Airbnb Scraper: Public Airbnb Listings via API

[
  {"id": "<room id 1>", "title": "Loft in Alfama", "price": 412, "currency": "€", "price_text": "€412",
   "checkin": "2026-12-10", "checkout": "2026-12-14", "nights": 4, "adults": 2, "country": "us"}
]

import statistics
from datetime import date, timedelta

DAYS = r"""(() => [...document.querySelectorAll('[data-testid^="calendar-day-"]')].map(d => ({
  date: d.getAttribute('data-testid').replace('calendar-day-', ''),
  blocked: d.getAttribute('data-is-day-blocked') === 'true'
})))()"""
MIN_NIGHTS = r"""(() => { const m = document.body.innerText.match(/(\d+)[- ]night minimum/i);
  return m ? Number(m[1]) : null; })()"""

def calendar(room_id, months=6):                      # up to 10 months: 2 * months - 1 + 1 <= 20 actions
    read = {"type": "evaluate", "script": DAYS}
    actions = [read] + [{"type": "click", "selector": "<next-month button>"}, read] * (months - 1)
    actions.append({"type": "evaluate", "script": MIN_NIGHTS})
    r = requests.post(API, headers=HEAD, timeout=180, json={
        "url": f"https://www.airbnb.com/rooms/{room_id}", "render": True,
        "country": COUNTRY, "output": "json", "wait_for": "[data-testid^='calendar-day-']",
        "actions": actions}).json()
    evals = [s for s in r["actions_log"] if s["type"] == "evaluate"]
    days = {}
    for step in evals[:-1]:
        for d in step.get("result") or []:
            days[d["date"]] = d                            # views overlap: one record per date
    blocked = sum(d["blocked"] for d in days.values())
    return {"id": room_id, "country": COUNTRY, "min_nights": evals[-1].get("result"),
            "days": list(days.values()),
            "summary": {"days": len(days), "open": len(days) - blocked, "blocked": blocked}}

PRICE = r"""(() => (document.body.innerText.match(/[A-Z]{0,3}[$€£¥₹]\s?\d[\d.,]*/) || [null])[0])()"""

def prices(room_id, first_checkin, n_dates=30, nights=2):   # 1 result per check-in date
    out = []
    for i in range(n_dates):
        ci = date.fromisoformat(first_checkin) + timedelta(days=i)
        co = ci + timedelta(days=nights)
        r = requests.post(API, headers=HEAD, timeout=180, json={
            "url": f"https://www.airbnb.com/rooms/{room_id}?check_in={ci}&check_out={co}&adults={ADULTS}",
            "render": True, "country": COUNTRY, "output": "json",
            "actions": [{"type": "evaluate", "script": PRICE}]}).json()
        price, currency = parse_price(r["actions_log"][0].get("result"))
        if price is not None:
            out.append({"checkin": str(ci), "checkout": str(co), "nights": nights,
                        "price": price, "currency": currency,
                        "price_text": r["actions_log"][0].get("result")})
    vals = [p["price"] for p in out]
    return {"id": room_id, "country": COUNTRY, "adults": ADULTS, "prices": out,
            "summary": {"dates": len(vals), "min": min(vals, default=None),
                        "median": statistics.median(vals) if vals else None,
                        "max": max(vals, default=None)}}

def listing(room_id, first_checkin, n_dates=8, nights=2, months=10):
    cal = calendar(room_id, months)                          # 1 result
    open_days = [d["date"] for d in sorted(cal["days"], key=lambda d: d["date"])
                 if not d["blocked"] and d["date"] >= first_checkin][:n_dates]
    priced = {}
    for ci in open_days:                                     # 1 result per open check-in date
        p = prices(room_id, ci, n_dates=1, nights=nights)["prices"]
        if p:
            priced[ci] = p[0]
    days = []
    for d in sorted(cal["days"], key=lambda d: d["date"]):
        row = {"date": d["date"], "blocked": d["blocked"]}
        if d["date"] in priced:
            row.update({k: priced[d["date"]][k] for k in ("price", "currency", "price_text")})
        days.append(row)
    vals = [p["price"] for p in priced.values()]
    return {"id": room_id, "country": COUNTRY, "adults": ADULTS, "nights": nights,
            "min_nights": cal["min_nights"], "days": days,
            "summary": {**cal["summary"], "priced": len(vals), "min": min(vals, default=None),
                        "median": statistics.median(vals) if vals else None,
                        "max": max(vals, default=None)}}

city = [listing(room_id, "2026-12-10") for room_id in listings]   # listings: from crawl() above
json.dump(city, open("lisbon-calendars.json", "w"), ensure_ascii=False, indent=1)