PolyGateway/src/polygateway/config.py

"""配置聚合与装配期校验(M1 设计 §8;ARCH §9)。

键族: 多源 `{SCOPE}__{PROVIDER}__{N}__{FIELD}`、scope 韧性键、平铺 LLM_*
简写(单 scope 迁移零改名)、PGW_* 装配键。缺关键配置直接报错,严禁默认
值兜底;scope 键与平铺键并存时 scope 键优先。

实现说明: 多源键族是动态键名,pydantic-settings 的静态字段模型无法表达,
故用其底层 python-dotenv 读 `.env` 并与 os.environ 合并(环境变量优先),
fail-loud 校验语义与 pydantic-settings 一致。
"""

from __future__ import annotations

import json
import os
from dataclasses import dataclass
from typing import TYPE_CHECKING

from dotenv import dotenv_values
from loguru import logger

from polygateway.types import (
    BackpressurePolicy,
    BreakerConfig,
    GlobalLimits,
    RetryPolicy,
    SourceConfig,
    coerce_effort,
)

if TYPE_CHECKING:
    from collections.abc import Mapping

# FIELD → (SourceConfig 属性, 类型);CHS config.py:95-104 全集 + M1 新增
_SOURCE_FIELDS: dict[str, tuple[str, str]] = {
    "BASE_URL": ("base_url", "str"),
    "API_KEY": ("api_key", "str"),
    "MODEL": ("model", "str"),
    "MAX_CONCURRENCY": ("max_concurrency", "int"),
    "RPM": ("rpm", "int"),
    "TPM": ("tpm", "int"),
    "EST_TOKENS": ("est_tokens", "int"),
    "TIMEOUT_S": ("timeout_s", "float"),
    "TTFT_TIMEOUT_S": ("ttft_timeout_s", "float"),
    "INTER_TOKEN_TIMEOUT_S": ("inter_token_timeout_s", "float"),
    "ENABLE_THINKING": ("enable_thinking", "bool"),
    # 档位两键(issue #20);值域校验分工: 档位在此(解析即校验,报错点得出 env 键名),
    # fallback 交给 SourceConfig 构造期(那道同时覆盖构造函数注入与 dataclasses.replace)
    "REASONING_EFFORT": ("reasoning_effort", "effort"),
    # 归一化(strip+lower)在 SourceConfig 构造期,与值域校验同处一点,故这里是裸 "str"
    "EFFORT_FALLBACK": ("effort_fallback", "str"),
    "MISSING_DONE": ("missing_done", "str"),
    "TRUST_ENV": ("trust_env", "bool"),
    "EXTRA_BODY": ("extra_body", "json"),
}
_RESERVED_SEGMENTS = frozenset({"GLOBAL", "RETRY", "BREAKER", "BACKPRESSURE"})
_SELECTORS = frozenset({"round_robin", "least_inflight", "health_aware"})
_QUOTA_FULL = frozenset({"wait", "fail_fast"})
# 熔断全拒时的处置(issue #14);值域与 _QUOTA_FULL 相同但语义不同——配额满是
# "排队等自己的份额"(必然轮到),熔断开路是"等源恢复"(未必恢复),故分列两键
_CIRCUIT_OPEN = frozenset({"wait", "fail_fast"})
# 后端合法域: env 解析与构造期校验共用一份定义,避免两处分叉
_LIMITER_BACKENDS = frozenset({"memory", "redis"})
_BREAKER_BACKENDS = frozenset({"memory", "redis"})
_CACHE_BACKENDS = frozenset({"redis", "memory", "none"})
_TELEMETRY_BACKENDS = frozenset({"sqlite", "postgres", "none"})
# 遥测 schema 档位(issue #13): auto 允许 recorder 给旧表 ALTER 补列,manual 不发 DDL
_SCHEMA_MODES = frozenset({"auto", "manual"})
_SCHEMA_MODE_KEY = "PGW_TELEMETRY_SCHEMA_MODE"
# 遥测正文字符上限(issue #12);二态键,未设 = 不截断
_TEXT_CAP_KEY = "PGW_TELEMETRY_TEXT_CAP"
# 遥测池的资源占用与写入预算(issue #15);缺省只写在这里,recorder 侧是必填参数
_POOL_MAX_KEY = "PGW_TELEMETRY_PG_POOL_MAX"
_WRITE_TIMEOUT_KEY = "PGW_TELEMETRY_PG_WRITE_TIMEOUT_S"
# 4 条实测约 15.6 行/秒(跨内网 RTT ≈ 123ms 的实验室 PG,50 行并发批耗时 3.2s)。
# **不要按 `pool_max / RTT` 折算**——那会乐观一倍(一次 INSERT 的往返比一次
# SELECT 1 重)。够单 client 十余并发;闲时占 0 条
_DEFAULT_PG_POOL_MAX = 4
# 实测稳态写入 123ms、首次含建连 513ms;5s 宽松且**有界**
_DEFAULT_PG_WRITE_TIMEOUT_S = 5.0
_REDIS_DEPENDENT_BACKENDS = ("limiter_backend", "breaker_backend", "cache_backend")
# 背压默认(M1 仅 poll 生效;CHS _BACKOFF_S=0.05 同源)
_DEFAULT_STALL_WINDOW_S = 300.0
_DEFAULT_POLL_INTERVAL_S = 0.05
_DEFAULT_LEASE_TTL_S = 1500.0  # CHS _DEFAULT_LEASE_TTL_MS 同源
_PROBE_GRACE_S = 5.0  # 半开探针租约相对最慢调用的清理宽限(CHS container.py:274-275)


def _cast(raw: str, kind: str, key: str) -> object:
    try:
        if kind == "int":
            return int(raw)
        if kind == "float":
            return float(raw)
        if kind == "bool":
            lowered = raw.strip().lower()
            if lowered in ("1", "true", "yes", "on"):
                return True
            if lowered in ("0", "false", "no", "off"):
                return False
            raise ValueError(f"非法布尔值: {raw!r}")
        if kind == "effort":
            # 归一化只有一份实现(`types.coerce_effort`),env 路与两条装配路同口径;
            # origin 传空串是因为 env 键名由下面统一的"配置 X 解析失败"补上
            return coerce_effort(raw, origin="")
        if kind == "json":
            # JSONDecodeError 是 ValueError 子类,复用下方的统一包装
            parsed = json.loads(raw)
            if not isinstance(parsed, dict):
                raise ValueError(f"必须是 JSON 对象(而非数组/标量): {raw!r}")
            return parsed
        return raw
    except ValueError as exc:
        raise ValueError(f"配置 {key} 解析失败: {exc}") from exc


def _first(env: Mapping[str, str], *keys: str) -> tuple[str, str] | None:
    for key in keys:
        raw = env.get(key)
        if raw is not None and raw != "":
            return key, raw
    return None


def _require(env: Mapping[str, str], *keys: str) -> tuple[str, str]:
    found = _first(env, *keys)
    if found is None:
        raise ValueError(f"缺关键配置: 需设置 {' 或 '.join(keys)}")
    return found


@dataclass(frozen=True)
class GatewaySettings:
    """一个 scope 的完整装配配置;**任何**构造路径都通过全部装配守卫(ARCH §7.3)。

    守卫校验的是**跨字段**不变量: 单看一个字段都合法,组合起来才会在运行时
    咬人(租约先于请求过期、正常慢首包被误判卡死、半开探针在途被接管)。
    types.py 各子配置的 `__post_init__` 只看得见自己的字段,故由本类把关。

    放在 `__post_init__` 而非某个工厂里: 这些约束是本类定义的一部分,不是
    某个入口的输入检查。挂在构造期,直接构造、`dataclasses.replace` 与全部
    装配工厂一并覆盖;挂在工厂里则每加一个工厂就多一处要同步。
    """

    scope: str
    sources: tuple[SourceConfig, ...]
    global_limits: GlobalLimits
    retry: RetryPolicy
    breaker: BreakerConfig
    backpressure: BackpressurePolicy
    selector: str
    quota_full: str
    # 熔断全拒时是当场判死还是等冷却过去(issue #14);缺省 fail_fast 保持
    # 存量下游的控制流不变,单源 scope 应显式配 wait
    circuit_open: str
    limiter_backend: str
    breaker_backend: str
    cache_backend: str
    cache_namespace: str | None
    cache_ttl_s: int | None
    telemetry_backend: str
    telemetry_sqlite_path: str | None
    telemetry_pg_dsn: str | None
    # 是否允许 recorder 给已存在的旧表自动 ALTER 补列(issue #13);env 的三态
    # 派生只写在 `_load_schema_mode` 一处,不与 recorder 的类签名漂移。
    # backend=none 时恒 False 这条跨字段不变量则由 `_validate_telemetry`
    # 把关,对直接构造与 `dataclasses.replace` 同样生效
    telemetry_auto_migrate: bool
    # 遥测落库正文的字符上限(issue #12);None = 不截断,与本字段出现之前逐字节相同。
    # 缺省不截断是人类决策: 截断后的遥测不再是审计证据、也无法用于复现与重放,而
    # 既有下游正依赖这一行为。值域(> 0)由 `_validate_telemetry` 把关,直接构造、
    # `dataclasses.replace` 与 env 三条路一并覆盖
    telemetry_text_cap: int | None
    # 遥测池对外声明的资源占用上限与整次写入的硬预算(issue #15)。库内每一处外部
    # 资源都按需建连,唯独遥测池此前预占 10 条(asyncpg 默认 `min_size`),共享实例
    # 余量紧张时先倒下的必然是它。这两个字段是库对自己占用的**显式表态**:
    # 稳态并发上限 = `pool_max`,闲时 0 条;单次写入(准备+取连接+执行)≤ 预算。
    # 值域由 `_validate_telemetry` 把关,直接构造、`dataclasses.replace` 与 env 三条路一致
    telemetry_pg_pool_max: int
    telemetry_pg_write_timeout_s: float
    redis_url: str | None
    pricing_path: str | None
    structured_max_retries: int
    lease_ttl_s: float

    def __post_init__(self) -> None:
        self._normalize()
        self._validate_identity()
        self._validate_backends()
        self._validate_cache()
        self._validate_telemetry()
        self._validate_lease()
        self._validate_stall()
        self._validate_probe()

    def _normalize(self) -> None:
        """把 `from_env` 一直在做的规范化补到构造路上,两条路必须产出同一个值。

        `scope` 的 strip 才是要紧的那一半: 它进 Redis key(`pgw:limit:{scope}:…`
        /`pgw:gate:{scope}:…`),而两个 Redis 后端在构造函数里只 `.lower()` **不 strip**
        ——`"llm "` 会产出 `pgw:limit:llm :…`,与 `from_env` 路的进程分裂成两套命名空间。
        大小写则不会: 后端自 v1.0.0 起各自 lower,`from_settings` 传 "LLM" 也落在同一
        套 key 上(此处 lower 只为让 `GatewaySettings.scope` 属性两路取值一致)。

        空串归 None 同理: 留着空串会骗过 `is None` 判断,把错误推迟到 redis 客户端
        抛连接串解析异常。`telemetry_pg_dsn` 的驱动后缀因为要看 backend 且需告警,
        规范化留在 `_validate_telemetry`。
        """
        normalized_scope = self.scope.strip().lower()
        if normalized_scope != self.scope:
            object.__setattr__(self, "scope", normalized_scope)
        for field in ("redis_url", "pricing_path"):
            if getattr(self, field) == "":
                object.__setattr__(self, field, None)

    def _validate_identity(self) -> None:
        """本类自身字段的基本域: 空 scope 会污染遥测与缓存命名空间;零源必然选源失败。"""
        if not self.scope.strip():
            raise ValueError("GatewaySettings.scope 不能为空")
        if not self.sources:
            raise ValueError("GatewaySettings.sources 不能为空: 至少一个源")
        if self.structured_max_retries < 0:
            raise ValueError(f"structured_max_retries 不能为负: {self.structured_max_retries}")

    def _validate_backends(self) -> None:
        """后端选择必须落在合法域内,取 redis 的还必须有连接串。

        域外取值此前只有 `from_env` 拦得住,直接构造会一路走到 `client.py` 的
        `_build_*`,落进 else 分支静默不建后端,或撞上那里的断言。
        """
        for field, allowed in (
            ("limiter_backend", _LIMITER_BACKENDS),
            ("breaker_backend", _BREAKER_BACKENDS),
            ("cache_backend", _CACHE_BACKENDS),
            ("telemetry_backend", _TELEMETRY_BACKENDS),
            ("selector", _SELECTORS),
            ("quota_full", _QUOTA_FULL),
            ("circuit_open", _CIRCUIT_OPEN),
        ):
            value = getattr(self, field)
            if value not in allowed:
                raise ValueError(f"{field} 非法值 {value!r};允许: {sorted(allowed)}")
        on_redis = [f for f in _REDIS_DEPENDENT_BACKENDS if getattr(self, f) == "redis"]
        if on_redis and self.redis_url is None:
            raise ValueError(f"{'、'.join(on_redis)} 取 redis 时必须提供 redis_url")

    def _validate_cache(self) -> None:
        """启用缓存必须有命名空间与正 TTL(缺命名空间即失去租户隔离,会毒化缓存)。"""
        if self.cache_backend == "none":
            return
        if not self.cache_namespace:
            raise ValueError("启用缓存时 cache_namespace 不能为空: 缓存 key 靠它做租户隔离")
        if self.cache_ttl_s is None or self.cache_ttl_s <= 0:
            raise ValueError(f"cache_ttl_s 必须 > 0(禁止永不过期): {self.cache_ttl_s}")

    def _validate_telemetry(self) -> None:
        """遥测后端各自的落点必填;顺带剥掉 asyncpg 不认的 SQLAlchemy 驱动后缀。

        剥而不是拒: 两条装配路对同一 DSN 应产出同一结果。但不静默——`from_env`
        那条路在 `_load_pg_dsn` 就剥干净了,能走到这里的只有手工构造的调用方,
        他有权知道库动了他给的值。

        `telemetry_auto_migrate` 同理归一化而非报错: backend=none 时根本没有
        recorder 消费它,True 是个自相矛盾却无害的状态。`from_env` 那条路的派生
        已经给出 False,归一化是为了直接构造与 `dataclasses.replace` 也一致——
        不变量挂在构造期,才不用每加一个装配工厂就多一处要同步。

        `telemetry_text_cap` 的值域则是**报错**而非归一化: 0 与负数都不是"不截断"
        的写法(不截断写 None),把它们悄悄改成 None 等于用默认值掩盖调用方的错误。
        报错文本同时点出字段名与 env 键名,两条装配路的调用方各看得懂自己那套。
        """
        if self.telemetry_text_cap is not None and self.telemetry_text_cap <= 0:
            raise ValueError(
                f"telemetry_text_cap({_TEXT_CAP_KEY})必须 > 0: {self.telemetry_text_cap};"
                "不截断请不设该键(None),0 只会让每条正文退化成一个省略标记"
            )
        self._validate_telemetry_pool()
        if self.telemetry_backend == "none" and self.telemetry_auto_migrate:
            object.__setattr__(self, "telemetry_auto_migrate", False)
        if self.telemetry_backend == "sqlite" and not self.telemetry_sqlite_path:
            raise ValueError("telemetry_backend=sqlite 时必须提供 telemetry_sqlite_path")
        if self.telemetry_backend != "postgres":
            return
        if not self.telemetry_pg_dsn:
            raise ValueError("telemetry_backend=postgres 时必须提供 telemetry_pg_dsn")
        stripped = _strip_dsn_driver(self.telemetry_pg_dsn)
        if stripped != self.telemetry_pg_dsn:
            # 只报 scheme 段: DSN 带密码,整串不得进日志(P5 敏感信息只走 .env)
            logger.warning(
                "telemetry_pg_dsn 的 scheme 含 asyncpg 不认的驱动后缀,已由 {} 剥为 {}",
                self.telemetry_pg_dsn.partition("://")[0],
                stripped.partition("://")[0],
            )
            object.__setattr__(self, "telemetry_pg_dsn", stripped)

    def _validate_telemetry_pool(self) -> None:
        """遥测池两个标量的值域(issue #15);与 backend 无关,三条装配路一并覆盖。

        不按 `telemetry_backend == "postgres"` 才校验: 值域错就是错,提前拦住
        比等到有人把 backend 切成 postgres 时才炸更接近"缺失关键配置直接报错"。
        报错文本同时点字段名与 env 键名(两类调用方各看得懂自己那套)。
        """
        if self.telemetry_pg_pool_max < 1:
            raise ValueError(
                f"telemetry_pg_pool_max({_POOL_MAX_KEY})必须 >= 1: "
                f"{self.telemetry_pg_pool_max};0 条上限等于永远取不到连接,遥测会全灭"
            )
        if self.telemetry_pg_write_timeout_s <= 0:
            raise ValueError(
                f"telemetry_pg_write_timeout_s({_WRITE_TIMEOUT_KEY})必须 > 0: "
                f"{self.telemetry_pg_write_timeout_s};预算 0 会让每一行当场超预算被丢弃"
            )

    def _validate_lease(self) -> None:
        """调用超时须 ≤ permit 租约 TTL,防租约先于请求过期使并发超出配额。"""
        slowest = max(s.timeout_s for s in self.sources)
        if slowest > self.lease_ttl_s:
            raise ValueError(
                f"源最大 timeout_s({slowest})超过 permit 租约 lease_ttl_s"
                f"({self.lease_ttl_s});调大 lease_ttl_s 或调小源的 timeout_s"
            )

    def _validate_stall(self) -> None:
        """stall 窗口须 ≥ 最慢源 TTFT 上限(保守冗余,见下)。

        原理由是"防把正常慢首包误判为卡死"。issue #8 起 stall 只累计**非
        生产性等待**(429 退避、配额轮询、熔断冷却),TTFT 等待属生产性时间、
        已不计入 stall 账,该误判在机制上不再可能。校验本身无害且不会误拒
        任何合理配置,故保留——删除它需同步改动 ARCHITECTURE.md §7.3 的契约
        补强 G6,超出 issue #8 的范围(2026-08-06 人类定夺)。
        """
        ttfts = [s.ttft_timeout_s for s in self.sources if s.ttft_timeout_s is not None]
        if ttfts and self.backpressure.stall_window_s < max(ttfts):
            raise ValueError(
                f"backpressure.stall_window_s({self.backpressure.stall_window_s})须 ≥ "
                f"最大源 ttft_timeout_s({max(ttfts)});调大 stall_window_s 或调小 ttft_timeout_s"
            )

    def _validate_probe(self) -> None:
        """半开探针租约须撑过一次最慢调用,否则探针在途即被接管(M2 设计 §3)。"""
        floor = max(s.timeout_s for s in self.sources) + _PROBE_GRACE_S
        if self.breaker.probe_ttl_s < floor:
            raise ValueError(
                f"breaker.probe_ttl_s({self.breaker.probe_ttl_s})须 ≥ 最慢源 "
                f"timeout_s + {_PROBE_GRACE_S}({floor});调大 probe_ttl_s 或调小源的 timeout_s"
            )

    @classmethod
    def from_env(
        cls,
        scope: str = "LLM",
        env: Mapping[str, str] | None = None,
        *,
        env_file: str = ".env",
    ) -> GatewaySettings:
        """聚合 env(缺省 .env + os.environ,后者优先);守卫由 `__post_init__` 执行。"""
        if env is None:
            env = {
                k: v for k, v in {**dotenv_values(env_file), **os.environ}.items() if v is not None
            }
        scope_u = scope.upper()
        sources = _load_sources(scope_u, env)
        global_limits = _load_global_limits(scope_u, env)
        retry = _load_retry(scope_u, env)
        breaker = _load_breaker(scope_u, env, sources, global_limits)
        return cls(
            scope=scope_u.lower(),
            sources=tuple(sources),
            global_limits=global_limits,
            retry=retry,
            breaker=breaker,
            backpressure=_load_backpressure(scope_u, env),
            selector=_load_choice(env, f"{scope_u}__SELECTOR", _SELECTORS, "health_aware"),
            quota_full=_load_choice(env, f"{scope_u}__QUOTA_FULL", _QUOTA_FULL, "wait"),
            circuit_open=_load_choice(env, f"{scope_u}__CIRCUIT_OPEN", _CIRCUIT_OPEN, "fail_fast"),
            **_load_pgw(env),
        )


def _load_sources(scope: str, env: Mapping[str, str]) -> list[SourceConfig]:
    specs: dict[tuple[str, str], dict[str, object]] = {}
    for key, raw in env.items():
        parts = key.split("__")
        if len(parts) != 4 or parts[0] != scope or parts[1] in _RESERVED_SEGMENTS:
            continue
        _, provider, index, field = parts
        if field not in _SOURCE_FIELDS:
            raise ValueError(f"未知源配置字段 {field}(键 {key});允许: {sorted(_SOURCE_FIELDS)}")
        attr, kind = _SOURCE_FIELDS[field]
        specs.setdefault((provider.lower(), index), {})[attr] = _cast(raw, kind, key)
    if not specs:
        raise ValueError(f"scope {scope} 未配置任何源({scope}__{{PROVIDER}}__{{N}}__{{FIELD}})")
    return [
        _build_source(scope, provider, index, spec, env)
        for (provider, index), spec in sorted(specs.items())
    ]


def _build_source(
    scope: str, provider: str, index: str, spec: dict[str, object], env: Mapping[str, str]
) -> SourceConfig:
    prefix = f"{scope}__{provider.upper()}__{index}"
    for field in ("BASE_URL", "API_KEY", "MODEL"):
        attr, _ = _SOURCE_FIELDS[field]
        if attr not in spec:
            raise ValueError(f"缺关键配置: {prefix}__{field}")
    if "timeout_s" not in spec:
        key, raw = _require(env, f"{prefix}__TIMEOUT_S", "LLM_TIMEOUT")
        spec["timeout_s"] = _cast(raw, "float", key)
    # 平铺看门狗键作为缺省(成对才生效,遵守 SourceConfig 不变式)
    if "ttft_timeout_s" not in spec and "inter_token_timeout_s" not in spec:
        flat = _first(env, "LLM_TTFT_TIMEOUT")
        flat_inter = _first(env, "LLM_INTER_TOKEN_TIMEOUT")
        if flat and flat_inter:
            spec["ttft_timeout_s"] = _cast(flat[1], "float", flat[0])
            spec["inter_token_timeout_s"] = _cast(flat_inter[1], "float", flat_inter[0])
    return SourceConfig(name=f"{provider}_{index}", provider=provider, **spec)


def _load_global_limits(scope: str, env: Mapping[str, str]) -> GlobalLimits:
    def read(field: str) -> int:
        found = _first(env, f"{scope}__GLOBAL__{field}")
        return int(_cast(found[1], "int", found[0])) if found else 0

    return GlobalLimits(max_concurrency=read("MAX_CONCURRENCY"), rpm=read("RPM"), tpm=read("TPM"))


def _load_retry(scope: str, env: Mapping[str, str]) -> RetryPolicy:
    key_a, attempts = _require(env, f"{scope}__RETRY__MAX_ATTEMPTS", "LLM_MAX_RETRIES")
    key_b, base = _require(env, f"{scope}__RETRY__BACKOFF_BASE_S", "LLM_RETRY_BASE_DELAY")
    key_m, max_d = _require(env, f"{scope}__RETRY__BACKOFF_MAX_S", "LLM_RETRY_MAX_DELAY")
    return RetryPolicy(
        max_attempts=int(_cast(attempts, "int", key_a)),
        backoff_base_s=float(_cast(base, "float", key_b)),
        backoff_max_s=float(_cast(max_d, "float", key_m)),
    )


def _load_breaker(
    scope: str, env: Mapping[str, str], sources: list[SourceConfig], global_limits: GlobalLimits
) -> BreakerConfig:
    key_t, thr = _require(env, f"{scope}__BREAKER__FAIL_THRESHOLD", "LLM_CIRCUIT_BREAKER_THRESHOLD")
    key_c, cool = _require(env, f"{scope}__BREAKER__COOLDOWN_S", "LLM_CIRCUIT_BREAKER_COOLDOWN")
    threshold = int(_cast(thr, "int", key_t))
    cooldown_s = float(_cast(cool, "float", key_c))
    # 有效阈值 = max(配置值, 源级并发×2)。M2.5 修正: 只看源级并发——M2 曾用
    # 全局并发抬升(SOAK 100→阈值 200)使熔断失灵(病灶 2,设计 2026-07-21-m25)
    concurrency = max((s.max_concurrency for s in sources), default=0)
    if concurrency > 0:
        threshold = max(threshold, concurrency * 2)
    slowest = max(s.timeout_s for s in sources)
    probe_floor = slowest + _PROBE_GRACE_S
    probe = _first(env, f"{scope}__BREAKER__PROBE_TTL_S")
    if probe is not None:
        # 配置值不在此校验: 探针租约下限是跨字段不变量,由 GatewaySettings._validate_probe
        # 统一把关(否则直接构造那条装配路会绕过)
        probe_ttl_s = float(_cast(probe[1], "float", probe[0]))
    else:
        # 派生规则: 探针租约须撑过一次最慢调用,且不短于冷却期(第三项保证守卫恒成立)
        probe_ttl_s = max(2 * slowest, cooldown_s, probe_floor)
    # M2.5 失败率通道参数(可选键,库缺省——韧性参数缺省先例同 backpressure)
    raw_min = env.get(f"{scope}__BREAKER__MIN_CALLS")
    min_calls = int(_cast(raw_min, "int", f"{scope}__BREAKER__MIN_CALLS")) if raw_min else 10
    fail_rate = _opt_float(env, f"{scope}__BREAKER__FAIL_RATE", 0.6)
    window_s = _opt_float(env, f"{scope}__BREAKER__WINDOW_S", 60.0)
    max_cooldown_s = _opt_float(env, f"{scope}__BREAKER__MAX_COOLDOWN_S", max(300.0, cooldown_s))
    return BreakerConfig(
        fail_threshold=threshold,
        cooldown_s=cooldown_s,
        probe_ttl_s=probe_ttl_s,
        min_calls=min_calls,
        fail_rate=fail_rate,
        window_s=window_s,
        max_cooldown_s=max_cooldown_s,
    )


def _opt_float(env: Mapping[str, str], key: str, default: float) -> float:
    """可选韧性参数: 缺省用库值,显式配置则解析(坏值 fail-loud)。"""
    raw = env.get(key)
    if raw is None or raw == "":
        return default
    return float(_cast(raw, "float", key))


def _load_backpressure(scope: str, env: Mapping[str, str]) -> BackpressurePolicy:
    stall = _first(env, f"{scope}__BACKPRESSURE__STALL_WINDOW_S")
    poll = _first(env, f"{scope}__BACKPRESSURE__POLL_INTERVAL_S")
    return BackpressurePolicy(
        stall_window_s=float(_cast(stall[1], "float", stall[0]))
        if stall
        else _DEFAULT_STALL_WINDOW_S,
        poll_interval_s=float(_cast(poll[1], "float", poll[0]))
        if poll
        else _DEFAULT_POLL_INTERVAL_S,
    )


def _load_choice(env: Mapping[str, str], key: str, allowed: frozenset[str], default: str) -> str:
    found = _first(env, key)
    value = found[1] if found else default
    if value not in allowed:
        raise ValueError(f"配置 {key} 非法值 {value!r};允许: {sorted(allowed)}")
    return value


def _load_pgw(env: Mapping[str, str]) -> dict[str, object]:
    # 合法域与构造期守卫共用常量;此处的检查保留是为了报错能点出 env 键名,
    # 构造期那道点的是字段名(两类调用方各看得懂自己那套)
    limiter_backend = _load_choice(env, "PGW_LIMITER_BACKEND", _LIMITER_BACKENDS, "memory")
    breaker_backend = _load_choice(env, "PGW_BREAKER_BACKEND", _BREAKER_BACKENDS, "memory")
    _, cache_backend = _require(env, "PGW_CACHE_BACKEND")
    _, telemetry_backend = _require(env, "PGW_TELEMETRY_BACKEND")
    if cache_backend not in _CACHE_BACKENDS:
        raise ValueError(f"PGW_CACHE_BACKEND 非法值 {cache_backend!r}")
    if telemetry_backend not in _TELEMETRY_BACKENDS:
        raise ValueError(f"PGW_TELEMETRY_BACKEND 非法值 {telemetry_backend!r}")
    redis_url = env.get("REDIS_URL") or None
    if "redis" in (limiter_backend, breaker_backend) and redis_url is None:
        raise ValueError("缺关键配置: 限流/熔断后端取 redis 需设置 REDIS_URL")
    auto_migrate = _load_schema_mode(env, telemetry_backend)
    return {
        "limiter_backend": limiter_backend,
        "breaker_backend": breaker_backend,
        "cache_backend": cache_backend,
        **_load_cache_keys(env, cache_backend, redis_url),
        "telemetry_backend": telemetry_backend,
        "telemetry_sqlite_path": _require(env, "PGW_TELEMETRY_SQLITE_PATH")[1]
        if telemetry_backend == "sqlite"
        else None,
        "telemetry_pg_dsn": _load_pg_dsn(env) if telemetry_backend == "postgres" else None,
        "telemetry_auto_migrate": auto_migrate,
        "telemetry_text_cap": _load_text_cap(env),
        "telemetry_pg_pool_max": _load_pool_max(env),
        "telemetry_pg_write_timeout_s": _load_write_timeout(env),
        "redis_url": redis_url,
        "pricing_path": env.get("PGW_PRICING_PATH") or None,
        "structured_max_retries": _load_structured_retries(env),
        "lease_ttl_s": _load_lease_ttl(env),
    }


def _load_schema_mode(env: Mapping[str, str], telemetry_backend: str) -> bool:
    """把 `PGW_TELEMETRY_SCHEMA_MODE` 的三态解成 `telemetry_auto_migrate`(issue #13)。

    三态: 键未设 → 按后端**不对称**派生;显式 auto/manual → 两侧都可覆盖。
    不对称的理由是两个后端的风险量级不同: SQLite 是下游自己的本地文件(没有
    DBA、没有迁移工具、没有第二个系统碰它),ALTER 是毫秒级元数据操作,要求
    手工跑 SQL 是给零运维场景强加运维步骤;PG 是共享的生产表,ALTER 取
    ACCESS EXCLUSIVE 锁会排在长事务后阻塞该表其后的所有查询,而遥测是业务
    路径上的内联 await。

    `_load_choice` 带 default,不能直接用来读这个键——default 会把"未设"和
    "设成默认值"抹平成同一种,三态就塌回两态,后端派生也就再没机会生效。故
    先用 `_first` 探"设没设",确认设了才交给 `_load_choice` 做值域校验(错误
    信息点出 env 键名这件事仍由它负责)。

    Args:
        env: 已合并的环境映射。
        telemetry_backend: 已校验过值域的遥测后端名。

    Returns:
        recorder 是否获准给旧表自动 ALTER 补列;backend=none 时无人消费,
        构造期守卫会再把它归一化为 False。
    """
    if _first(env, _SCHEMA_MODE_KEY) is None:
        return telemetry_backend == "sqlite"
    return _load_choice(env, _SCHEMA_MODE_KEY, _SCHEMA_MODES, "auto") == "auto"


def _load_text_cap(env: Mapping[str, str]) -> int | None:
    """读 `PGW_TELEMETRY_TEXT_CAP`(issue #12);键未设即 None = 不截断。

    与相邻的 `PGW_TELEMETRY_SCHEMA_MODE` 不同,这个键是**二态**而非三态:
    "未设"本身就是最终答案(不截断),没有需要按后端派生的第二种缺省,故不必像
    那边一样先探"设没设"再分两条路取值,读到什么解什么即可。

    值域(> 0)刻意不在此处判: 构造期守卫那道同时覆盖直接构造与
    `dataclasses.replace`,而报错文本已点出本键名,env 路的调用方不会看丢。

    Args:
        env: 已合并的环境映射。

    Returns:
        遥测正文的字符上限;键未设或为空串时返回 None(不截断)。
    """
    found = _first(env, _TEXT_CAP_KEY)
    if found is None:
        return None
    return int(_cast(found[1], "int", found[0]))


def _load_pool_max(env: Mapping[str, str]) -> int:
    """读 `PGW_TELEMETRY_PG_POOL_MAX`(issue #15);未设即缺省 4。

    与 `_load_text_cap` 同为二态键,只是"未设"落到一个具体缺省而非 None:
    池上限没有"不设上限"这一档——不表态就是继承第三方默认值,而那正是本 issue
    的病灶。值域(>= 1)留给构造期守卫,它同时覆盖直接构造与 `dataclasses.replace`。

    Args:
        env: 已合并的环境映射。

    Returns:
        遥测池允许的最大连接数。
    """
    found = _first(env, _POOL_MAX_KEY)
    if found is None:
        return _DEFAULT_PG_POOL_MAX
    return int(_cast(found[1], "int", found[0]))


def _load_write_timeout(env: Mapping[str, str]) -> float:
    """读 `PGW_TELEMETRY_PG_WRITE_TIMEOUT_S`(issue #15);未设即缺省 5.0 秒。

    这个值同时是 connect、acquire 与整次写入的上界: 遥测是业务路径上的内联
    await,"不设预算"不是一个允许存在的档位(铁律"丢一条 < 拖垮调用")。

    Args:
        env: 已合并的环境映射。

    Returns:
        单次遥测写入的硬预算(秒)。
    """
    found = _first(env, _WRITE_TIMEOUT_KEY)
    if found is None:
        return _DEFAULT_PG_WRITE_TIMEOUT_S
    return float(_cast(found[1], "float", found[0]))


def _strip_dsn_driver(dsn: str) -> str:
    """剥 SQLAlchemy 风格的 `+driver` 后缀(asyncpg 不认);已干净的原样返回。"""
    scheme, sep, rest = dsn.partition("://")
    return f"{scheme.partition('+')[0]}{sep}{rest}"


def _load_pg_dsn(env: Mapping[str, str]) -> str:
    """读取 Postgres DSN 并剥驱动后缀。

    env 路在此剥干净,构造期那道就无事可做——三项目 `.env` 里的 SQLAlchemy
    写法不会每次装配都刷一条 warning。
    """
    _, dsn = _require(env, "PGW_TELEMETRY_PG_DSN")
    return _strip_dsn_driver(dsn)


def _load_cache_keys(
    env: Mapping[str, str], cache_backend: str, redis_url: str | None
) -> dict[str, object]:
    if cache_backend == "none":
        return {"cache_namespace": None, "cache_ttl_s": None}
    _, namespace = _require(env, "PGW_CACHE_NAMESPACE")
    key, ttl_raw = _require(env, "PGW_CACHE_TTL_S")
    ttl = int(_cast(ttl_raw, "int", key))
    if ttl <= 0:
        raise ValueError("PGW_CACHE_TTL_S(TTL)必须 > 0,禁止永不过期")
    if cache_backend == "redis" and redis_url is None:
        raise ValueError("缺关键配置: PGW_CACHE_BACKEND=redis 需设置 REDIS_URL")
    return {"cache_namespace": namespace, "cache_ttl_s": ttl}


def _load_structured_retries(env: Mapping[str, str]) -> int:
    found = _first(env, "PGW_STRUCTURED_MAX_RETRIES")
    value = int(_cast(found[1], "int", found[0])) if found else 2  # M2.5 迭代4: 1→2
    if value < 0:
        raise ValueError("PGW_STRUCTURED_MAX_RETRIES 不能为负")
    return value


def _load_lease_ttl(env: Mapping[str, str]) -> float:
    found = _first(env, "PGW_LEASE_TTL_S")
    return float(_cast(found[1], "float", found[0])) if found else _DEFAULT_LEASE_TTL_S


@dataclass(frozen=True)
class EmbeddingSettings:
    """Embedding scope 装配配置(M2 §7): 复用 GatewaySettings + embedding 专用键。

    专用键不进 GatewaySettings(LLM scope 不受影响): `{SCOPE}__BATCH_SIZE`
    必填(分批是行为关键,不设默认)、`{SCOPE}__NORMALIZE`/`{SCOPE}__EXPECTED_DIM`
    可选。cache/structured 键对 embedding 无意义,装配时忽略。
    """

    gateway: GatewaySettings
    batch_size: int
    normalize: bool = False
    expected_dim: int | None = None

    def __post_init__(self) -> None:
        """自身字段的域校验;内嵌的 gateway 由 `GatewaySettings.__post_init__` 自己把关。"""
        if self.batch_size < 1:
            raise ValueError(f"EmbeddingSettings.batch_size 必须 ≥ 1: {self.batch_size}")
        if self.expected_dim is not None and self.expected_dim < 1:
            raise ValueError(f"EmbeddingSettings.expected_dim 必须 ≥ 1: {self.expected_dim}")

    @classmethod
    def from_env(
        cls,
        scope: str = "EMBED",
        env: Mapping[str, str] | None = None,
        *,
        env_file: str = ".env",
    ) -> EmbeddingSettings:
        if env is None:
            env = {
                k: v for k, v in {**dotenv_values(env_file), **os.environ}.items() if v is not None
            }
        scope_u = scope.upper()
        gateway = GatewaySettings.from_env(scope_u, env=env)
        key, raw = _require(env, f"{scope_u}__BATCH_SIZE")
        batch_size = int(_cast(raw, "int", key))
        if batch_size < 1:
            raise ValueError(f"{scope_u}__BATCH_SIZE 必须 ≥ 1")
        norm = _first(env, f"{scope_u}__NORMALIZE")
        dim = _first(env, f"{scope_u}__EXPECTED_DIM")
        expected_dim = int(_cast(dim[1], "int", dim[0])) if dim else None
        if expected_dim is not None and expected_dim < 1:
            raise ValueError(f"{scope_u}__EXPECTED_DIM 必须 ≥ 1")
        return cls(
            gateway=gateway,
            batch_size=batch_size,
            normalize=bool(_cast(norm[1], "bool", norm[0])) if norm else False,
            expected_dim=expected_dim,
        )


@dataclass(frozen=True)
class OcrSettings:
    """OCR scope 装配配置(M3 设计 §3.4): 复用 GatewaySettings,无 OCR 专用键。

    cache/structured/pricing 键对 OCR 无意义,装配时忽略;TPM 闸不启用
    (tpm=0)、看门狗键不配(非流式)由源键缺省自然达成。api_key 惯例填
    "none"(MonkeyOCR 无鉴权,SourceConfig 非空校验用占位)。
    """

    gateway: GatewaySettings

    @classmethod
    def from_env(
        cls,
        scope: str = "OCR",
        env: Mapping[str, str] | None = None,
        *,
        env_file: str = ".env",
    ) -> OcrSettings:
        if env is None:
            env = {
                k: v for k, v in {**dotenv_values(env_file), **os.environ}.items() if v is not None
            }
        return cls(gateway=GatewaySettings.from_env(scope.upper(), env=env))