous(root, roll=roll, adjust=adj, depth=a.depth, timeframe=a.timeframe, start=a.start, end=a.end)
            except hfmd.HfmdError as e:
                print(f"\n[{name}] {e}", file=sys.stderr)
                if e.status == 404:
                    print("  → the v2 continuous endpoint is not deployed on this server yet; use scripts/stitch_local.py", file=sys.stderr)
                continue
            describe_rolls(name, df, meta)
            if not df.empty:
                series[name] = df.set_index("datetime")["close"].astype(float)

    if a.vendor:
        name = f"vendor {a.vendor}"
        try:
            vdf = hfmd.bars("futures", root, a.timeframe, a.start, a.end, a.vendor)
            describe_rolls(name, vdf, {})
            if not vdf.empty:
                series[name] = vdf.set_index("datetime")["close"].astype(float)
        except hfmd.HfmdError as e:
            print(f"\n[{name}] {e}", file=sys.stderr)

    if not series:
        print("\nno series retrieved", file=sys.stderr)
        return 1

    wide = pd.concat(series, axis=1).sort_index()
    rets = wide.pct_change()
    print("\nreturn correlation between series (daily pct changes; 'none'/UNadj variants include roll gaps):")
    with pd.option_context("display.width", 200):
        print(rets.corr().round(4).to_string())
    if len(series) > 1:
        names = list(series)
        base = names[-1]
        print(f"\nannualised tracking difference of log-returns vs '{base}':")
        for n in names[:-1]:
            d = (np.log1p(rets[n]) - np.log1p(rets[base])).dropna()
            print(f"  {n:<45} mean={d.mean() * 252:+.4%}/yr  std={d.std() * np.sqrt(252):.4%}  n={len(d)}")

    if a.out:
        wide.to_csv(a.out)
        print(f"\nwrote {a.out} ({len(wide):,} rows × {wide.shape[1]} series)")
    if a.plot:
        try:
            import matplotlib
            matplotlib.use("Agg")
            import matplotlib.pyplot as plt
        except ImportError:
            print("matplotlib not installed", file=sys.stderr)
            return 0
        fig, ax = plt.subplots(figsize=(11, 5))
        for n, s in series.items():
            ax.plot(s.index, s / s.dropna().iloc[0] * 100, lw=1.0, label=n)
        ax.set_title(f"{root} continuous series rebased to 100 (differences = roll/adjust methodology)")
        ax.grid(alpha=0.25)
        ax.legend(fontsize=8)
        fig.tight_layout()
        fig.savefig(a.plot, dpi=130)
        print(f"wrote {a.plot}")
    return 0


if __name__ == "__main__":
    sys.exit(main())