"""Rate-limited WhatsApp interim progress during long channel turns.""" from __future__ import annotations import os import re import threading import time from typing import Any, Callable _TOOLS_DONE_RE = re.compile(r"tools done\s*\((\d+)\s*ms\)", re.IGNORECASE) # Tools that commonly exceed ~10s on production WA turns. _LONG_TOOL_LABELS_ZH: dict[str, str] = { "execmanagedne": "正在设备上执行命令，请稍候…", "listclitargets": "正在枚举 CLI 目标，请稍候…", "umealarmxlsxreport": "正在生成告警 Excel，请稍候…", "queryumealarms": "正在查询 UME 告警，请稍候…", "queryumealarmsraw": "正在查询 UME 告警，请稍候…", "aggregateumealarms": "正在汇总 UME 告警，请稍候…", "aggregateumealarmsraw": "正在汇总 UME 告警，请稍候…", "runumediagnostics": "正在跑 UME 诊断，请稍候…", "sqlqueryume": "正在执行 UME SQL，请稍候…", "findtopologypaths": "正在查拓扑路径，请稍候…", "writexlsx": "正在写 Excel，请稍候…", "getmanagedne": "正在读取纳管网元信息，请稍候…", "getumene": "正在读取 UME 网元详情，请稍候…", } _LONG_TOOL_LABELS_EN: dict[str, str] = { "execmanagedne": "Running device CLI, please wait…", "listclitargets": "Listing CLI targets, please wait…", "umealarmxlsxreport": "Building alarm Excel, please wait…", "queryumealarms": "Querying UME alarms, please wait…", "queryumealarmsraw": "Querying UME alarms, please wait…", "aggregateumealarms": "Aggregating UME alarms, please wait…", "aggregateumealarmsraw": "Aggregating UME alarms, please wait…", "runumediagnostics": "Running UME diagnostics, please wait…", "sqlqueryume": "Running UME SQL, please wait…", "findtopologypaths": "Finding topology paths, please wait…", "writexlsx": "Writing Excel, please wait…", "getmanagedne": "Loading managed NE, please wait…", "getumene": "Loading UME NE detail, please wait…", } def whatsapp_turn_progress_enabled() -> bool: raw = str(os.environ.get("OCLAW_WHATSAPP_TURN_PROGRESS") or "").strip().lower() if not raw: return True return raw not in {"0", "false", "no", "off"} def progress_min_interval_sec() -> float: raw = str(os.environ.get("OCLAW_WHATSAPP_PROGRESS_MIN_INTERVAL_SEC") or "").strip() try: return max(3.0, min(float(raw), 120.0)) except Exception: return 12.0 def normalize_tool_key(name: str) -> str: s = str(name or "").strip() if not s: return "" low = s.lower() if low.startswith("mcp__"): # mcp__netx__execManagedNe -> execmanagedne parts = low.split("__") if len(parts) >= 3: low = parts[-1] elif low.startswith("netx_"): low = low[len("netx_") :] return low.replace("_", "") def humanize_long_tool(*, tool_name: str, lang: str = "en") -> str | None: key = normalize_tool_key(tool_name) if not key: return None # Field WhatsApp is English-first; only use Chinese when lang is explicitly zh. table = _LONG_TOOL_LABELS_ZH if str(lang or "").strip().lower().startswith("zh") else _LONG_TOOL_LABELS_EN return table.get(key) def should_forward_progress_text(text: str) -> bool: """Filter noisy think/retry ticks; keep meaningful wait signals.""" t = str(text or "").strip() if not t: return False low = t.lower() if low in {"oclaw: running…", "oclaw: running...", "oclaw: finalize…", "oclaw: finalize..."}: return False if "think (" in low or low.startswith("oclaw: think"): return False if "retry-empty" in low or "retry-native-tool-calls" in low: return False m = _TOOLS_DONE_RE.search(t) if m: try: return int(m.group(1)) >= 8000 except Exception: return False # Specialist / other explicit progress lines if low.startswith("oclaw:"): return True return len(t) >= 8 def humanize_progress_text(*, text: str, lang: str = "en") -> str: t = str(text or "").strip() is_zh = str(lang or "").strip().lower().startswith("zh") m = _TOOLS_DONE_RE.search(t) if m: if is_zh: return "工具已完成，正在整理回复…" return "Tools finished; composing the reply…" if t.lower().startswith("oclaw:"): body = t.split(":", 1)[-1].strip() if is_zh: return f"处理中：{body}" if body else "处理中，请稍候…" return body or "Still working, please wait…" return t def build_whatsapp_group_progress_metadata(*, inbound: Any) -> dict[str, Any]: """@sender without quoting (avoids sticky quote spam on interim ticks).""" from runtime.orchestration.group_ingest import build_whatsapp_group_reply_metadata meta = build_whatsapp_group_reply_metadata(inbound=inbound) for key in ( "quote_remote_jid", "quote_stanza_id", "quote_participant", "quote_text", "quote_push_name", ): meta.pop(key, None) return meta class WhatsappTurnProgressPublisher: """Enqueue throttled interim WhatsApp texts during a channel turn.""" def __init__( self, *, enqueue: Callable[[str, dict[str, Any] | None], None], lang: str = "en", is_group: bool = False, inbound: Any = None, min_interval_sec: float | None = None, enabled: bool | None = None, clock: Callable[[], float] | None = None, ) -> None: self._enqueue = enqueue self._lang = str(lang or "zh") self._is_group = bool(is_group) self._inbound = inbound self._min_interval = float(min_interval_sec if min_interval_sec is not None else progress_min_interval_sec()) self._enabled = bool(whatsapp_turn_progress_enabled() if enabled is None else enabled) self._clock = clock or time.monotonic self._lock = threading.Lock() self._last_sent_at = 0.0 self._last_text = "" self._sent_count = 0 @property def sent_count(self) -> int: return int(self._sent_count) def on_progress(self, text: str) -> None: if not self._enabled: return if not should_forward_progress_text(text): return msg = humanize_progress_text(text=text, lang=self._lang) self._maybe_send(msg) def on_tool_ui(self, event: str, payload: dict[str, Any] | None) -> None: if not self._enabled: return if str(event or "").strip() != "tool_use_call": return pl = payload if isinstance(payload, dict) else {} label = humanize_long_tool(tool_name=str(pl.get("tool_name") or ""), lang=self._lang) if not label: return self._maybe_send(label) def _reply_metadata(self) -> dict[str, Any] | None: if not self._is_group or self._inbound is None: return None try: return build_whatsapp_group_progress_metadata(inbound=self._inbound) except Exception: return None def _maybe_send(self, text: str) -> None: msg = str(text or "").strip() if not msg: return with self._lock: now = float(self._clock()) if msg == self._last_text and self._sent_count > 0: return if self._sent_count > 0 and (now - self._last_sent_at) < self._min_interval: return try: self._enqueue(msg, self._reply_metadata()) except Exception: return self._last_sent_at = now self._last_text = msg self._sent_count += 1 __all__ = [ "WhatsappTurnProgressPublisher", "build_whatsapp_group_progress_metadata", "humanize_long_tool", "humanize_progress_text", "normalize_tool_key", "progress_min_interval_sec", "should_forward_progress_text", "whatsapp_turn_progress_enabled", ]