#!/usr/bin/env python3
"""Compare OSRM road distances against previously-recorded (Google) distances.

Routing distance feeds `riderkms` -> `ridercharges`, so a systematic difference
between the old and new routing backend changes what riders get paid. Run this
against a sample of completed deliveries BEFORE rolling out.

Input CSV (header required), one row per completed delivery:

    origin_lat,origin_lng,dest_lat,dest_lng,recorded_km
    11.0168,76.9558,10.9984,76.9776,4.6
    11.0510,76.9910,11.0168,76.9558,6.1

Usage:

    python3 tool/compare_route_distances.py deliveries.csv
    python3 tool/compare_route_distances.py deliveries.csv \\
        --osrm https://osrm.your-domain.in
"""

import argparse
import csv
import json
import statistics
import subprocess
import sys
import urllib.parse
import urllib.request

DEFAULT_OSRM = "https://routing.openstreetmap.de/routed-car"
USER_AGENT = "nearle-distance-check"


def _fetch(url, timeout):
    """GET a URL, falling back to curl when Python's SSL stack is too old.

    Some systems ship a Python linked against an OpenSSL that fails the
    handshake against modern servers; curl on the same box usually works.
    """
    try:
        req = urllib.request.Request(url, headers={"User-Agent": USER_AGENT})
        with urllib.request.urlopen(req, timeout=timeout) as resp:
            return resp.read().decode()
    except Exception:                                     # noqa: BLE001
        result = subprocess.run(
            ["curl", "-sS", "--max-time", str(timeout), "-A", USER_AGENT, url],
            capture_output=True, text=True,
        )
        if result.returncode != 0:
            raise RuntimeError(result.stderr.strip() or "curl failed") from None
        return result.stdout


def osrm_km(base, o_lat, o_lng, d_lat, d_lng, timeout=25):
    """Road distance in km, or None. Note OSRM takes lon,lat."""
    url = (
        f"{base}/route/v1/driving/"
        f"{o_lng},{o_lat};{d_lng},{d_lat}"
        f"?{urllib.parse.urlencode({'overview': 'false'})}"
    )
    data = json.loads(_fetch(url, timeout))
    if data.get("code") != "Ok" or not data.get("routes"):
        return None
    return data["routes"][0]["distance"] / 1000.0


def main():
    ap = argparse.ArgumentParser(description=__doc__,
                                 formatter_class=argparse.RawDescriptionHelpFormatter)
    ap.add_argument("csv_path", help="CSV of completed deliveries")
    ap.add_argument("--osrm", default=DEFAULT_OSRM,
                    help=f"OSRM base URL (default: {DEFAULT_OSRM})")
    args = ap.parse_args()

    deltas, failures = [], 0

    print(f"{'recorded':>10} {'osrm':>10} {'delta':>10} {'pct':>8}")
    print("-" * 42)

    with open(args.csv_path, newline="") as fh:
        for row in csv.DictReader(fh):
            try:
                recorded = float(row["recorded_km"])
                km = osrm_km(
                    args.osrm,
                    float(row["origin_lat"]), float(row["origin_lng"]),
                    float(row["dest_lat"]), float(row["dest_lng"]),
                )
            except Exception as exc:                      # noqa: BLE001
                print(f"  ERROR: {exc}", file=sys.stderr)
                failures += 1
                continue

            if km is None:
                print("  no route found", file=sys.stderr)
                failures += 1
                continue

            delta = km - recorded
            pct = (delta / recorded * 100) if recorded else 0.0
            deltas.append(pct)
            print(f"{recorded:9.2f}k {km:9.2f}k {delta:+9.2f}k {pct:+7.1f}%")

    if not deltas:
        print("\nNo comparable rows.", file=sys.stderr)
        return 1

    mean = statistics.fmean(deltas)
    median = statistics.median(deltas)
    print("-" * 42)
    print(f"rows compared : {len(deltas)}  (failed: {failures})")
    print(f"mean delta    : {mean:+.1f}%")
    print(f"median delta  : {median:+.1f}%")
    if len(deltas) > 1:
        print(f"std dev       : {statistics.stdev(deltas):.1f}%")

    print()
    if abs(median) < 3:
        print("OK — no meaningful bias. Rider pay is unaffected.")
    elif abs(median) < 8:
        print("MILD BIAS — acceptable, but tell finance the average trip "
              f"distance shifts {median:+.1f}%.")
    else:
        print(f"SIGNIFICANT BIAS ({median:+.1f}%) — investigate before rollout. "
              "Usually means missing OSM roads in your service area; consider "
              "contributing the missing roads to OpenStreetMap, which fixes it "
              "permanently and for everyone.")
    return 0


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