hermes-agent/gateway/platforms/weixin.py

"""Weixin (WeChat personal account) adapter over Tencent's iLink Bot API. Long-poll ``getupdates`` drives inbound
delivery; every outbound reply must echo the peer's latest ``context_token``; media moves through an AES-128-ECB
encrypted CDN; ``qr_login`` backs the gateway setup wizard."""

from __future__ import annotations

import asyncio, base64, contextlib, hashlib, json, logging, mimetypes, os, re, secrets, tempfile, textwrap, time, uuid  # noqa: E401
from datetime import datetime
from functools import partial
from pathlib import Path
from typing import Any, Awaitable, Callable, Dict, List, Optional, Tuple
from urllib.parse import quote, urlparse

logger = logging.getLogger(__name__)
WEIXIN_COPY_LINE_WIDTH = 120

try:
    import aiohttp
except ImportError:  # pragma: no cover - dependency gate
    aiohttp = None  # type: ignore[assignment]
AIOHTTP_AVAILABLE = aiohttp is not None

try:
    from cryptography.hazmat.backends import default_backend
    from cryptography.hazmat.primitives.ciphers import Cipher, algorithms, modes
except ImportError:  # pragma: no cover - dependency gate
    default_backend = Cipher = algorithms = modes = None  # type: ignore[assignment]
CRYPTO_AVAILABLE = Cipher is not None

from gateway.config import Platform, PlatformConfig
from gateway.platforms.helpers import MessageDeduplicator, cancel_task, greedy_pack_blocks
from gateway.platforms.access_policy_mixin import OwnAccessPolicyMixin
from agent.i18n import t
from gateway.platforms.base import (
    _IMAGE_EXTS, _VIDEO_EXTS, gateway_trust_env, BasePlatformAdapter, SendResult,
    cache_audio_from_bytes_async, cache_document_from_bytes_async, cache_image_from_bytes_async,
)
from gateway.platforms.event import MessageEvent, MessageType
from hermes_constants import get_hermes_home
from utils import atomic_json_write
from gateway.platforms._shared import extra_or_secret as _extra_or_env, get_scoped_secret as _wx_secret


def _extra_or_secret(extra: Dict[str, Any], key: str, default: str = "") -> str:
    """``config.extra[key]`` first, else the scoped ``WEIXIN_<KEY>``; stripped."""
    return str(_extra_or_env(extra, key, f"WEIXIN_{key.upper()}", default)).strip()


ILINK_BASE_URL = "https://ilinkai.weixin.qq.com"
WEIXIN_CDN_BASE_URL = "https://novac2c.cdn.weixin.qq.com/c2c"
ILINK_APP_ID, CHANNEL_VERSION, ILINK_APP_CLIENT_VERSION = "bot", "2.2.0", (2 << 16) | (2 << 8) | 0
EP_GET_UPDATES, EP_SEND_MESSAGE, EP_SEND_TYPING = "ilink/bot/getupdates", "ilink/bot/sendmessage", "ilink/bot/sendtyping"
EP_GET_CONFIG, EP_GET_UPLOAD_URL = "ilink/bot/getconfig", "ilink/bot/getuploadurl"
EP_GET_BOT_QR, EP_GET_QR_STATUS = "ilink/bot/get_bot_qrcode", "ilink/bot/get_qrcode_status"
LONG_POLL_TIMEOUT_MS, API_TIMEOUT_MS, CONFIG_TIMEOUT_MS, QR_TIMEOUT_MS = 35_000, 15_000, 10_000, 35_000
MAX_CONSECUTIVE_FAILURES, RETRY_DELAY_SECONDS, BACKOFF_DELAY_SECONDS = 3, 2, 30
SESSION_EXPIRED_ERRCODE, RATE_LIMIT_ERRCODE = -14, -2  # -2: iLink frequency limit — backoff and retry
MESSAGE_DEDUP_TTL_SECONDS = 300
MEDIA_IMAGE, MEDIA_VIDEO, MEDIA_FILE, MEDIA_VOICE = 1, 2, 3, 4  # getuploadurl media_type
ITEM_TEXT, ITEM_IMAGE, ITEM_VOICE, ITEM_FILE, ITEM_VIDEO = 1, 2, 3, 4, 5  # item_list entry types
MSG_TYPE_BOT, MSG_STATE_FINISH = 2, 2
TYPING_START, TYPING_STOP = 1, 2
_LIVE_ADAPTERS: Dict[str, Any] = {}
_HEADER_RE = re.compile(r"^(#{1,6})\s+(.+?)\s*$")
_TABLE_RULE_RE = re.compile(r"^\s*\|?(?:\s*:?-{3,}:?\s*\|)+\s*:?-{3,}:?\s*\|?\s*$")
_FENCE_RE = re.compile(r"^```([^\n`]*)\s*$")


def _is_stale_session_ret(ret: "Optional[int]", errcode: "Optional[int]", errmsg: "Optional[str]") -> bool:
    """Recognize stale-session variants of iLink's ``-2`` response, not real rate limits."""
    return (ret == RATE_LIMIT_ERRCODE or errcode == RATE_LIMIT_ERRCODE) and (errmsg or "").lower() in {
        "unknown error",
        "prepare failed",
    }


def _is_session_expired(resp: Dict[str, Any], ret: Any, errcode: Any) -> bool:
    return SESSION_EXPIRED_ERRCODE in (ret, errcode) or _is_stale_session_ret(ret, errcode, resp.get("errmsg") or resp.get("msg"))


def _session_not_ready_error(ret: Any, errcode: Any, errmsg: Any) -> RuntimeError:
    """The stale-session ``-2`` after the tokenless re-send is exhausted (or with no token to drop): iLink will not
    prepare a bot-initiated send until this peer messages the bot again. Deterministic, so it is neither retried nor
    fed to the rate-limit breaker (#80125). The text must not contain "rate limit" — ``classify_send_error`` would
    route it back into the rate-limited redelivery lane."""
    return RuntimeError(
        f"iLink sendmessage session not ready: ret={ret} errcode={errcode} errmsg={errmsg or 'unknown error'}"
        " — the user must send the bot a message first (or re-pair)")


def _make_ssl_connector() -> Optional["aiohttp.TCPConnector"]:
    """TCPConnector with certifi's CA bundle (``ilinkai.weixin.qq.com`` fails some system stores, e.g. Homebrew
    OpenSSL); None without certifi so aiohttp's default (honors ``SSL_CERT_FILE`` under trust_env) applies.
    ``keepalive_timeout=2`` + ``enable_cleanup_closed`` drain idle CLOSE_WAIT sockets behind proxies like Warp.

    Uses a tight ``keepalive_timeout=2`` (default aiohttp: 30s) so idle connections drain promptly behind
    proxies like Cloudflare Warp that leave peer-initiated FIN in ``CLOSE_WAIT`` (same class as #18451).
    ``enable_cleanup_closed=True`` helps the connector clean up sockets that the remote side has already
    closed.
    """
    try:
        import ssl
        import certifi
    except ImportError:
        return None
    if not AIOHTTP_AVAILABLE:
        return None
    return aiohttp.TCPConnector(ssl=ssl.create_default_context(cafile=certifi.where()), keepalive_timeout=2, enable_cleanup_closed=True)


def _new_session(**kwargs: Any) -> "aiohttp.ClientSession":
    return aiohttp.ClientSession(trust_env=gateway_trust_env(), connector=_make_ssl_connector(), **kwargs)


def check_weixin_requirements() -> bool:
    return AIOHTTP_AVAILABLE and CRYPTO_AVAILABLE


def _safe_id(value: Optional[str], keep: int = 8) -> str:
    raw = str(value or "").strip()
    return raw[:keep] if raw else "?"


def _pkcs7_pad(data: bytes, block_size: int = 16) -> bytes:
    pad_len = block_size - (len(data) % block_size)
    return data + bytes([pad_len] * pad_len)


def _aes128_ecb_encrypt(plaintext: bytes, key: bytes) -> bytes:
    encryptor = Cipher(algorithms.AES(key), modes.ECB(), backend=default_backend()).encryptor()
    return encryptor.update(_pkcs7_pad(plaintext)) + encryptor.finalize()


def _aes128_ecb_decrypt(ciphertext: bytes, key: bytes) -> bytes:
    """Decrypt and strip PKCS#7 padding when it is well-formed (else return the raw block output)."""
    decryptor = Cipher(algorithms.AES(key), modes.ECB(), backend=default_backend()).decryptor()
    padded = decryptor.update(ciphertext) + decryptor.finalize()
    pad_len = padded[-1] if padded else 0
    if 1 <= pad_len <= 16 and padded.endswith(bytes([pad_len]) * pad_len):
        return padded[:-pad_len]
    return padded


def _headers(token: Optional[str], body: str) -> Dict[str, str]:
    uin = base64.b64encode(str(int.from_bytes(secrets.token_bytes(4), "big")).encode("utf-8")).decode("ascii")
    return {
        "Content-Type": "application/json", "AuthorizationType": "ilink_bot_token",
        "Content-Length": str(len(body.encode("utf-8"))), "X-WECHAT-UIN": uin,
        "iLink-App-Id": ILINK_APP_ID, "iLink-App-ClientVersion": str(ILINK_APP_CLIENT_VERSION),
        **({"Authorization": f"Bearer {token}"} if token else {}),
    }


def _account_dir(hermes_home: str) -> Path:
    path = Path(hermes_home) / "weixin" / "accounts"
    path.mkdir(parents=True, exist_ok=True)
    return path


def _read_json(path: Path) -> Any:
    try:
        # utf-8-sig (ours): tolerate BOM-persisted JSON files.
        return json.loads(path.read_text(encoding="utf-8-sig")) if path.exists() else None
    except Exception:
        return None


def save_weixin_account(hermes_home: str, *, account_id: str, token: str, base_url: str, user_id: str = "") -> None:
    path = _account_dir(hermes_home) / f"{account_id}.json"
    saved_at = time.strftime("%Y-%m-%dT%H:%M:%SZ", time.gmtime())
    atomic_json_write(path, {"token": token, "base_url": base_url, "user_id": user_id, "saved_at": saved_at})
    with contextlib.suppress(OSError):
        path.chmod(0o600)


def load_weixin_account(hermes_home: str, account_id: str) -> Optional[Dict[str, Any]]:
    return _read_json(_account_dir(hermes_home) / f"{account_id}.json")


class ContextTokenStore:
    """Disk-backed ``context_token`` cache keyed by account + peer."""

    def __init__(self, hermes_home: str):
        self._root = _account_dir(hermes_home)
        self._cache: Dict[str, str] = {}
        # Serializes the offloaded flushes so two concurrent set() calls
        # cannot land their writes out of order (last-writer-wins would drop
        # the newer token from disk).
        self._persist_lock = asyncio.Lock()

    @staticmethod
    def _key(account_id: str, user_id: str) -> str:
        return f"{account_id}:{user_id}"

    def restore(self, account_id: str) -> None:
        path = self._root / f"{account_id}.context-tokens.json"
        if not path.exists():
            return
        try:
            data = json.loads(path.read_text(encoding="utf-8-sig"))
        except Exception as exc:
            logger.warning("weixin: failed to restore context tokens for %s: %s", _safe_id(account_id), exc)
            return
        restored = {self._key(account_id, u): t for u, t in data.items() if isinstance(t, str) and t}
        self._cache.update(restored)
        if restored:
            logger.info("weixin: restored %d context token(s) for %s", len(restored), _safe_id(account_id))

    def get(self, account_id: str, user_id: str) -> Optional[str]:
        return self._cache.get(self._key(account_id, user_id))

    async def set(self, account_id: str, user_id: str, token: str) -> None:
        self._cache[self._key(account_id, user_id)] = token
        # atomic_json_write() fsyncs, so the flush is offloaded off the loop; the payload is snapshotted
        # here (the worker never iterates ``_cache`` mid-mutation) and the lock keeps flushes in order.
        async with self._persist_lock:
            prefix = f"{account_id}:"
            payload = {key[len(prefix):]: value for key, value in self._cache.items() if key.startswith(prefix)}
            await asyncio.to_thread(self._persist, account_id, payload)

    def _persist(self, account_id: str, payload: Dict[str, str]) -> None:
        try:
            atomic_json_write(self._root / f"{account_id}.context-tokens.json", payload)
        except Exception as exc:
            logger.warning("weixin: failed to persist context tokens for %s: %s", _safe_id(account_id), exc)


class TypingTicketCache:
    """Short-lived typing ticket cache from ``getconfig``."""

    def __init__(self, ttl_seconds: float = 600.0):
        self._ttl_seconds = ttl_seconds
        self._cache: Dict[str, Tuple[str, float]] = {}

    def get(self, user_id: str) -> Optional[str]:
        entry = self._cache.get(user_id)
        if entry and time.time() - entry[1] < self._ttl_seconds:
            return entry[0]
        self._cache.pop(user_id, None)
        return None

    def set(self, user_id: str, ticket: str) -> None:
        self._cache[user_id] = (ticket, time.time())


def _parse_aes_key(aes_key_b64: str) -> bytes:
    decoded = base64.b64decode(aes_key_b64)
    if len(decoded) == 16:
        return decoded
    text = decoded.decode("ascii", errors="ignore") if len(decoded) == 32 else ""
    if text and all(ch in "0123456789abcdefABCDEF" for ch in text):
        return bytes.fromhex(text)
    raise ValueError(f"unexpected aes_key format ({len(decoded)} decoded bytes)")


def _guess_chat_type(message: Dict[str, Any], account_id: str) -> Tuple[str, str]:
    room_id = str(message.get("room_id") or message.get("chat_room_id") or "").strip()
    to_user_id = str(message.get("to_user_id") or "").strip()
    if room_id or (to_user_id and account_id and to_user_id != account_id and message.get("msg_type") == 1):
        return "group", room_id or to_user_id or str(message.get("from_user_id") or "")
    return "dm", str(message.get("from_user_id") or "")


# HTTP helpers enforce timeouts via asyncio.wait_for(), not aiohttp ClientTimeout, which raises
# "Timeout context manager should be used inside a task" under run_coroutine_threadsafe() from cron.
async def _api_request(
    session: "aiohttp.ClientSession", method: str, *, base_url: str, endpoint: str, headers: Dict[str, str], timeout_ms: int, body: Optional[str] = None,
) -> Dict[str, Any]:
    async def _do() -> Dict[str, Any]:
        kwargs = {"data": body} if body is not None else {}
        async with getattr(session, method.lower())(f"{base_url.rstrip('/')}/{endpoint}", headers=headers, **kwargs) as response:
            raw = await response.text()
            if not response.ok:
                raise RuntimeError(f"iLink {method} {endpoint} HTTP {response.status}: {raw[:200]}")
            return json.loads(raw)
    return await asyncio.wait_for(_do(), timeout=timeout_ms / 1000)


async def _api_post(
    session: "aiohttp.ClientSession", *, base_url: str, endpoint: str, payload: Dict[str, Any], token: Optional[str], timeout_ms: int,
) -> Dict[str, Any]:
    body = json.dumps({**payload, "base_info": {"channel_version": CHANNEL_VERSION}}, ensure_ascii=False, separators=(",", ":"))
    return await _api_request(session, "POST", base_url=base_url, endpoint=endpoint, headers=_headers(token, body), timeout_ms=timeout_ms, body=body)


async def _api_get(session: "aiohttp.ClientSession", *, base_url: str, endpoint: str, timeout_ms: int) -> Dict[str, Any]:
    headers = {"iLink-App-Id": ILINK_APP_ID, "iLink-App-ClientVersion": str(ILINK_APP_CLIENT_VERSION)}
    return await _api_request(session, "GET", base_url=base_url, endpoint=endpoint, headers=headers, timeout_ms=timeout_ms)


async def _get_updates(session: "aiohttp.ClientSession", *, base_url: str, token: str, sync_buf: str, timeout_ms: int) -> Dict[str, Any]:
    try:
        return await _api_post(session, base_url=base_url, endpoint=EP_GET_UPDATES, payload={"get_updates_buf": sync_buf}, token=token, timeout_ms=timeout_ms)
    except asyncio.TimeoutError:
        return {"ret": 0, "msgs": [], "get_updates_buf": sync_buf}


async def _send_items(
    session: "aiohttp.ClientSession", *, base_url: str, token: str, to: str, item_list: List[Dict[str, Any]], context_token: Optional[str], client_id: str,
) -> Dict[str, Any]:
    message: Dict[str, Any] = {
        "from_user_id": "", "to_user_id": to, "client_id": client_id, "message_type": MSG_TYPE_BOT, "message_state": MSG_STATE_FINISH,
        "item_list": item_list}
    if context_token:
        message["context_token"] = context_token
    return await _api_post(session, base_url=base_url, endpoint=EP_SEND_MESSAGE, payload={"msg": message}, token=token, timeout_ms=API_TIMEOUT_MS)


async def _send_message(
    session: "aiohttp.ClientSession", *, base_url: str, token: str, to: str, text: str, context_token: Optional[str], client_id: str,
) -> Dict[str, Any]:
    if not text or not text.strip():
        raise ValueError("_send_message: text must not be empty")
    item_list = [{"type": ITEM_TEXT, "text_item": {"text": text}}]
    return await _send_items(session, base_url=base_url, token=token, to=to, item_list=item_list, context_token=context_token, client_id=client_id)


async def _get_config(session: "aiohttp.ClientSession", *, base_url: str, token: str, user_id: str, context_token: Optional[str]) -> Dict[str, Any]:
    payload: Dict[str, Any] = {"ilink_user_id": user_id}
    if context_token:
        payload["context_token"] = context_token
    return await _api_post(session, base_url=base_url, endpoint=EP_GET_CONFIG, payload=payload, token=token, timeout_ms=CONFIG_TIMEOUT_MS)


async def _get_upload_url(
    session: "aiohttp.ClientSession", *, base_url: str, token: str, to_user_id: str, media_type: int, filekey: str, rawsize: int,
    rawfilemd5: str, filesize: int, aeskey_hex: str,
) -> Dict[str, Any]:
    payload = {
        "filekey": filekey, "media_type": media_type, "to_user_id": to_user_id, "rawsize": rawsize, "rawfilemd5": rawfilemd5,
        "filesize": filesize, "no_need_thumb": True, "aeskey": aeskey_hex}
    return await _api_post(session, base_url=base_url, endpoint=EP_GET_UPLOAD_URL, payload=payload, token=token, timeout_ms=API_TIMEOUT_MS)


async def _upload_ciphertext(session: "aiohttp.ClientSession", *, ciphertext: bytes, upload_url: str) -> str:
    async def _do() -> str:
        async with session.post(upload_url, data=ciphertext, headers={"Content-Type": "application/octet-stream"}) as response:
            encrypted_param = response.headers.get("x-encrypted-param") if response.status == 200 else None
            if encrypted_param:
                await response.read()
                return encrypted_param
            raw = (await response.text())[:200]
            raise RuntimeError(f"CDN upload missing x-encrypted-param header: {raw}" if response.status == 200 else f"CDN upload HTTP {response.status}: {raw}")
    return await asyncio.wait_for(_do(), timeout=120)


async def _download_bytes(session: "aiohttp.ClientSession", *, url: str, timeout_seconds: float = 60.0) -> bytes:
    async def _do() -> bytes:
        async with session.get(url) as response:
            response.raise_for_status()
            return await response.read()
    return await asyncio.wait_for(_do(), timeout=timeout_seconds)


_WEIXIN_CDN_ALLOWLIST: frozenset[str] = frozenset({
    "novac2c.cdn.weixin.qq.com", "ilinkai.weixin.qq.com", "wx.qlogo.cn", "thirdwx.qlogo.cn", "res.wx.qq.com", "mmbiz.qpic.cn", "mmbiz.qlogo.cn"})


def _assert_weixin_cdn_url(url: str) -> None:
    try:
        parsed = urlparse(url)
        scheme, host = parsed.scheme.lower(), parsed.hostname or ""
    except Exception as exc:  # noqa: BLE001
        raise ValueError(f"Unparseable media URL: {url!r}") from exc
    if scheme not in {"http", "https"}:
        raise ValueError(f"Media URL has disallowed scheme {scheme!r}; only http/https are permitted.")
    if host not in _WEIXIN_CDN_ALLOWLIST:
        raise ValueError(f"Media URL host {host!r} is not in the WeChat CDN allowlist. Refusing to fetch to prevent SSRF.")


async def _download_and_decrypt_media(
    session: "aiohttp.ClientSession", *, cdn_base_url: str, encrypted_query_param: Optional[str], aes_key_b64: Optional[str],
    full_url: Optional[str], timeout_seconds: float,
) -> bytes:
    if encrypted_query_param:
        url = f"{cdn_base_url.rstrip('/')}/download?encrypted_query_param={quote(encrypted_query_param, safe='')}"
    elif full_url:
        _assert_weixin_cdn_url(full_url)
        url = full_url
    else:
        raise RuntimeError("media item had neither encrypt_query_param nor full_url")
    raw = await _download_bytes(session, url=url, timeout_seconds=timeout_seconds)
    return _aes128_ecb_decrypt(raw, _parse_aes_key(aes_key_b64)) if aes_key_b64 else raw


def _walk_markdown_lines(content: str):
    """Yield ``(rstripped line, is_fence, in_code_block)`` per line; ``in_code_block`` is the state *before* a fence toggles it."""
    in_code_block = False
    for raw_line in content.splitlines():
        line = raw_line.rstrip()
        is_fence = bool(_FENCE_RE.match(line.strip()))
        yield line, is_fence, in_code_block
        if is_fence:
            in_code_block = not in_code_block


def _normalize_markdown_blocks(content: str) -> str:
    """rstrip lines and collapse blank runs to one, leaving fenced code untouched."""
    result: List[str] = []
    blank_run = 0
    for line, is_fence, in_code_block in _walk_markdown_lines(content):
        if not is_fence and not in_code_block and not line.strip():
            blank_run += 1
            if blank_run <= 1:
                result.append("")
            continue
        if is_fence or not in_code_block:
            blank_run = 0
        result.append(line)
    return "\n".join(result).strip()


def _wrap_copy_friendly_lines_for_weixin(content: str) -> str:
    """Wrap long display lines that are hard to copy in WeChat clients (not code/table lines)."""
    if not content:
        return content
    wrapped: List[str] = []
    for line, is_fence, in_code_block in _walk_markdown_lines(content):
        stripped = line.strip()
        keep = is_fence or in_code_block or len(line) <= WEIXIN_COPY_LINE_WIDTH or not stripped or stripped.startswith("|")
        if keep or _TABLE_RULE_RE.match(stripped):
            wrapped.append(line)
        else:
            wrapped.extend(textwrap.wrap(
                line, width=WEIXIN_COPY_LINE_WIDTH, break_long_words=False, break_on_hyphens=False, replace_whitespace=False,
                drop_whitespace=True) or [line])
    return "\n".join(wrapped).strip()


def _split_markdown_blocks(content: str) -> List[str]:
    blocks: List[str] = []
    current: List[str] = []

    def flush() -> None:
        if current:
            blocks.append("\n".join(current).strip())
            current.clear()
    for line, is_fence, in_code_block in _walk_markdown_lines(content):
        if is_fence and not in_code_block:  # opening fence starts a fresh block
            flush()
        if is_fence or in_code_block or line.strip():
            current.append(line)
        else:
            flush()
        if is_fence and in_code_block:  # closing fence ends the block
            flush()
    flush()
    return [block for block in blocks if block]


def _split_delivery_units_for_weixin(content: str) -> List[str]:
    """Top-level lines become units; fenced code stays intact; indented continuation lines attach to the previous line."""
    units: List[str] = []
    for block in _split_markdown_blocks(content):
        if _FENCE_RE.match(block.splitlines()[0].strip()):
            units.append(block)
            continue
        current: List[str] = []
        for raw_line in block.splitlines():
            line = raw_line.rstrip()
            if current and line.strip() and raw_line.startswith((" ", "\t")):
                current.append(line)
                continue
            units.append("\n".join(current).strip())
            current = [line] if line.strip() else []
        units.append("\n".join(current).strip())
    return [unit for unit in units if unit]


def _looks_like_chatty_line_for_weixin(line: str) -> bool:
    stripped = line.strip()
    return bool(
        stripped and len(stripped) <= 48 and not line.startswith((" ", "\t")) and not stripped.startswith((">", "-", "*", "【", "#", "|"))
        and not _TABLE_RULE_RE.match(stripped) and not re.match(r"^\*\*[^*]+\*\*$", stripped) and not re.match(r"^\d+\.\s", stripped))


def _should_split_short_chat_block_for_weixin(block: str) -> bool:
    """Split only chat-like multiline blocks (2-6 chatty lines, first line not a heading) into separate bubbles."""
    lines = [line for line in block.splitlines() if line.strip()]
    if not 2 <= len(lines) <= 6:
        return False
    first = lines[0].strip()
    if _HEADER_RE.match(first) or (len(first) <= 24 and first.endswith((":", "："))):
        return False
    return all(_looks_like_chatty_line_for_weixin(line) for line in lines)


def _pack_markdown_blocks_for_weixin(content: str, max_length: int) -> List[str]:
    if len(content) <= max_length:
        return [content]
    # Block extraction stays weixin-local (anchored _FENCE_RE + per-line rstrip); packing is shared.
    overflow = lambda block: BasePlatformAdapter.truncate_message(block, max_length)  # noqa: E731
    return greedy_pack_blocks(_split_markdown_blocks(content), max_length, overflow=overflow)


def _split_text_for_weixin_delivery(content: str, max_length: int, split_per_line: bool = False) -> List[str]:
    """Compact (default): one message when it fits, unless it reads as a short chatty exchange (separate
    bubbles). Per-line (legacy ``extra.split_multiline_messages`` / ``WEIXIN_SPLIT_MULTILINE_MESSAGES``):
    top-level line breaks become separate messages. Oversized units use block-aware packing."""
    if not content:
        return []
    if split_per_line:
        if len(content) <= max_length and "\n" not in content:
            return [content]
        units = _split_delivery_units_for_weixin(content)
        chunks = [c for u in units for c in ([u] if len(u) <= max_length else _pack_markdown_blocks_for_weixin(u, max_length))]
        return [c for c in chunks if c] or [content]
    if len(content) > max_length:
        return _pack_markdown_blocks_for_weixin(content, max_length) or [content]
    return _split_delivery_units_for_weixin(content) if _should_split_short_chat_block_for_weixin(content) else [content]


def _coerce_bool(value: Any, default: bool = True) -> bool:
    """Coerce a config value to bool, tolerating strings like ``"true"``; unknown -> ``default``."""
    if value is None:
        return default
    if isinstance(value, (int, float)):
        return bool(value)
    text = str(value).strip().lower()
    return True if text in {"1", "true", "yes", "on"} else False if text in {"0", "false", "no", "off"} else default


def _extract_text(item_list: List[Dict[str, Any]]) -> str:
    for item in item_list:
        if item.get("type") == ITEM_TEXT:
            text = str((item.get("text_item") or {}).get("text") or "")
            ref = item.get("ref_msg") or {}
            ref_item = ref.get("message_item") or {}
            if ref_item.get("type") in {ITEM_IMAGE, ITEM_VIDEO, ITEM_FILE, ITEM_VOICE}:
                title = ref.get("title") or ""
                return f"[引用媒体: {title}]\n{text}".strip() if title else f"[引用媒体]\n{text}".strip()
            if ref_item:
                parts = [p for p in (str(ref["title"]) if ref.get("title") else "", _extract_text([ref_item])) if p]
                if parts:
                    return f"[引用: {' | '.join(parts)}]\n{text}".strip()
            return text
    for item in item_list:
        if item.get("type") == ITEM_VOICE:
            # Tencent's ``voice_item.text`` is their STT output and is wrong for non-Chinese audio.
            # When raw audio exists return "" so gateway/run.py's central STT transcribes the download;
            # otherwise use Weixin's transcript but mark its voice origin.
            # #27300: Tencent Cloud's `voice_item.text` is their STT output, which is wrong for any
            # non-Chinese audio (the original report was a Russian voice message that came back as English
            # gibberish). Return empty so the central STT pipeline in ``gateway/run.py`` produces the body
            # from the downloaded audio instead.
            voice_item = item.get("voice_item") or {}
            # Use it, but preserve the voice origin so the agent can distinguish this from text the user
            # typed (#65022).
            voice_text = str(voice_item.get("text") or "")
            if not (voice_item.get("media") or {}) and voice_text:
                return f"[Voice transcription provided by Weixin]\n{voice_text}"
    return ""


_MIME_PREFIX_TYPES = (("image/", MessageType.PHOTO), ("video/", MessageType.VIDEO), ("audio/", MessageType.VOICE))


def _message_type_from_media(media_types: List[str], text: str) -> MessageType:
    for prefix, message_type in _MIME_PREFIX_TYPES:
        if any(m.startswith(prefix) for m in media_types):
            return message_type
    return MessageType.DOCUMENT if media_types else MessageType.COMMAND if text.startswith("/") else MessageType.TEXT


def _load_sync_buf(hermes_home: str, account_id: str) -> str:
    data = _read_json(_account_dir(hermes_home) / f"{account_id}.sync.json")
    return data.get("get_updates_buf", "") if isinstance(data, dict) else ""


def _save_sync_buf(hermes_home: str, account_id: str, sync_buf: str) -> None:
    atomic_json_write(_account_dir(hermes_home) / f"{account_id}.sync.json", {"get_updates_buf": sync_buf})


async def _fetch_qr(session: "aiohttp.ClientSession", bot_type: str) -> Tuple[str, str]:
    qr_resp = await _api_get(session, base_url=ILINK_BASE_URL, endpoint=f"{EP_GET_BOT_QR}?bot_type={bot_type}", timeout_ms=QR_TIMEOUT_MS)
    return str(qr_resp.get("qrcode") or ""), str(qr_resp.get("qrcode_img_content") or "")


def _print_qr(qrcode_value: str, qrcode_url: str, *, report_render_error: bool) -> None:
    """Print the QR URL + ASCII render; WeChat must scan the liteapp URL (``qrcode_img_content``), not the bare hex token."""
    if qrcode_url:
        print(qrcode_url)
    try:
        import qrcode
        qr = qrcode.QRCode()
        qr.add_data(qrcode_url or qrcode_value)
        qr.make(fit=True)
        qr.print_ascii(invert=True)
    except Exception as _qr_exc:
        if report_render_error:
            print(f"（终端二维码渲染失败: {_qr_exc}，请直接打开上面的二维码链接）")


async def qr_login(hermes_home: str, *, bot_type: str = "3", timeout_seconds: int = 480) -> Optional[Dict[str, str]]:
    if not AIOHTTP_AVAILABLE:
        raise RuntimeError("aiohttp is required for Weixin QR login")
    async with _new_session() as session:
        try:
            qrcode_value, qrcode_url = await _fetch_qr(session, bot_type)
        except Exception as exc:
            logger.error("weixin: failed to fetch QR code: %s", exc)
            return None
        if not qrcode_value:
            logger.error("weixin: QR response missing qrcode")
            return None
        print("\n请使用微信扫描以下二维码：")
        _print_qr(qrcode_value, qrcode_url, report_render_error=True)
        deadline = time.monotonic() + timeout_seconds
        current_base_url, refresh_count = ILINK_BASE_URL, 0
        while time.monotonic() < deadline:
            try:
                status_resp = await _api_get(session, base_url=current_base_url, endpoint=f"{EP_GET_QR_STATUS}?qrcode={qrcode_value}", timeout_ms=QR_TIMEOUT_MS)
            except Exception as exc:
                if not isinstance(exc, asyncio.TimeoutError):
                    logger.warning("weixin: QR poll error: %s", exc)
                await asyncio.sleep(1)
                continue
            status = str(status_resp.get("status") or "wait")
            if status == "wait":
                print(".", end="", flush=True)
            elif status == "scaned":
                print("\n已扫码，请在微信里确认...")
            elif status == "scaned_but_redirect" and status_resp.get("redirect_host"):
                current_base_url = f"https://{status_resp['redirect_host']}"
            elif status == "expired":
                refresh_count += 1
                if refresh_count > 3:
                    print("\n二维码多次过期，请重新执行登录。")
                    return None
                print(f"\n二维码已过期，正在刷新... ({refresh_count}/3)")
                try:
                    qrcode_value, qrcode_url = await _fetch_qr(session, bot_type)
                    _print_qr(qrcode_value, qrcode_url, report_render_error=False)
                except Exception as exc:
                    logger.error("weixin: QR refresh failed: %s", exc)
                    return None
            elif status == "confirmed":
                creds = {
                    "account_id": str(status_resp.get("ilink_bot_id") or ""), "token": str(status_resp.get("bot_token") or ""),
                    "base_url": str(status_resp.get("baseurl") or ILINK_BASE_URL), "user_id": str(status_resp.get("ilink_user_id") or "")}
                if not creds["account_id"] or not creds["token"]:
                    logger.error("weixin: QR confirmed but credential payload was incomplete")
                    return None
                save_weixin_account(hermes_home, **creds)
                print(f"\n微信连接成功，account_id={creds['account_id']}")
                return creds
            await asyncio.sleep(1)
        print("\n微信登录超时。")
        return None


# Outbound item shapes: item type -> (item key, extra fields after the shared ``media`` block).
_ITEM_SHAPES: Dict[int, Tuple[str, Callable[[Dict[str, Any]], Dict[str, Any]]]] = {
    ITEM_FILE: ("file_item", lambda kw: {"file_name": kw["filename"], "len": str(kw["plaintext_size"])}),
    ITEM_IMAGE: ("image_item", lambda kw: {"mid_size": kw["ciphertext_size"]}),
    ITEM_VIDEO: ("video_item", lambda kw: {
        "video_size": kw["ciphertext_size"], "play_length": kw.get("play_length", 0), "video_md5": kw.get("rawfilemd5", "")}),
    ITEM_VOICE: ("voice_item", lambda kw: {
        "encode_type": kw.get("encode_type"), "bits_per_sample": kw.get("bits_per_sample"), "sample_rate": kw.get("sample_rate"),
        "playtime": kw.get("playtime", 0)}),
}


def _media_item(item_type: int, **kw: Any) -> Dict[str, Any]:
    key, fields = _ITEM_SHAPES[item_type]
    media = {"encrypt_query_param": kw["encrypt_query_param"], "aes_key": kw["aes_key_for_api"], "encrypt_type": 1}
    return {"type": item_type, key: {"media": media, **fields(kw)}}


_file_item, _image_item = partial(_media_item, ITEM_FILE), partial(_media_item, ITEM_IMAGE)
_video_item, _voice_item = partial(_media_item, ITEM_VIDEO), partial(_media_item, ITEM_VOICE)


# Inbound media dispatch: item type -> (item key, download timeout, cache fn, mime or None (= guess from
# file_name), log label). Cache fns are lambdas so monkeypatching the module names takes effect at call time.
_INBOUND_MEDIA: Dict[int, Tuple[str, float, Callable[[bytes, str], Awaitable[str]], Optional[str], str]] = {
    ITEM_IMAGE: ("image_item", 30.0, lambda data, _name: cache_image_from_bytes_async(data, ".jpg"), "image/jpeg", "image"),
    ITEM_VIDEO: ("video_item", 120.0, lambda data, _name: cache_document_from_bytes_async(data, "video.mp4"), "video/mp4", "video"),
    ITEM_FILE: ("file_item", 60.0, lambda data, name: cache_document_from_bytes_async(data, name), None, "file"),
    ITEM_VOICE: ("voice_item", 60.0, lambda data, _name: cache_audio_from_bytes_async(data, ".silk"), "audio/silk", "voice"),
}

# Outbound local-file dispatch by extension: (extensions, sender method, path kwarg); default = send_document.
_OUTBOUND_BY_EXT: Tuple[Tuple[frozenset, str, str], ...] = (
    (frozenset({".ogg", ".opus", ".mp3", ".wav", ".m4a", ".flac"}), "send_voice", "audio_path"),  # no .m2a, unlike base
    (_VIDEO_EXTS, "send_video", "video_path"),
    (_IMAGE_EXTS, "send_image_file", "image_path"),
)
_DIRECT_IMAGE_EXTS = {".jpg", ".jpeg", ".png", ".gif", ".webp", ".bmp"}


class WeixinAdapter(OwnAccessPolicyMixin, BasePlatformAdapter):
    ALLOW_ALL_ENV_PREFIX = "WEIXIN"
    supports_code_blocks = True  # Weixin renders fenced code blocks
    splits_long_messages = True  # send() chunks via _split_text()
    MAX_MESSAGE_LENGTH = 2000
    SUPPORTS_MESSAGE_EDITING = False  # WeChat cannot edit; streaming must send-final-only so the cursor (▉) is never left visible
    _SPLIT_THRESHOLD = 1800  # iLink chunks at ~2048 chars

    def __init__(self, config: PlatformConfig):
        super().__init__(config, Platform.WEIXIN)
        extra = config.extra or {}
        self._hermes_home = hermes_home = str(get_hermes_home())
        self._token_store = ContextTokenStore(hermes_home)
        self._typing_cache = TypingTicketCache()
        self._poll_session = self._send_session = None  # type: Optional[aiohttp.ClientSession]
        self._poll_task: Optional[asyncio.Task] = None
        self._dedup = MessageDeduplicator(ttl_seconds=MESSAGE_DEDUP_TTL_SECONDS)
        self._account_id = _extra_or_secret(extra, "account_id")
        self._token = str(config.token or extra.get("token") or _wx_secret("WEIXIN_TOKEN", "")).strip()
        self._base_url = _extra_or_secret(extra, "base_url", ILINK_BASE_URL).rstrip("/")
        self._cdn_base_url = _extra_or_secret(extra, "cdn_base_url", WEIXIN_CDN_BASE_URL).rstrip("/")
        # Tunables: ``extra.<key>`` else env ``WEIXIN_<KEY>`` (e.g. WEIXIN_SEND_CHUNK_RETRIES).
        self._send_chunk_delay_seconds = float(_extra_or_secret(extra, "send_chunk_delay_seconds", "1.5"))
        self._send_chunk_retries = int(_extra_or_secret(extra, "send_chunk_retries", "4"))
        self._send_chunk_retry_delay_seconds = float(_extra_or_secret(extra, "send_chunk_retry_delay_seconds", "1.0"))
        self._send_text_gate = asyncio.Lock()
        self._rate_limit_circuit_threshold = max(1, int(_extra_or_secret(extra, "rate_limit_circuit_threshold", "1")))
        self._rate_limit_circuit_window_seconds = float(_extra_or_secret(extra, "rate_limit_circuit_window_seconds", "30.0"))
        self._rate_limit_circuit_open_seconds = float(_extra_or_secret(extra, "rate_limit_circuit_open_seconds", "30.0"))
        self._rate_limit_circuit_until, self._rate_limit_events = 0.0, []  # type: float, List[float]
        self._dm_policy = _extra_or_secret(extra, "dm_policy", "pairing").lower()
        self._group_policy = _extra_or_secret(extra, "group_policy", "disabled").lower()
        # ``extra`` wins even when falsy (an explicit empty list disables the env allowlist).
        allow_from, group_allow_from = extra.get("allow_from"), extra.get("group_allow_from")
        self._allow_from = self._coerce_list(_wx_secret("WEIXIN_ALLOWED_USERS", "") if allow_from is None else allow_from)
        self._group_allow_from = self._coerce_list(_wx_secret("WEIXIN_GROUP_ALLOWED_USERS", "") if group_allow_from is None else group_allow_from)
        self._split_multiline_messages = _coerce_bool(_extra_or_secret(extra, "split_multiline_messages", ""), default=False)
        # Text debounce batching (Telegram pattern): iLink delivers messages individually, so rapid bursts would each
        # trigger a separate agent run. Telegram cadence and ceilings (#44883); ``0`` dispatches immediately.
        self._configure_text_batch_delays()
        persisted = load_weixin_account(hermes_home, self._account_id) if self._account_id and not self._token else None
        if persisted:
            self._token = str(persisted.get("token") or "").strip()
            self._base_url = str(persisted.get("base_url") or self._base_url).strip().rstrip("/")

    @staticmethod
    def _coerce_list(value: Any) -> List[str]:
        if value is None:
            return []
        items = value.split(",") if isinstance(value, str) else value if isinstance(value, (list, tuple, set)) else [value]
        return [str(item).strip() for item in items if str(item).strip()]

    async def connect(self, *, is_reconnect: bool = False) -> bool:
        preflight = (
            (check_weixin_requirements(), "weixin_missing_dependency", "aiohttp and cryptography are required"),
            (self._token, "weixin_missing_token", "WEIXIN_TOKEN is required"),
            (self._account_id, "weixin_missing_account", "WEIXIN_ACCOUNT_ID is required"))
        for ok, code, reason in preflight:
            if not ok:
                self._set_fatal_error(code, f"Weixin startup failed: {reason}", retryable=False)
                logger.warning("[%s] Weixin startup failed: %s", self.name, reason)
                return False
        try:
            if not self._acquire_platform_lock('weixin-bot-token', self._token, 'Weixin bot token'):
                return False
        except Exception as exc:
            logger.debug("[%s] Token lock unavailable (non-fatal): %s", self.name, exc)
        self._poll_session = _new_session()
        # total=None disables aiohttp's ClientTimeout so send() works via run_coroutine_threadsafe() from cron;
        # _api_post/_api_get enforce timeouts with asyncio.wait_for() instead.
        self._send_session = _new_session(timeout=aiohttp.ClientTimeout(total=None, connect=None, sock_connect=None, sock_read=None))
        self._token_store.restore(self._account_id)
        self._poll_task = asyncio.create_task(self._poll_loop(), name="weixin-poll")
        self._mark_connected()
        _LIVE_ADAPTERS[self._token] = self
        logger.info("[%s] Connected account=%s base=%s", self.name, _safe_id(self._account_id), self._base_url)
        if self._group_policy != "disabled":
            logger.warning(
                "[%s] WEIXIN_GROUP_POLICY=%s is set, but QR-login connects an iLink bot identity (e.g. ...@im.bot) "
                "which typically cannot be invited into ordinary WeChat groups. iLink usually does not deliver "
                "ordinary-group events for these accounts, so group messages may never reach Hermes regardless of "
                "this policy. If group delivery doesn't work, the limitation is on the iLink side, not in Hermes.",
                self.name, self._group_policy)
        self._wire_plugin_handlers(None)  # plugin-registered native handlers
        return True

    async def disconnect(self) -> None:
        _LIVE_ADAPTERS.pop(self._token, None)
        self._running = False
        for task in self._pending_text_batch_tasks.values():
            if not task.done():
                task.cancel()
        self._pending_text_batches.clear()
        self._pending_text_batch_tasks.clear()
        await cancel_task(self._poll_task)
        self._poll_task = None
        for attr in ("_poll_session", "_send_session"):
            session = getattr(self, attr)
            if session and not session.closed:
                await session.close()
            setattr(self, attr, None)
        self._release_platform_lock()
        self._mark_disconnected()
        logger.info("[%s] Disconnected", self.name)

    async def _poll_loop(self) -> None:
        assert self._poll_session is not None
        sync_buf = _load_sync_buf(self._hermes_home, self._account_id)
        timeout_ms = LONG_POLL_TIMEOUT_MS
        consecutive_failures = 0

        async def backoff() -> int:
            """Sleep for the failure streak; returns the new streak count (0 after a full streak)."""
            streak_done = consecutive_failures >= MAX_CONSECUTIVE_FAILURES
            await asyncio.sleep(BACKOFF_DELAY_SECONDS if streak_done else RETRY_DELAY_SECONDS)
            return 0 if streak_done else consecutive_failures

        while self._running:
            try:
                response = await _get_updates(self._poll_session, base_url=self._base_url, token=self._token, sync_buf=sync_buf, timeout_ms=timeout_ms)
                suggested_timeout = response.get("longpolling_timeout_ms")
                if isinstance(suggested_timeout, int) and suggested_timeout > 0:
                    timeout_ms = suggested_timeout
                ret, errcode = response.get("ret", 0), response.get("errcode", 0)
                if ret not in {0, None} or errcode not in {0, None}:
                    if _is_session_expired(response, ret, errcode):
                        logger.error("[%s] Session expired; pausing for 10 minutes", self.name)
                        await asyncio.sleep(600)
                        consecutive_failures = 0
                        continue
                    consecutive_failures += 1
                    logger.warning("[%s] getUpdates failed ret=%s errcode=%s errmsg=%s (%d/%d)", self.name, ret, errcode,
                                   response.get("errmsg", ""), consecutive_failures, MAX_CONSECUTIVE_FAILURES)
                    consecutive_failures = await backoff()
                    continue
                consecutive_failures = 0
                # Dispatch before persisting: the off-loop write is an await, and a disconnect that
                # cancels it must not leave the advanced cursor on disk with this batch undelivered.
                for message in response.get("msgs") or []:
                    asyncio.create_task(self._process_message_safe(message))
                # atomic_json_write fsyncs + renames: persist off the loop, and only when the cursor
                # moved (an empty long-poll echoes the same buffer back every cycle).
                if response.get("get_updates_buf") and str(response["get_updates_buf"]) != sync_buf:
                    sync_buf = str(response["get_updates_buf"])
                    await asyncio.to_thread(_save_sync_buf, self._hermes_home, self._account_id, sync_buf)
            except asyncio.CancelledError:
                break
            except Exception as exc:
                consecutive_failures += 1
                logger.error("[%s] poll error (%d/%d): %s", self.name, consecutive_failures, MAX_CONSECUTIVE_FAILURES, exc)
                consecutive_failures = await backoff()
                if consecutive_failures == 0:
                    # Full failure streak: recycle the session. Failed connects through a local proxy (e.g.
                    # Clash) strand sockets the keepalive reaper never sees; on macOS the 256-fd soft limit
                    # then yields EMFILE and a crash. Closing the session tears down every socket.
                    # Clash on 127.0.0.1:7890) can strand sockets that never return to the connector's
                    # keepalive pool, so the tight keepalive_timeout never reaps them. On macOS the default
                    # 256-fd soft limit turns that drip into `[Errno 24] Too many open files` and a gateway
                    # crash (#79889). Closing the session tears down its connector and every socket it
                    # holds; a fresh session starts the next attempt from zero fds.
                    await self._recycle_poll_session()

    async def _recycle_poll_session(self) -> None:
        """Swap in a fresh ``_poll_session`` *then* close the old one, so concurrent tasks never see a closed session."""
        if not self._running or aiohttp is None:
            return
        old, self._poll_session = self._poll_session, _new_session()
        if old is not None and not old.closed:
            try:
                await old.close()
            except Exception as exc:
                logger.debug("[%s] old poll session close failed: %s", self.name, exc)

    async def _process_message_safe(self, message: Dict[str, Any]) -> None:
        try:
            await self._process_message(message)
        except Exception as exc:
            logger.error("[%s] unhandled inbound error from=%s: %s", self.name, _safe_id(message.get("from_user_id")), exc, exc_info=True)

    async def _process_message(self, message: Dict[str, Any]) -> None:
        assert self._poll_session is not None
        sender_id = str(message.get("from_user_id") or "").strip()
        message_id = str(message.get("message_id") or "").strip()
        if not sender_id or sender_id == self._account_id or (message_id and self._dedup.is_duplicate(message_id)):
            return
        # Secondary content-fingerprint dedup: upstream re-sends identical text under new message_ids.
        item_list = message.get("item_list") or []
        text = _extract_text(item_list)
        if text and self._dedup.is_duplicate(f"content:{sender_id}:{hashlib.md5(text.encode()).hexdigest()}"):
            logger.debug("[%s] Content-dedup: skipping duplicate message from %s", self.name, sender_id)
            return
        chat_type, effective_chat_id = _guess_chat_type(message, self._account_id)
        if chat_type == "group":
            if not self._is_group_allowed(effective_chat_id):
                return
        elif not self._is_dm_intake_allowed(sender_id):
            return
        context_token = str(message.get("context_token") or "").strip()
        if context_token:
            await self._token_store.set(self._account_id, sender_id, context_token)
        if self._poll_session and self._token and not self._typing_cache.get(sender_id):
            asyncio.create_task(self._fetch_typing_ticket(self._poll_session, sender_id, context_token or None, "getConfig failed"))
        media_paths, media_types = [], []  # type: List[str], List[str]
        for item in item_list:
            ref_item = (item.get("ref_msg") or {}).get("message_item")
            for candidate in (item, ref_item) if isinstance(ref_item, dict) else (item,):
                await self._collect_media(candidate, media_paths, media_types)
        if not text and not media_paths:
            return
        source = self.build_source(chat_id=effective_chat_id, chat_type=chat_type, user_id=sender_id, user_name=sender_id)
        event = MessageEvent(
            text=text, message_type=_message_type_from_media(media_types, text), source=source, raw_message=message,
            message_id=message_id or None, media_urls=media_paths, media_types=media_types, timestamp=datetime.now())
        logger.info("[%s] inbound from=%s type=%s media=%d", self.name, _safe_id(sender_id), source.chat_type, len(media_paths))
        if event.message_type == MessageType.TEXT:
            self._enqueue_text_event(event)
        else:
            await self.handle_message(event)

    async def _collect_media(self, item: Dict[str, Any], media_paths: List[str], media_types: List[str]) -> None:
        spec = _INBOUND_MEDIA.get(item.get("type"))
        path, mime = await self._download_media(item, spec) if spec else (None, "")
        if path:
            media_paths.append(path)
            media_types.append(mime)

    async def _download_media(self, item: Dict[str, Any], spec: Tuple[Any, ...]) -> Tuple[Optional[str], str]:
        """Download + decrypt one inbound media item -> (cached path or None, mime). Voice is always downloaded
        (never trust Tencent's ``voice_item.text``) so gateway/run.py's central STT re-transcribes."""
        item_key, timeout_seconds, cache_fn, mime, label = spec
        payload = item.get(item_key) or {}
        media = payload.get("media") or {}
        filename = str(payload.get("file_name") or "document.bin")
        mime = mime or mimetypes.guess_type(filename)[0] or "application/octet-stream"
        try:
            aes_key_b64 = media.get("aes_key")
            if item_key == "image_item" and payload.get("aeskey"):  # image_item may carry a raw hex ``aeskey`` beside the media block
                aes_key_b64 = base64.b64encode(bytes.fromhex(str(payload.get("aeskey")))).decode("ascii") or aes_key_b64
            # #27300: previously short-circuited when ``voice_item.text`` was set on the assumption that
            # Tencent Cloud's STT was good enough. For non-Chinese audio that text is garbage (e.g. a
            # Russian message comes back as English phonemes) — we must always download the raw audio so
            # ``gateway/run.py``'s central STT pipeline can re-transcribe with the user's configured
            # mlx-whisper / whisper.cpp / faster-whisper backend.
            data = await _download_and_decrypt_media(
                self._poll_session, cdn_base_url=self._cdn_base_url, encrypted_query_param=media.get("encrypt_query_param"),
                aes_key_b64=aes_key_b64, full_url=media.get("full_url"), timeout_seconds=timeout_seconds)
            return await cache_fn(data, filename), mime
        except Exception as exc:
            logger.warning("[%s] %s download failed: %s", self.name, label, exc)
            return None, mime

    async def _fetch_typing_ticket(self, session: Any, user_id: str, context_token: Optional[str], failure_label: str) -> Optional[str]:
        try:
            response = await _get_config(session, base_url=self._base_url, token=self._token, user_id=user_id, context_token=context_token)
            typing_ticket = str(response.get("typing_ticket") or "")
            if typing_ticket:
                self._typing_cache.set(user_id, typing_ticket)
                return typing_ticket
        except Exception as exc:
            logger.debug("[%s] %s for %s: %s", self.name, failure_label, _safe_id(user_id), exc)
        return None

    def _split_text(self, content: str) -> List[str]:
        return _split_text_for_weixin_delivery(content, self.MAX_MESSAGE_LENGTH, self._split_multiline_messages)

    def _rate_limit_cooldown_remaining(self) -> float:
        return max(0.0, self._rate_limit_circuit_until - time.monotonic())

    def _record_rate_limit_event(self) -> bool:
        now = time.monotonic()
        self._rate_limit_events = [ts for ts in self._rate_limit_events if ts >= now - self._rate_limit_circuit_window_seconds] + [now]
        if len(self._rate_limit_events) < self._rate_limit_circuit_threshold:
            return False
        if self._rate_limit_circuit_open_seconds > 0:
            self._rate_limit_circuit_until = max(self._rate_limit_circuit_until, time.monotonic() + self._rate_limit_circuit_open_seconds)
        return self._rate_limit_cooldown_remaining() > 0

    async def _send_text_chunk(self, *, chat_id: str, chunk: str, context_token: Optional[str], client_id: str) -> None:
        """Send one text chunk with retry/backoff under the adapter-wide text gate. A stale-session response (``-14``,
        or ``-2`` with ``prepare failed``/``unknown error``) is re-sent once *without* ``context_token`` — iLink accepts
        tokenless sends as a degraded fallback, which keeps cron pushes working when no user message refreshed the
        session. A ``-2`` that survives that fails fast via ``_session_not_ready_error``."""
        async with self._send_text_gate:
            last_error: Optional[Exception] = None
            retried_without_token = False
            attempt = 0  # counts real failures only — the tokenless re-send must not eat the retry budget
            while True:
                if self._rate_limit_cooldown_remaining() > 0:
                    raise RuntimeError(f"iLink sendmessage rate limited; cooldown active for {self._rate_limit_cooldown_remaining():.1f}s")
                try:
                    resp = await _send_message(
                        self._send_session, base_url=self._base_url, token=self._token, to=chat_id, text=chunk,
                        context_token=context_token, client_id=client_id)
                    ret, errcode = (resp.get("ret"), resp.get("errcode")) if resp and isinstance(resp, dict) else (None, None)
                    if (ret is not None and ret != 0) or (errcode is not None and errcode != 0):
                        errmsg = resp.get("errmsg") or resp.get("msg")
                        if _is_session_expired(resp, ret, errcode) and not retried_without_token and context_token:
                            retried_without_token, context_token = True, None
                            self._token_store._cache.pop(self._token_store._key(self._account_id, chat_id), None)
                            logger.warning("[%s] session expired for %s; retrying without context_token", self.name, _safe_id(chat_id))
                            continue
                        if _is_stale_session_ret(ret, errcode, errmsg):
                            # break, not raise: a raise here is caught below and re-enters the retry ladder.
                            last_error = _session_not_ready_error(ret, errcode, errmsg)
                            break
                        if ret != RATE_LIMIT_ERRCODE and errcode != RATE_LIMIT_ERRCODE:
                            raise RuntimeError(f"iLink sendmessage error: ret={ret} errcode={errcode} errmsg={errmsg or 'unknown error'}")
                        # Keep a descriptive error for when the loop exhausts while still limited.
                        last_error = RuntimeError(f"iLink sendmessage rate limited: ret={ret} errcode={errcode} errmsg={errmsg or 'rate limited'}")
                        if self._record_rate_limit_event():
                            last_error = RuntimeError(
                                f"iLink sendmessage rate limited (ret={ret} errcode={errcode} errmsg={errmsg or 'rate limited'}); "
                                f"cooldown active for {self._rate_limit_cooldown_remaining():.1f}s")
                            break
                        if attempt >= self._send_chunk_retries:
                            break
                        attempt += 1
                        wait = self._send_chunk_retry_delay_seconds * 3  # 3x backoff for rate limit
                        logger.warning("[%s] rate limited for %s; backing off %.1fs before retry", self.name, _safe_id(chat_id), wait)
                        await asyncio.sleep(wait)
                        continue
                    self._rate_limit_events.clear()
                    self._rate_limit_circuit_until = 0.0
                    return
                except Exception as exc:
                    last_error = exc
                    if attempt >= self._send_chunk_retries:
                        break
                    attempt += 1
                    wait = self._send_chunk_retry_delay_seconds * attempt
                    logger.warning("[%s] send chunk failed to=%s attempt=%d/%d, retrying in %.2fs: %s",
                                   self.name, _safe_id(chat_id), attempt, self._send_chunk_retries + 1, wait, exc)
                    if wait > 0:
                        await asyncio.sleep(wait)
            assert last_error is not None
            raise last_error

    async def send(self, chat_id: str, content: str, reply_to: Optional[str] = None, metadata: Optional[Dict[str, Any]] = None) -> SendResult:
        if not self._send_session or not self._token:
            return SendResult(success=False, error="Not connected")
        context_token = self._token_store.get(self._account_id, chat_id)
        last_message_id: Optional[str] = None
        # Extract MEDIA: tags and bare local file paths before text delivery, under the routed
        # profile's scope: Docker MEDIA translation infers the sandbox from the active profile (#109024).
        with self._media_delivery_scope(self.build_source(chat_id=chat_id)):
            media_files, cleaned_content = self.extract_media(content)
            local_files, final_content = self.extract_local_files(self.extract_images(cleaned_content)[1])
            deliveries = [(p, v, "media") for p, v in self.filter_media_delivery_paths(media_files)]
            deliveries += [(p, False, "local file") for p in self.filter_local_delivery_paths(local_files)]
        try:
            for path, is_voice, label in deliveries:
                ext = Path(path).suffix.lower()
                sender, key = next(((m, k) for exts, m, k in _OUTBOUND_BY_EXT if is_voice or ext in exts), ("send_document", "file_path"))
                try:
                    await getattr(self, sender)(chat_id=chat_id, metadata=metadata, **{key: path})
                except Exception as exc:
                    logger.warning("[%s] %s delivery failed for %s: %s", self.name, label, path, exc)
            chunks = [c for c in self._split_text(self.format_message(final_content)) if c and c.strip()]
            for idx, chunk in enumerate(chunks):
                client_id = f"hermes-weixin-{uuid.uuid4().hex}"
                await self._send_text_chunk(chat_id=chat_id, chunk=chunk, context_token=context_token, client_id=client_id)
                last_message_id = client_id
                if idx < len(chunks) - 1 and self._send_chunk_delay_seconds > 0:
                    await asyncio.sleep(self._send_chunk_delay_seconds)
            return SendResult(success=True, message_id=last_message_id)
        except Exception as exc:
            logger.error("[%s] send failed to=%s: %s", self.name, _safe_id(chat_id), exc)
            return SendResult(success=False, error=str(exc))

    async def _ensure_typing_ticket(self, chat_id: str) -> Optional[str]:
        """Return a valid typing ticket, refreshing via getConfig once the 600s TTL evicts it —
        otherwise ``stop_typing`` no-ops and the WeChat client shows the indicator forever."""
        ticket = self._typing_cache.get(chat_id)
        if ticket or not self._send_session or not self._token:
            return ticket or None
        return await self._fetch_typing_ticket(
            self._send_session, chat_id, self._token_store.get(self._account_id, chat_id), "typing ticket refresh failed")

    async def _set_typing(self, chat_id: str, status: int, label: str) -> None:
        typing_ticket = await self._ensure_typing_ticket(chat_id)
        if not typing_ticket:
            return
        try:
            await _api_post(
                self._send_session, base_url=self._base_url, endpoint=EP_SEND_TYPING, token=self._token, timeout_ms=CONFIG_TIMEOUT_MS,
                payload={"ilink_user_id": chat_id, "typing_ticket": typing_ticket, "status": status})
        except Exception as exc:
            logger.debug("[%s] typing %s failed for %s: %s", self.name, label, _safe_id(chat_id), exc)

    async def send_typing(self, chat_id: str, metadata: Optional[Dict[str, Any]] = None) -> None:
        await self._set_typing(chat_id, TYPING_START, "start")

    async def stop_typing(self, chat_id: str) -> None:
        await self._set_typing(chat_id, TYPING_STOP, "stop")

    async def send_image(self, chat_id: str, image_url: str, caption: str, reply_to: Optional[str] = None, metadata=None) -> SendResult:
        cleanup = image_url.startswith(("http://", "https://"))
        file_path = await self._download_remote_media(image_url) if cleanup else image_url.replace("file://", "")
        if not cleanup and not os.path.isabs(file_path):
            file_path = os.path.abspath(file_path)
        try:
            return await self.send_document(chat_id, file_path, caption=caption, metadata=metadata)
        finally:
            if cleanup and file_path and os.path.exists(file_path):
                with contextlib.suppress(OSError):
                    os.unlink(file_path)

    async def send_image_file(self, chat_id: str, image_path: str, caption: Optional[str] = None, reply_to=None, metadata=None, **kwargs) -> SendResult:
        return await self.send_document(chat_id=chat_id, file_path=image_path, caption=caption, metadata=metadata)

    async def _send_file_result(self, chat_id: str, path: str, caption: str, label: str, **kwargs: Any) -> SendResult:
        if not self._send_session or not self._token:
            return SendResult(success=False, error="Not connected")
        try:
            return SendResult(success=True, message_id=await self._send_file(chat_id, path, caption, **kwargs))
        except Exception as exc:
            logger.error("[%s] %s failed to=%s: %s", self.name, label, _safe_id(chat_id), exc)
            return SendResult(success=False, error=str(exc))

    async def send_document(
        self, chat_id: str, file_path: str, caption: Optional[str] = None, file_name: Optional[str] = None, reply_to: Optional[str] = None,
        metadata: Optional[Dict[str, Any]] = None, **kwargs,
    ) -> SendResult:
        return await self._send_file_result(chat_id, file_path, caption or "", "send_document")

    async def send_video(self, chat_id: str, video_path: str, caption: Optional[str] = None, reply_to=None, metadata=None) -> SendResult:
        return await self._send_file_result(chat_id, video_path, caption or "", "send_video")

    async def send_voice(self, chat_id: str, audio_path: str, caption: Optional[str] = None, reply_to=None, metadata=None, **kwargs) -> SendResult:
        # Native outbound voice bubbles are not proven-working upstream; a file attachment at least plays (even .silk).
        return await self._send_file_result(chat_id, audio_path, caption or self.warning_text(t("platform.weixin.voice_as_attachment")), "send_voice", force_file_attachment=True)

    async def _download_remote_media(self, url: str) -> str:
        from tools.url_safety import is_safe_url
        if not is_safe_url(url):
            raise ValueError(f"Blocked unsafe URL (SSRF protection): {url}")
        assert self._send_session is not None
        data = await _download_bytes(self._send_session, url=url, timeout_seconds=30)
        with tempfile.NamedTemporaryFile(delete=False, suffix=Path(url.split("?", 1)[0]).suffix or ".bin") as handle:
            handle.write(data)
            return handle.name

    async def _send_file(self, chat_id: str, path: str, caption: str, force_file_attachment: bool = False) -> str:
        assert self._send_session is not None and self._token is not None
        plaintext = Path(path).read_bytes()
        media_type, item_builder = self._outbound_media_builder(path, force_file_attachment=force_file_attachment)
        filekey, aes_key = secrets.token_hex(16), secrets.token_bytes(16)
        rawsize, rawfilemd5 = len(plaintext), hashlib.md5(plaintext).hexdigest()
        upload_response = await _get_upload_url(
            self._send_session, base_url=self._base_url, token=self._token, to_user_id=chat_id, media_type=media_type, filekey=filekey,
            rawsize=rawsize, rawfilemd5=rawfilemd5, filesize=((rawsize + 16) // 16) * 16, aeskey_hex=aes_key.hex())
        upload_param = str(upload_response.get("upload_param") or "")
        ciphertext = _aes128_ecb_encrypt(plaintext, aes_key)
        # Prefer upload_full_url (direct CDN), else construct from upload_param. Both use POST — PUT 404s on the CDN.
        upload_url = str(upload_response.get("upload_full_url") or "") or (upload_param and (
            f"{self._cdn_base_url.rstrip('/')}/upload?encrypted_query_param={quote(upload_param, safe='')}&filekey={quote(filekey, safe='')}"))
        if not upload_url:
            raise RuntimeError(f"getUploadUrl returned neither upload_param nor upload_full_url: {upload_response}")
        encrypted_query_param = await _upload_ciphertext(self._send_session, ciphertext=ciphertext, upload_url=upload_url)
        context_token = self._token_store.get(self._account_id, chat_id)
        # iLink expects aes_key as base64(hex_string), not base64(raw_bytes) — otherwise images render as grey boxes.
        item_kwargs = {
            "encrypt_query_param": encrypted_query_param, "aes_key_for_api": base64.b64encode(aes_key.hex().encode("ascii")).decode("ascii"),
            "ciphertext_size": len(ciphertext), "plaintext_size": rawsize, "filename": Path(path).name, "rawfilemd5": rawfilemd5}
        if media_type == MEDIA_VOICE and path.endswith(".silk"):
            item_kwargs.update(encode_type=6, sample_rate=24000, bits_per_sample=16)
        item_lists: List[List[Dict[str, Any]]] = [[item_builder(**item_kwargs)]]
        if caption:
            item_lists.insert(0, [{"type": ITEM_TEXT, "text_item": {"text": self.format_message(caption)}}])
        last_message_id = ""
        for item_list in item_lists:
            last_message_id = f"hermes-weixin-{uuid.uuid4().hex}"
            while True:
                resp = await _send_items(
                    self._send_session, base_url=self._base_url, token=self._token, to=chat_id, item_list=item_list,
                    context_token=context_token, client_id=last_message_id)
                ret, errcode = (resp.get("ret"), resp.get("errcode")) if resp and isinstance(resp, dict) else (None, None)
                if (ret is None or ret == 0) and (errcode is None or errcode == 0):
                    break
                # Same stale-session fallback as _send_text_chunk: re-send once without context_token. Clearing the
                # token also covers the remaining item lists (caption, then media) and bounds this loop.
                if _is_session_expired(resp, ret, errcode) and context_token:
                    context_token = None
                    self._token_store._cache.pop(self._token_store._key(self._account_id, chat_id), None)
                    logger.warning("[%s] session expired for %s; re-sending media without context_token", self.name, _safe_id(chat_id))
                    continue
                errmsg = resp.get("errmsg") or resp.get("msg")
                if _is_stale_session_ret(ret, errcode, errmsg):
                    raise _session_not_ready_error(ret, errcode, errmsg)
                raise RuntimeError(f"iLink sendmessage error: ret={ret} errcode={errcode} errmsg={errmsg or 'unknown error'}")
        return last_message_id

    def _outbound_media_builder(self, path: str, force_file_attachment: bool = False):
        mime = mimetypes.guess_type(path)[0] or "application/octet-stream"
        if mime.startswith("image/"):
            return MEDIA_IMAGE, _image_item
        if mime.startswith("video/"):
            return MEDIA_VIDEO, _video_item
        if path.endswith(".silk") and not force_file_attachment:
            return MEDIA_VOICE, _voice_item
        return MEDIA_FILE, _file_item  # audio/* and everything else ship as file attachments

    async def get_chat_info(self, chat_id: str) -> Dict[str, Any]:
        return {"name": chat_id, "type": "group" if chat_id.endswith("@chatroom") else "dm", "chat_id": chat_id}

    def format_message(self, content: Optional[str]) -> str:
        return "" if content is None else _wrap_copy_friendly_lines_for_weixin(_normalize_markdown_blocks(content))


async def _deliver_direct(
    adapter: WeixinAdapter, chat_id: str, message: str, media_files: Optional[List[Tuple[str, bool]]], context_token: Optional[str],
) -> Dict[str, Any]:
    last_result: Optional[SendResult] = None
    cleaned = adapter.format_message(message)
    if cleaned:
        last_result = await adapter.send(chat_id, cleaned)
        if not last_result.success:
            return {"error": f"Weixin send failed: {last_result.error}"}
    for media_path, _is_voice in media_files or []:
        sender = adapter.send_image_file if Path(media_path).suffix.lower() in _DIRECT_IMAGE_EXTS else adapter.send_document
        last_result = await sender(chat_id, media_path)
        if not last_result.success:
            return {"error": f"Weixin media send failed: {last_result.error}"}
    message_id = last_result.message_id if last_result else None
    return {"success": True, "platform": "weixin", "chat_id": chat_id, "message_id": message_id, "context_token_used": bool(context_token)}


async def send_weixin_direct(
    *, extra: Dict[str, Any], token: Optional[str], chat_id: str, message: str, media_files: Optional[List[Tuple[str, bool]]] = None,
) -> Dict[str, Any]:
    """One-shot send for ``send_message``/cron: reuse the live adapter's session on this loop, else a throwaway adapter."""
    account_id = _extra_or_secret(extra, "account_id")
    base_url = _extra_or_secret(extra, "base_url", ILINK_BASE_URL).rstrip("/")
    cdn_base_url = _extra_or_secret(extra, "cdn_base_url", WEIXIN_CDN_BASE_URL).rstrip("/")
    resolved_token = str(token or extra.get("token") or _wx_secret("WEIXIN_TOKEN", "")).strip()
    if not resolved_token:
        return {"error": "Weixin token missing. Configure WEIXIN_TOKEN or platforms.weixin.token."}
    if not account_id:
        return {"error": "Weixin account ID missing. Configure WEIXIN_ACCOUNT_ID or platforms.weixin.extra.account_id."}
    token_store = ContextTokenStore(str(get_hermes_home()))
    token_store.restore(account_id)
    context_token = token_store.get(account_id, chat_id)
    live_adapter = _LIVE_ADAPTERS.get(resolved_token)
    send_session = getattr(live_adapter, '_send_session', None)
    if send_session is not None and not send_session.closed and send_session._loop is asyncio.get_running_loop():
        return await _deliver_direct(live_adapter, chat_id, message, media_files, context_token)
    async with _new_session() as session:
        merged = {**dict(extra or {}), "account_id": account_id, "base_url": base_url, "cdn_base_url": cdn_base_url}
        adapter = WeixinAdapter(PlatformConfig(enabled=True, token=resolved_token, extra=merged))
        adapter._send_session = adapter._session = session
        adapter._token_store = token_store
        return await _deliver_direct(adapter, chat_id, message, media_files, context_token)