site/WESP_REL/app/services/client_log_upload.py

"""Сохранение логов, присланных полевыми клиентами на центральный сервер."""

from __future__ import annotations

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

from sqlalchemy import or_, select

_LOG_NAME_RE = re.compile(r"^[\w.\-]{1,120}$")


def safe_path_segment(raw: str, max_len: int = 96) -> str:
    s = (raw or "").strip()[:max_len]
    if not s:
        return "unknown"
    return "".join(c if (c.isalnum() or c in "-_") else "_" for c in s)


def validate_log_basename(name: str) -> str:
    n = (name or "").strip() or "wesp.log"
    if not _LOG_NAME_RE.match(n):
        raise ValueError("Недопустимое имя файла лога")
    return n


def store_client_log(
    *,
    upload_dir: str,
    client_id: str,
    log_basename: str,
    raw: bytes,
) -> Dict[str, Any]:
    """
    Пишет байты в upload_dir/<safe_client>/<timestamp>_<log_basename>.
    upload_dir должен быть абсолютным или относительным к cwd процесса.
    """
    if not upload_dir or not str(upload_dir).strip():
        raise ValueError("Каталог выгрузки логов не задан")
    cid = safe_path_segment(client_id, max_len=80)
    base = validate_log_basename(log_basename)
    root = Path(upload_dir).resolve()
    dest_dir = root / cid
    dest_dir.mkdir(parents=True, exist_ok=True)
    ts = time.strftime("%Y%m%d_%H%M%S", time.gmtime())
    dest = dest_dir / f"{ts}_{base}"
    dest.write_bytes(raw)
    try:
        rel = dest.relative_to(root)
    except ValueError:
        rel = dest
    return {
        "path": str(dest),
        "relative": str(rel).replace("\\", "/"),
        "bytes_written": len(raw),
    }


def client_logs_root_resolved(upload_dir: str, base_dir: str) -> Path:
    raw = (upload_dir or "").strip()
    if not raw:
        raise ValueError("Каталог выгрузки логов не задан")
    p = Path(raw)
    if not p.is_absolute():
        p = Path(base_dir) / p
    return p.resolve()


def _is_inside(parent: Path, child: Path) -> bool:
    try:
        child.resolve().relative_to(parent.resolve())
        return True
    except ValueError:
        return False


def client_upload_log_dir(root: Path, client_id: str) -> Path:
    seg = safe_path_segment(client_id, max_len=80)
    d = (root / seg).resolve()
    if not _is_inside(root, d) and d != root:
        raise ValueError("Некорректный client_id")
    return d


def list_uploaded_client_log_packages(root: Path) -> List[Dict[str, Any]]:
    if not root.is_dir():
        return []
    out: List[Dict[str, Any]] = []
    for sub in sorted(root.iterdir(), key=lambda x: x.name.lower()):
        if not sub.is_dir():
            continue
        files: List[Dict[str, Any]] = []
        for f in sorted(sub.iterdir(), key=lambda x: x.name):
            if not f.is_file() or not _LOG_NAME_RE.match(f.name):
                continue
            st = f.stat()
            files.append(
                {
                    "name": f.name,
                    "size": st.st_size,
                    "mtime": datetime.fromtimestamp(st.st_mtime, tz=timezone.utc).isoformat(),
                }
            )
        out.append({"dir_segment": sub.name, "file_count": len(files), "files": files})
    return out


def build_uploaded_logs_index(db_session, root: Path) -> List[Dict[str, Any]]:
    """Список каталогов выгрузки + сопоставление с SyncClient по safe_path_segment(node_id)."""
    from app.models import SyncClient, SyncClientDisplayName

    packages = list_uploaded_client_log_packages(root)
    clients = list(
        db_session.execute(
            select(SyncClient).where(
                or_(SyncClient.is_deleted.is_(False), SyncClient.is_deleted.is_(None))
            )
        ).scalars().all()
    )
    names = {
        n.node_id: n.display_name
        for n in db_session.execute(select(SyncClientDisplayName)).scalars().all()
    }
    by_seg = {safe_path_segment(c.node_id, max_len=80): c for c in clients}
    result: List[Dict[str, Any]] = []
    for p in packages:
        seg = p["dir_segment"]
        sc = by_seg.get(seg)
        nid = sc.node_id if sc else None
        result.append(
            {
                "dir_segment": seg,
                "node_id": nid,
                "client_name": (sc.client_name if sc else None),
                "display_name": (names.get(nid) if nid else None),
                "file_count": p["file_count"],
                "files": p["files"],
                "api_client_id": nid or seg,
            }
        )
    return result


def latest_uploaded_log_file(client_dir: Path) -> Optional[Path]:
    if not client_dir.is_dir():
        return None
    candidates = [
        f
        for f in client_dir.iterdir()
        if f.is_file() and _LOG_NAME_RE.match(f.name)
    ]
    if not candidates:
        return None
    return max(candidates, key=lambda p: p.stat().st_mtime)


def resolve_uploaded_client_log_file(root: Path, client_id: str, filename: str) -> Path:
    d = client_upload_log_dir(root, client_id)
    if not d.is_dir():
        raise FileNotFoundError("Каталог клиента не найден")
    fn = Path(filename).name
    if not _LOG_NAME_RE.match(fn):
        raise ValueError("Недопустимое имя файла")
    fpath = (d / fn).resolve()
    if not _is_inside(d, fpath) and fpath != d:
        raise ValueError("Некорректный путь")
    if not fpath.is_file():
        raise FileNotFoundError("Файл не найден")
    return fpath