iptv/M3U8/scrapers/webcast.py

import asyncio
import re
from functools import partial

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]] = {}

TAG = "WEBCAST"

CACHE_FILE = Cache(TAG, exp=10_800)

HTML_CACHE = Cache(f"{TAG}-html", exp=19_800)

BASE_URLS = {
    "MLB": "https://mlbwebcast.com",
    # "NFL": "https://nflwebcast.com",
    "NHL": "https://slapstreams.com",
}


def fix_event(s: str) -> str:
    return " vs ".join(s.split("@"))


async def process_event(url: str, url_num: int) -> str | None:
    if not (event_data := await network.request(url, log=log)):
        log.info(f"URL {url_num}) Failed to load url.")

        return

    soup = HTMLParser(event_data.content)

    if not (iframe := soup.css_first('iframe[name="srcFrame"]')):
        log.warning(f"URL {url_num}) No iframe element found.")

        return

    if not (iframe_src := iframe.attributes.get("src")):
        log.warning(f"URL {url_num}) No iframe source found.")

        return

    if not (
        iframe_src_data := await network.request(
            iframe_src,
            headers={"Referer": url},
            log=log,
        )
    ):
        log.info(f"URL {url_num}) Failed to load iframe source.")

        return

    pattern = re.compile(r"source:\s+(\'|\")(.*)(\'|\")", re.I)

    if not (match := pattern.search(iframe_src_data.text)):
        log.warning(f"URL {url_num}) No Clappr source found.")

        return

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

    return match[2]


async def refresh_html_cache(url: str) -> dict[str, dict[str, str | float]]:
    events = {}

    if not (html_data := await network.request(url, log=log)):
        return events

    now = Time.clean(Time.now())

    soup = HTMLParser(html_data.content)

    sport = next((k for k, v in BASE_URLS.items() if v == url), "Live Event")

    date_text = now.strftime("%B %d, %Y")

    if date_row := soup.css_first("tr.mdatetitle"):
        if mtdate_span := date_row.css_first("span.mtdate"):
            date_text = mtdate_span.text(strip=True)

    for row in soup.css("tr.singele_match_date"):
        if not (time_node := row.css_first("td.matchtime")):
            continue

        time = time_node.text(strip=True)

        if not (vs_node := row.css_first("td.teamvs a")):
            continue

        event_name = vs_node.text(strip=True)

        for span in vs_node.css("span.mtdate"):
            date = span.text(strip=True)

            event_name = event_name.replace(date, "").strip()

        if not (href := vs_node.attributes.get("href")):
            continue

        event_dt = Time.from_str(f"{date_text} {time} PM", timezone="EST")

        event = fix_event(event_name)

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

        events[key] = {
            "sport": sport,
            "event": event,
            "link": href,
            "event_ts": event_dt.timestamp(),
            "timestamp": now.timestamp(),
        }

    return events


async def get_events(cached_keys: list[str]) -> list[dict[str, str]]:
    now = Time.clean(Time.now())

    if not (events := HTML_CACHE.load()):
        log.info("Refreshing HTML cache")

        tasks = [refresh_html_cache(url) for url in BASE_URLS.values()]

        results = await asyncio.gather(*tasks)

        events = {k: v for data in results for k, v in data.items()}

        HTML_CACHE.write(events)

    live = []

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

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

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

        live.append(v)

    return live


async def scrape() -> None:
    cached_urls = CACHE_FILE.load()

    valid_urls = {k: v for k, v in cached_urls.items() if v["url"]}

    valid_count = cached_count = len(valid_urls)

    urls.update(valid_urls)

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

    log.info(f'Scraping from "{' & '.join(BASE_URLS.values())}"')

    events = await get_events(cached_urls.keys())

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

        for i, ev in enumerate(events, start=1):
            handler = partial(
                process_event,
                url=(link := ev["link"]),
                url_num=i,
            )

            url = await network.safe_process(
                handler,
                url_num=i,
                semaphore=network.PW_S,
                log=log,
            )

            sport, event, ts = (
                ev["sport"],
                ev["event"],
                ev["event_ts"],
            )

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

            tvg_id, logo = leagues.get_tvg_info(sport, event)

            entry = {
                "url": url,
                "logo": logo,
                "base": BASE_URLS[sport],
                "timestamp": ts,
                "id": tvg_id or "Live.Event.us",
                "link": link,
            }

            cached_urls[key] = entry

            if url:
                valid_count += 1

                urls[key] = entry

    if new_count := valid_count - 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)