ani-web/server/providers/anikoto.py

"""providers/anikoto.py — anikototv.to (AniKoto / Hianime) fallback source.

Flow: search page -> watch page (numeric anime id) -> ajax episode list
(per-episode token) -> ajax server list (sub/dub groups) -> ajax/server?get=
(MegaPlay embed page) -> getSourcesNew?id= (master m3u8).

MegaPlay's CDN domains rotate per getSources call and some (e.g. norami.top)
hard-403 non-browser clients, so _master_m3u8 re-rolls the playlist a few
times until a CDN domain actually serves us.
"""
import re

from ..upstream import _cached, _get, megaplay_sources

BASE = "https://anikototv.to"
XHR = {"X-Requested-With": "XMLHttpRequest"}
_MEGAPLAY_ID = re.compile(r'data-id="(\d+)"')


def _search(q):
    html = _get(f"{BASE}/search", params={"keyword": q}).text
    out, seen = [], set()
    for m in re.finditer(
            r'href="https://anikototv\.to/watch/([a-z0-9-]+)[^"]*"[^>]*>\s*([^<]{2,80})</a>',
            html):
        slug, name = m.group(1), m.group(2).strip()
        if slug in seen:
            continue
        seen.add(slug)
        imgs = re.findall(r'<img src="(https://cdn\.anipixcdn\.co/[^"]+)"',
                          html[max(0, m.start() - 600):m.start()])
        out.append({"id": 0, "slug_id": slug, "name": name,
                    "poster": imgs[-1] if imgs else None})
    return out


def _details(slug):
    t = _get(f"{BASE}/watch/{slug}").text
    h1 = re.search(r'<h1[^>]*>(.*?)</h1>', t, re.S)
    syn = re.search(r'class="synopsis[^"]*"[\s\S]*?class="content"[\s\S]*?>(.*?)</div>', t)
    post = re.search(r'(https://cdn\.anipixcdn\.co/thumbnail/[^"\']+)', t)
    aid = re.search(r'data-anime-id="(\d+)"', t)
    return {"id": int(aid.group(1)) if aid else 0,
            "name": re.sub(r"<[^>]+>", "", h1.group(1)).strip() if h1 else None,
            "synopsis": re.sub(r"<[^>]+>", "", syn.group(1)).strip() if syn else None,
            "mal_id": None,
            "poster": post.group(1) if post else None,
            "seasons": []}


def _ep_entries(anime_id):
    d = _get(f"{BASE}/ajax/episode/list/{anime_id}", headers=XHR).json()
    out = []
    for tag in re.findall(r'<a[^>]+>', d.get("result") or ""):
        num = re.search(r'data-num="(\d+)"', tag)
        tok = re.search(r'data-ids="([^"]+)"', tag)
        if num and tok:
            out.append({"num": int(num.group(1)), "token": tok.group(1),
                        "sub": 'data-sub="1"' in tag, "dub": 'data-dub="1"' in tag})
    return out


def _ep_entry(ep_id):
    # ep_id encodes (anime_id, num): anime_id * 1000 + num (num < 1000)
    anime_id, num = divmod(ep_id, 1000)
    return next((e for e in _ep_entries(anime_id) if e["num"] == num), None)


def _episodes(anime_id):
    return [{"id": anime_id * 1000 + e["num"], "number": e["num"], "filler": False}
            for e in _ep_entries(anime_id)]


def _languages(ep_id):
    e = _ep_entry(ep_id)
    if not e:
        return []
    out = []
    if e["sub"]:
        out.append({"code": "jpn", "name": "Sub", "embed_url": "sub"})
    if e["dub"]:
        out.append({"code": "eng", "name": "Dub", "embed_url": "dub"})
    return out


def _server_link_ids(token, want):
    d = _get(f"{BASE}/ajax/server/list", params={"servers": token}, headers=XHR).json()
    res = d.get("result") or ""
    for m in re.finditer(r'data-type="(\w+)"[\s\S]*?<ul>([\s\S]*?)</ul>', res):
        if m.group(1) == want:
            return re.findall(r'data-link-id="([^"]+)"', m.group(2))
    return re.findall(r'data-link-id="([^"]+)"', res)  # no group split: take all


def _player_file(embed_url):
    """MegaPlay AND its re-skins (vidtube etc.): same player libs, so the
    same getSourcesNew endpoint on the embed's own host."""
    page = _get(embed_url, timeout=15)
    m = _MEGAPLAY_ID.search(page.text)
    if not m:
        return None
    host = re.match(r'https?://([^/]+)/', embed_url).group(1)
    d = megaplay_sources(_get(f"https://{host}/stream/getSourcesNew?id={m.group(1)}",
             headers={"Referer": embed_url, "X-Requested-With": "XMLHttpRequest"}).json())
    return (d.get("sources") or {}).get("file")


def _master_m3u8(ep_id, lang="eng"):
    e = _ep_entry(ep_id)
    if not e:
        return None
    want = "sub" if lang in ("jpn", "sub") else "dub"
    ids = _server_link_ids(e["token"], want) or _server_link_ids(e["token"], "sub")
    errs = []
    for lid in ids[:3]:
        d = _get(f"{BASE}/ajax/server", params={"get": lid}, headers=XHR).json()
        url = (d.get("result") or {}).get("url")
        if not url:
            continue
        for _ in range(2):  # re-roll: CDN domains rotate, most 403 our IP
            try:
                f = _player_file(url)
                if not f:
                    break
                _get(f, timeout=10)  # prove this CDN serves us before committing
                return f
            except Exception as ex:
                errs.append(str(ex)[:60])
    raise RuntimeError("anikoto: no working server: " + "; ".join(errs[:3]))


# Provider interface (cache keys namespaced "anikoto:").
search = lambda q: _cached(f"anikoto:search:{q.lower()}", 300, _search, q)
details = lambda s: _cached(f"anikoto:details:{s}", 3600, _details, s)
episodes = lambda i: _cached(f"anikoto:episodes:{i}", 300, _episodes, i)
languages = lambda e: _cached(f"anikoto:languages:{e}", 3600, _languages, e)
master_m3u8 = lambda e, l="eng": _cached(f"anikoto:master:{e}:{l}", 3600, _master_m3u8, e, l)