reyna-cli/backup/macmini-automation-baseline-2026-08-03/hermes-scripts/chiapas_flight_price_monitor.py

#!/usr/bin/env python3
"""Silent watchdog for Chiapas flight prices.

Checks the user's preferred KAYAK search URL and only prints an alert when a
meaningful price drop is detected. Empty stdout means no WhatsApp message.
"""
import json
import os
import re
import subprocess
import sys
import tempfile
from datetime import datetime, timezone
from pathlib import Path

URL = "https://www.kayak.com/flights/MIA,MCO-TGZ/2026-09-21/2026-10-09/2adults/children-5-8?ucs=uwkx32&sort=bestflight_a"
STATE_PATH = Path.home() / ".hermes" / "state" / "chiapas_flight_monitor.json"
CHROMIUM = "/snap/bin/chromium"

# Alert policy: establish a baseline silently, then alert only if the best price
# is materially better than what we have seen before.
DROP_PERCENT = 0.20          # at least 20% lower than baseline/best seen
DROP_ABSOLUTE = 300          # or at least $300 lower for the searched party
VERY_GOOD_TOTAL = 450        # always alert at or below this displayed fare
MIN_REASONABLE = 400         # ignore UI noise like baggage/filter slider prices
MAX_REASONABLE = 8000        # ignore unrelated large numbers


def load_state():
    if STATE_PATH.exists():
        try:
            return json.loads(STATE_PATH.read_text())
        except Exception:
            return {}
    return {}


def save_state(state):
    STATE_PATH.parent.mkdir(parents=True, exist_ok=True)
    STATE_PATH.write_text(json.dumps(state, indent=2, sort_keys=True))


def fetch_html():
    if not Path(CHROMIUM).exists():
        return ""
    with tempfile.NamedTemporaryFile(prefix="chiapas_kayak_", suffix=".html", delete=False) as tmp:
        out_path = tmp.name
    try:
        cmd = [
            CHROMIUM,
            "--headless",
            "--disable-gpu",
            "--no-sandbox",
            "--disable-dev-shm-usage",
            "--user-agent=Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 Chrome/120 Safari/537.36",
            "--virtual-time-budget=60000",
            "--dump-dom",
            URL,
        ]
        result = subprocess.run(cmd, stdout=subprocess.PIPE, stderr=subprocess.DEVNULL, text=True, timeout=120)
        return result.stdout or ""
    except Exception:
        return ""
    finally:
        try:
            os.unlink(out_path)
        except Exception:
            pass


def extract_prices(html):
    # KAYAK may render prices as visible text or escaped JSON. Keep this simple
    # and conservative: only currency-looking amounts in a plausible full-party range.
    raw = re.findall(r"\$\s*([0-9][0-9,]{2,5})", html)
    prices = []
    for s in raw:
        try:
            val = int(s.replace(",", ""))
        except ValueError:
            continue
        if MIN_REASONABLE <= val <= MAX_REASONABLE:
            prices.append(val)
    return sorted(set(prices))


def main():
    now = datetime.now(timezone.utc).isoformat()
    state = load_state()
    html = fetch_html()
    prices = extract_prices(html)

    state["last_checked_utc"] = now
    state["url"] = URL

    if not prices:
        state["last_status"] = "no_parseable_prices"
        save_state(state)
        return 0

    current = min(prices)
    state["last_prices_seen"] = prices[:10]
    state["last_lowest"] = current

    baseline = state.get("baseline_lowest")
    best_seen = state.get("best_seen")

    if baseline is None:
        state["baseline_lowest"] = current
        state["best_seen"] = current
        state["last_status"] = "baseline_set_silent"
        save_state(state)
        return 0

    if best_seen is None or current < best_seen:
        state["best_seen"] = current

    threshold_from_baseline = int(baseline * (1 - DROP_PERCENT))
    threshold_from_best = int(best_seen * (1 - DROP_PERCENT)) if best_seen else threshold_from_baseline
    absolute_threshold = max(0, baseline - DROP_ABSOLUTE)

    should_alert = (
        current <= VERY_GOOD_TOTAL
        or current <= threshold_from_baseline
        or current <= threshold_from_best
        or current <= absolute_threshold
    )

    already_alerted_at = state.get("last_alert_price")
    if should_alert and (already_alerted_at is None or current < already_alerted_at):
        state["last_alert_price"] = current
        state["last_status"] = "alerted"
        save_state(state)
        print(
            "Chiapas flight price alert: KAYAK is showing a notably better fare.\n"
            f"Route/search: MIA or MCO → TGZ, Sept 21–Oct 9, 2026, family search.\n"
            f"Lowest price I could parse today: ${current:,}.\n"
            f"Baseline: ${baseline:,}. Best previously seen: ${best_seen:,}.\n"
            f"Check/book here: {URL}"
        )
        return 0

    state["last_status"] = "checked_no_alert"
    save_state(state)
    return 0


if __name__ == "__main__":
    raise SystemExit(main())