iptv/M3U8/scrapers/ppv.py

#!/usr/bin/env python3

import asyncio
import json
import re
from datetime import datetime, timedelta
from functools import partial
from pathlib import Path
from urllib.parse import urljoin

import httpx
from playwright.async_api import async_playwright

from .utils import (
    LOGOS,
    TZ,
    capture_req,
    get_base,
    get_logger,
    load_cache,
    new_browser,
    now,
    safe_process_event,
    write_cache,
)

log = get_logger(__name__)

urls: dict[str, dict[str, str | float]] = {}

API_FILE = Path(__file__).parent / "caches" / "ppv_api.json"

CACHE_FILE = Path(__file__).parent / "caches" / "ppv.json"

MIRRORS = [
    "https://ppvs.su",
    "https://ppv.to",
    "https://ppv.wtf",
    "https://ppv.land",
    "https://freeppv.fun",
]


async def refresh_api_cache(
    client: httpx.AsyncClient, url: str
) -> dict[str, dict[str, str]]:
    log.info("Refreshing API cache")

    try:
        r = await client.get(url)
        r.raise_for_status()
    except Exception as e:
        log.error(f'Failed to fetch "{url}"\n{e}')
        return {}

    return r.json()


async def process_event(url: str, url_num: int) -> str | None:
    async with async_playwright() as p:
        browser, context = await new_browser(p)

        page = await context.new_page()

        captured: list[str] = []

        got_one = asyncio.Event()

        handler = partial(capture_req, captured=captured, got_one=got_one)

        page.on("request", handler)

        try:
            await page.goto(url, wait_until="domcontentloaded", timeout=15_000)

            wait_task = asyncio.create_task(got_one.wait())

            try:
                await asyncio.wait_for(wait_task, timeout=10)
            except asyncio.TimeoutError:
                log.warning(f"URL {url_num}) Timed out waiting for M3U8.")
                return

            finally:
                if not wait_task.done():
                    wait_task.cancel()

                    try:
                        await wait_task
                    except asyncio.CancelledError:
                        pass

            if captured:
                log.info(f"URL {url_num}) Captured M3U8")

                return captured[-1]

            log.warning(f"URL {url_num}) No M3U8 captured after waiting.")
            return

        except Exception as e:
            log.warning(f"URL {url_num}) Exception while processing: {e}")
            return

        finally:
            page.remove_listener("request", handler)
            await page.close()
            await browser.close()


async def get_events(
    client: httpx.AsyncClient,
    api_url: str,
    cached_keys: set[str],
) -> list[dict[str, str]]:

    events: list[dict[str, str]] = []

    base_url = re.match(r"(https?://.+?)/", api_url)[1]

    if not (
        api_data := load_cache(
            API_FILE,
            exp=86400,
            nearest_hr=True,
            per_entry=False,
        )
    ):
        api_data = await refresh_api_cache(client, api_url)
        API_FILE.write_text(json.dumps(api_data, indent=2), encoding="utf-8")

    for stream_group in api_data["streams"]:
        sport = stream_group["category"]

        if sport == "24/7 Streams":
            continue

        for event in stream_group["streams"]:
            name, start_ts, end_ts, logo, uri_name = (
                event["name"],
                event["starts_at"],
                event["ends_at"],
                event.get("poster", LOGOS["default"]),
                event["uri_name"],
            )

            key = f"[{sport}] {name} (PPV)"

            if cached_keys & {key}:
                continue

            start_dt = datetime.fromtimestamp(start_ts, tz=TZ) - timedelta(minutes=30)

            end_dt = datetime.fromtimestamp(end_ts, tz=TZ) + timedelta(minutes=30)

            if not start_dt <= now < end_dt:
                continue

            events.append(
                {
                    "sport": sport,
                    "event": name,
                    "link": urljoin(base_url, f"live/{uri_name}"),
                    "logo": logo,
                }
            )

    return events


async def main(client: httpx.AsyncClient) -> None:
    cached_urls = load_cache(CACHE_FILE, exp=10800)
    cached_count = len(cached_urls)
    urls.update(cached_urls)

    log.info(f"Collected {cached_count} event(s) from cache")

    if not (base_url := await get_base(client, MIRRORS)):
        log.warning("No working PPV mirrors")
        write_cache(CACHE_FILE, cached_urls)
        return

    log.info(f'Scraping from "{base_url}"')

    events = await get_events(
        client,
        urljoin(base_url, "api/streams"),
        set(cached_urls.keys()),
    )

    log.info(f"Processing {len(events)} new URL(s)")

    for i, ev in enumerate(events, start=1):
        url = await safe_process_event(
            lambda: process_event(ev["link"], url_num=i),
            url_num=i,
            log=log,
        )

        if url:
            key = f"[{ev['sport']}] {ev['event']} (PPV)"

            entry = {
                "url": url,
                "logo": ev["logo"],
                "base": base_url,
                "timestamp": now.timestamp(),
            }

            urls[key] = cached_urls[key] = entry

    if new_count := len(cached_urls) - cached_count:
        log.info(f"Collected and cached {new_count} new event(s)")
    else:
        log.info("No new events found")

    write_cache(CACHE_FILE, cached_urls)


# works if no cloudflare bot detection