services/config.py

from __future__ import annotations

from dataclasses import dataclass
import json
import os
import sys
from pathlib import Path
import time

from services.storage.base import StorageBackend

BASE_DIR = Path(__file__).resolve().parents[1]


def _resolve_data_dir() -> Path:
    """选择数据目录：env DATA_DIR > 可写的 /data(持久盘) > 本地 ./data(临时)。

    HF Space 开启 Persistent Storage 后会把持久盘挂在 /data，重启不丢；
    未挂载时退回仓库内 ./data（临时，重启清空）。
    """
    env_dir = os.getenv("DATA_DIR", "").strip()
    if env_dir:
        return Path(env_dir)
    persistent = Path("/data")
    try:
        if persistent.is_dir() and os.access(persistent, os.W_OK):
            return persistent
    except OSError:
        pass
    return BASE_DIR / "data"


DATA_DIR = _resolve_data_dir()
# 始终容器本地盘（临时，重启清空）。放无需持久化的高频写文件（如调用日志），
# 避免每次写都打到挂载的对象桶 /data 上拖慢请求。
LOCAL_DATA_DIR = BASE_DIR / "data"
# 仓库内默认配置（打进镜像），作首启种子。
SEED_CONFIG_FILE = BASE_DIR / "config.json"
# 持久化配置：写到 DATA_DIR（挂了 /data 桶就持久）。UI 改的设置重启不丢。
CONFIG_FILE = DATA_DIR / "config.json"
VERSION_FILE = BASE_DIR / "VERSION"
BACKUP_STATE_FILE = DATA_DIR / "backup_state.json"

DEFAULT_BACKUP_INCLUDE = {
    "config": True,
    "register": True,
    "cpa": True,
    "sub2api": True,
    "logs": True,
    "image_tasks": True,
    "accounts_snapshot": True,
    "auth_keys_snapshot": True,
    "images": False,
}


def _normalize_bool(value: object, default: bool = False) -> bool:
    if isinstance(value, str):
        lowered = value.strip().lower()
        if lowered in {"1", "true", "yes", "on"}:
            return True
        if lowered in {"0", "false", "no", "off"}:
            return False
    if value is None:
        return default
    return bool(value)


def _normalize_positive_int(value: object, default: int, minimum: int = 0) -> int:
    try:
        normalized = int(value)
    except (TypeError, ValueError):
        normalized = default
    return max(minimum, normalized)


def _normalize_backup_include(value: object) -> dict[str, bool]:
    source = value if isinstance(value, dict) else {}
    normalized = dict(DEFAULT_BACKUP_INCLUDE)
    for key in normalized:
        normalized[key] = _normalize_bool(source.get(key), normalized[key])
    return normalized


def _normalize_backup_settings(value: object) -> dict[str, object]:
    source = value if isinstance(value, dict) else {}
    return {
        "enabled": _normalize_bool(source.get("enabled"), False),
        "provider": "cloudflare_r2",
        "account_id": str(source.get("account_id") or "").strip(),
        "access_key_id": str(source.get("access_key_id") or "").strip(),
        "secret_access_key": str(source.get("secret_access_key") or "").strip(),
        "bucket": str(source.get("bucket") or "").strip(),
        "prefix": str(source.get("prefix") or "backups").strip().strip("/") or "backups",
        "interval_minutes": _normalize_positive_int(source.get("interval_minutes"), 360, 1),
        "rotation_keep": _normalize_positive_int(source.get("rotation_keep"), 10, 0),
        "encrypt": _normalize_bool(source.get("encrypt"), False),
        "passphrase": str(source.get("passphrase") or "").strip(),
        "include": _normalize_backup_include(source.get("include")),
    }


def _normalize_backup_state(value: object) -> dict[str, object]:
    source = value if isinstance(value, dict) else {}
    return {
        "last_started_at": str(source.get("last_started_at") or "").strip() or None,
        "last_finished_at": str(source.get("last_finished_at") or "").strip() or None,
        "last_status": str(source.get("last_status") or "idle").strip() or "idle",
        "last_error": str(source.get("last_error") or "").strip() or None,
        "last_object_key": str(source.get("last_object_key") or "").strip() or None,
    }


@dataclass(frozen=True)
class LoadedSettings:
    auth_key: str
    refresh_account_interval_minute: int


def _normalize_auth_key(value: object) -> str:
    return str(value or "").strip()


def _is_invalid_auth_key(value: object) -> bool:
    return _normalize_auth_key(value) == ""


def _read_json_object(path: Path, *, name: str) -> dict[str, object]:
    if not path.exists():
        return {}
    if path.is_dir():
        print(
            f"Warning: {name} at '{path}' is a directory, ignoring it and falling back to other configuration sources.",
            file=sys.stderr,
        )
        return {}
    try:
        data = json.loads(path.read_text(encoding="utf-8"))
    except Exception:
        return {}
    return data if isinstance(data, dict) else {}


def _load_settings() -> LoadedSettings:
    DATA_DIR.mkdir(parents=True, exist_ok=True)
    raw_config = _read_json_object(CONFIG_FILE, name="config.json")
    auth_key = _normalize_auth_key(os.getenv("CHATGPT2API_AUTH_KEY") or raw_config.get("auth-key"))
    if _is_invalid_auth_key(auth_key):
        raise ValueError(
            "❌ auth-key 未设置！\n"
            "请在环境变量 CHATGPT2API_AUTH_KEY 中设置，或者在 config.json 中填写 auth-key。"
        )

    try:
        refresh_interval = int(raw_config.get("refresh_account_interval_minute", 5))
    except (TypeError, ValueError):
        refresh_interval = 5

    return LoadedSettings(
        auth_key=auth_key,
        refresh_account_interval_minute=refresh_interval,
    )


class ConfigStore:
    def __init__(self, path: Path):
        self.path = path
        DATA_DIR.mkdir(parents=True, exist_ok=True)
        self.data = self._load()
        self._storage_backend: StorageBackend | None = None
        if _is_invalid_auth_key(self.auth_key):
            raise ValueError(
                "❌ auth-key 未设置！\n"
                "请按以下任意一种方式解决：\n"
                "1. 在 Render 的 Environment 变量中添加：\n"
                "   CHATGPT2API_AUTH_KEY = your_real_auth_key\n"
                "2. 或者在 config.json 中填写：\n"
                '   "auth-key": "your_real_auth_key"'
            )

    def _load(self) -> dict[str, object]:
        # 优先读持久化配置；不存在（首启/未挂持久盘）则用仓库内默认 config.json 作种子。
        data = _read_json_object(self.path, name="config.json")
        if data:
            return data
        return _read_json_object(SEED_CONFIG_FILE, name="config.json")

    def _save(self) -> None:
        self.path.write_text(json.dumps(self.data, ensure_ascii=False, indent=2) + "\n", encoding="utf-8")

    @property
    def auth_key(self) -> str:
        return _normalize_auth_key(os.getenv("CHATGPT2API_AUTH_KEY") or self.data.get("auth-key"))

    @property
    def accounts_file(self) -> Path:
        return DATA_DIR / "accounts.json"

    @property
    def refresh_account_interval_minute(self) -> int:
        try:
            return int(self.data.get("refresh_account_interval_minute", 5))
        except (TypeError, ValueError):
            return 5

    @property
    def image_retention_days(self) -> int:
        try:
            return max(1, int(self.data.get("image_retention_days", 30)))
        except (TypeError, ValueError):
            return 30

    @property
    def image_poll_timeout_secs(self) -> int:
        try:
            return max(1, int(self.data.get("image_poll_timeout_secs", 120)))
        except (TypeError, ValueError):
            return 120

    @property
    def image_account_concurrency(self) -> int:
        try:
            return max(1, int(self.data.get("image_account_concurrency", 3)))
        except (TypeError, ValueError):
            return 3

    @property
    def blocking_threadpool_size(self) -> int:
        # 所有阻塞型接口（生图/文本/内容过滤）都跑在 anyio 的 run_in_threadpool 线程池里，
        # 默认只有 40，是生图同时并发的硬顶（见 conversation.stream_image_outputs_with_pool）。
        # 这里允许通过 env BLOCKING_THREADPOOL_SIZE 或 config.json 抬高；I/O 密集，2C16G 抬到
        # 一两百只占很少内存，真正天花板会让位给 账号数×image_account_concurrency 与上游限流。
        raw = os.getenv("BLOCKING_THREADPOOL_SIZE") or self.data.get("blocking_threadpool_size") or 40
        try:
            value = int(raw)
        except (TypeError, ValueError):
            value = 40
        # 不低于 anyio 默认 40（避免误降并发），上限防手滑打爆内存。
        return max(40, min(2048, value))

    @property
    def auto_remove_invalid_accounts(self) -> bool:
        value = self.data.get("auto_remove_invalid_accounts", False)
        if isinstance(value, str):
            return value.strip().lower() in {"1", "true", "yes", "on"}
        return bool(value)

    @property
    def auto_remove_rate_limited_accounts(self) -> bool:
        value = self.data.get("auto_remove_rate_limited_accounts", False)
        if isinstance(value, str):
            return value.strip().lower() in {"1", "true", "yes", "on"}
        return bool(value)

    @property
    def log_levels(self) -> list[str]:
        levels = self.data.get("log_levels")
        if not isinstance(levels, list):
            return []
        allowed = {"debug", "info", "warning", "error"}
        return [level for item in levels if (level := str(item or "").strip().lower()) in allowed]

    @property
    def sensitive_words(self) -> list[str]:
        words = self.data.get("sensitive_words")
        return [word for item in words if (word := str(item or "").strip())] if isinstance(words, list) else []

    @property
    def ai_review(self) -> dict[str, object]:
        value = self.data.get("ai_review")
        return value if isinstance(value, dict) else {}

    @property
    def global_system_prompt(self) -> str:
        return str(self.data.get("global_system_prompt") or "").strip()

    @property
    def images_dir(self) -> Path:
        # 图片存本地临时盘，不放持久桶 /data：对象桶 FUSE 读取慢（TTFB 秒级），
        # 预览很卡。代价：重启丢图（图片非核心数据）。
        path = LOCAL_DATA_DIR / "images"
        path.mkdir(parents=True, exist_ok=True)
        return path

    @property
    def image_thumbnails_dir(self) -> Path:
        path = LOCAL_DATA_DIR / "image_thumbnails"
        path.mkdir(parents=True, exist_ok=True)
        return path

    def cleanup_old_images(self) -> int:
        cutoff = time.time() - self.image_retention_days * 86400
        removed = 0
        for path in self.images_dir.rglob("*"):
            if path.is_file() and path.stat().st_mtime < cutoff:
                path.unlink()
                removed += 1
        for path in sorted((p for p in self.images_dir.rglob("*") if p.is_dir()), key=lambda p: len(p.parts), reverse=True):
            try:
                path.rmdir()
            except OSError:
                pass
        return removed

    @property
    def base_url(self) -> str:
        return str(
            os.getenv("CHATGPT2API_BASE_URL")
            or self.data.get("base_url")
            or ""
        ).strip().rstrip("/")

    @property
    def app_version(self) -> str:
        try:
            value = VERSION_FILE.read_text(encoding="utf-8").strip()
        except FileNotFoundError:
            return "0.0.0"
        return value or "0.0.0"

    def get(self) -> dict[str, object]:
        data = dict(self.data)
        data["refresh_account_interval_minute"] = self.refresh_account_interval_minute
        data["image_retention_days"] = self.image_retention_days
        data["image_poll_timeout_secs"] = self.image_poll_timeout_secs
        data["image_account_concurrency"] = self.image_account_concurrency
        data["blocking_threadpool_size"] = self.blocking_threadpool_size
        data["auto_remove_invalid_accounts"] = self.auto_remove_invalid_accounts
        data["auto_remove_rate_limited_accounts"] = self.auto_remove_rate_limited_accounts
        data["log_levels"] = self.log_levels
        data["sensitive_words"] = self.sensitive_words
        data["ai_review"] = self.ai_review
        data["global_system_prompt"] = self.global_system_prompt
        data["backup"] = self.get_backup_settings()
        data.pop("auth-key", None)
        return data

    def get_proxy_settings(self) -> str:
        return str(self.data.get("proxy") or "").strip()

    def update(self, data: dict[str, object]) -> dict[str, object]:
        next_data = dict(self.data)
        next_data.update(dict(data or {}))
        if "backup" in next_data:
            next_data["backup"] = _normalize_backup_settings(next_data.get("backup"))
        next_data.pop("backup_state", None)
        self.data = next_data
        self._save()
        return self.get()

    def get_backup_settings(self) -> dict[str, object]:
        return _normalize_backup_settings(self.data.get("backup"))

    def get_storage_backend(self) -> StorageBackend:
        """获取存储后端实例（单例）"""
        if self._storage_backend is None:
            from services.storage.factory import create_storage_backend
            self._storage_backend = create_storage_backend(DATA_DIR)
        return self._storage_backend


def load_backup_state() -> dict[str, object]:
    return _normalize_backup_state(_read_json_object(BACKUP_STATE_FILE, name="backup_state.json"))


def save_backup_state(state: dict[str, object]) -> dict[str, object]:
    normalized = _normalize_backup_state(state)
    BACKUP_STATE_FILE.write_text(json.dumps(normalized, ensure_ascii=False, indent=2) + "\n", encoding="utf-8")
    return normalized


config = ConfigStore(CONFIG_FILE)