site/WESP_REL/app/services/admin_activity_feed.py

"""Объединённая лента админки: JSONL (действия UI) + хвост wesp.log (WARNING+)."""

from __future__ import annotations

import json
import re
import time
from datetime import datetime
from pathlib import Path
from typing import Any, Dict, List, Optional

from flask import Flask

from app.services.admin_dashboard_service import tail_text_file
from app.services.admin_llm_activity_log import (
    format_llm_log_line_text,
    read_admin_llm_activity,
)
from app.services.admin_ui_activity_log import _activity_log_path, _line_unix_ts

_LOG_LINE_RE = re.compile(
    r"^(?:(\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2}) )?(DEBUG|INFO|WARNING|ERROR|CRITICAL):([^:]+):(.*)$"
)


def _resolve_path(app: Flask, raw: str) -> Path:
    p = Path(raw)
    if not p.is_absolute():
        p = Path(app.config.get("BASE_DIR", ".")) / p
    return p


def _tail_jsonl_objects(path: Path, max_lines: int, max_bytes: int = 384_000) -> List[Dict[str, Any]]:
    if not path.is_file():
        return []
    lines, _err = tail_text_file(path, max_lines=max_lines + 80, max_bytes=max_bytes)
    out: List[Dict[str, Any]] = []
    for ln in lines:
        raw = ln.strip()
        if not raw:
            continue
        try:
            obj = json.loads(raw)
        except json.JSONDecodeError:
            continue
        if isinstance(obj, dict) and "u" in obj and "text" in obj:
            out.append(obj)
    return out[-max_lines:]


def _parse_log_line(line: str, fallback_ts_ms: int) -> Optional[Dict[str, Any]]:
    m = _LOG_LINE_RE.match(line.strip())
    if not m:
        return None
    time_s, sev, _name, msg = m.group(1), m.group(2), m.group(3), m.group(4)
    if sev in ("DEBUG", "INFO"):
        return None
    level = "warn" if sev == "WARNING" else "err"
    ts_ms = fallback_ts_ms
    if time_s:
        try:
            dt = datetime.strptime(time_s, "%Y-%m-%d %H:%M:%S")
            ts_ms = int(dt.timestamp() * 1000)
        except ValueError:
            pass
    text = f"{sev}:{m.group(3)}:{msg}".strip()
    return {
        "ts": ts_ms,
        "level": level,
        "source": "log",
        "text": text[:2000],
        "user": "wesp.log",
    }


def _dedupe_sorted(entries: List[Dict[str, Any]]) -> List[Dict[str, Any]]:
    """Один ключ (level + начало текста) — оставляем запись с большим ts (новее)."""
    by_key: Dict[tuple[str, str], Dict[str, Any]] = {}
    for e in entries:
        lvl = str(e.get("level") or "ok")
        txt = str(e.get("text") or "").strip()[:240]
        key = (lvl, txt)
        prev = by_key.get(key)
        ts = int(e.get("ts") or 0)
        if prev is None or ts >= int(prev.get("ts") or 0):
            by_key[key] = e
    out = list(by_key.values())
    out.sort(key=lambda x: int(x.get("ts") or 0))
    return out


def build_activity_feed(app: Flask) -> tuple[List[Dict[str, Any]], str]:
    """
    События за окно retention UI (суток), без дублирования level+text.
    ts — миллисекунды (Unix*1000 или синтетика для строк без asctime).
    """
    days = int(app.config.get("WESP_ADMIN_UI_ACTIVITY_RETENTION_DAYS", 3) or 3)
    days = max(1, min(days, 30))
    cutoff_ms = int((time.time() - float(days * 86400)) * 1000)

    items: List[Dict[str, Any]] = []

    jsonl_raw = _activity_log_path(app)
    if jsonl_raw and not app.config.get("TESTING"):
        jpath = _resolve_path(app, jsonl_raw)
        for obj in _tail_jsonl_objects(jpath, max_lines=400):
            u = _line_unix_ts(obj)
            if u is None:
                continue
            ts_ms = int(u * 1000)
            if ts_ms < cutoff_ms:
                continue
            lvl = str(obj.get("level") or "ok")
            if lvl not in ("ok", "err", "warn"):
                lvl = "ok"
            items.append(
                {
                    "ts": ts_ms,
                    "level": lvl,
                    "source": "ui",
                    "text": str(obj.get("text") or "")[:2000],
                    "user": str(obj.get("user") or "")[:128] or None,
                }
            )

    llm_days = int(app.config.get("WESP_ADMIN_LLM_LOG_RETENTION_DAYS", 7) or 7)
    llm_days = max(1, min(llm_days, 30))
    llm_cutoff_ms = int((time.time() - float(llm_days * 86400)) * 1000)
    llm_rows, _llm_path = read_admin_llm_activity(app, max_lines=400)
    for obj in llm_rows:
        u = obj.get("u")
        if not isinstance(u, (int, float)):
            continue
        ts_ms = int(float(u) * 1000)
        if ts_ms < llm_cutoff_ms:
            continue
        lvl = str(obj.get("level") or "ok")
        if lvl not in ("ok", "err", "warn"):
            lvl = "ok"
        items.append(
            {
                "ts": ts_ms,
                "level": lvl,
                "source": "llm",
                "text": format_llm_log_line_text(obj),
                "user": str(obj.get("user") or "")[:128] or None,
            }
        )

    log_raw = (app.config.get("WESP_ADMIN_LOG_PATH") or "").strip()
    if log_raw and not app.config.get("TESTING"):
        lpath = _resolve_path(app, log_raw)
        log_lines, _err = tail_text_file(lpath, max_lines=500, max_bytes=512_000)
        base_ms = int(time.time() * 1000)
        n = len(log_lines)
        for i, line in enumerate(log_lines):
            fb = base_ms - (n - 1 - i) * 15
            parsed = _parse_log_line(line, fb)
            if not parsed:
                continue
            if int(parsed["ts"]) < cutoff_ms:
                continue
            items.append(parsed)

    merged = _dedupe_sorted(items)
    llm_log_raw = (app.config.get("WESP_ADMIN_LLM_LOG_PATH") or "").strip()
    meta = (
        f"jsonl={bool(jsonl_raw)} llm={bool(llm_log_raw)} log={bool(log_raw)} "
        f"days={days} llm_days={llm_days}"
    )
    return merged, meta