iptv/M3U8/scrapers/streamed.py

import asyncio
import re
from functools import partial
from pathlib import Path

import httpx
from playwright.async_api import BrowserContext, async_playwright
from selectolax.parser import HTMLParser

from .utils import Cache, Time, get_logger, leagues, network

log = get_logger(__name__)

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

CACHE_FILE = Cache(Path(__file__).parent / "caches" / "streamed.json", exp=10_800)

HTML_CACHE = Cache(Path(__file__).parent / "caches" / "streamed_php.json", exp=86_400)

BASE_URL = "https://streamed.site/webmaster.php"


async def process_event(
    url: str,
    url_num: int,
    context: BrowserContext,
) -> str | None:
    page = await context.new_page()

    captured: list[str] = []

    got_one = asyncio.Event()

    handler = partial(network.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()


async def refresh_html_cache(client: httpx.AsyncClient, url: str) -> dict[str, str]:
    try:
        r = await client.get(url)
        r.raise_for_status()
    except Exception as e:
        log.error(f'Failed to fetch "{url}"\n{e}')

        return []

    soup = HTMLParser(r.text)
    events = {}
    now = Time.now()

    for row in soup.css("div.wrap div.row"):
        if not (date := row.css_first("div.date")):
            continue

        event_dt = Time.from_str(date.text(strip=True))

        if event_dt.date() != now.date():
            continue

        league = row.css_first("div.league")

        title = row.css_first("div.title")

        hds_a = row.css_first("div.hds a")

        if not (league and title and hds_a):
            continue

        sport, event = league.text(strip=True), title.text(strip=True)

        onclick = hds_a.attributes.get("onclick", "")

        if not (m := re.search(r"openPlayerPopup\(\s*(\d+)\s*\)", onclick)):
            continue

        key = f"[{sport}] {event} (STRMED)"

        events[key] = {
            "sport": sport,
            "event": event,
            "link": f"https://streamed.site/set.php?{m[1]}",
            "timestamp": now.timestamp(),
            "event_ts": event_dt.timestamp(),
        }

    return events


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

    if not (events := HTML_CACHE.load()):
        events = await refresh_html_cache(client, url)
        HTML_CACHE.write(events)

    live = []

    now = Time.clean(Time.now())
    start_ts = now.delta(minutes=-30).timestamp()
    end_ts = now.delta(minutes=30).timestamp()

    for k, v in events.items():
        if cached_keys & {k}:
            continue

        if not start_ts <= v["event_ts"] <= end_ts:
            continue

        live.append({**v})

    return live


async def scrape(client: httpx.AsyncClient) -> None:
    cached_urls = CACHE_FILE.load()
    cached_count = len(cached_urls)
    urls.update(cached_urls)

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

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

    events = await get_events(
        client,
        BASE_URL,
        set(cached_urls.keys()),
    )

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

    async with async_playwright() as p:
        browser, context = await network.browser(p, browser="brave")

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

            if url:
                sport, event, ts = ev["sport"], ev["event"], ev["event_ts"]

                tvg_id, logo = leagues.info(sport)

                key = f"[{sport}] {event} (STRMD)"

                entry = {
                    "url": url,
                    "logo": logo,
                    "base": "",
                    "timestamp": ts,
                    "id": tvg_id or "Live.Event.us",
                }

                urls[key] = cached_urls[key] = entry

        await browser.close()

    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")

    CACHE_FILE.write(cached_urls)