from __future__ import annotations

from functools import lru_cache
from pathlib import Path

from pydantic import Field, SecretStr
from pydantic_settings import BaseSettings, SettingsConfigDict


class Settings(BaseSettings):
    """Runtime settings. Secrets are `SecretStr` so that `repr()`/logging never prints them."""

    model_config = SettingsConfigDict(env_prefix="FINDATA_", env_file=".env", env_file_encoding="utf-8", extra="ignore")

    env: str = "development"
    root: Path = Field(default_factory=lambda: Path.cwd())
    lake: Path = Field(default_factory=lambda: Path.cwd() / "data" / "lake")
    database_url: SecretStr = SecretStr("postgresql://findata:findata-local@127.0.0.1:5432/findata")
    bind_host: str = "127.0.0.1"
    bind_port: int = 8600
    public_url: str = "http://127.0.0.1:8600"
    secret_key: SecretStr = SecretStr("dev-only-change-me-dev-only-change-me")
    lake_soft_limit_gb: float = 250.0
    lake_hard_min_free_gb: float = 60.0
    # second lake root on the external SSD (cold datasets + raw archives); must contain .findata-cold-root
    lake_cold: Path | None = None
    lake_cold_hard_min_free_gb: float = 100.0
    hot_datasets: str = ("market_bars.stock.1m.raw,market_bars.etf.1m.raw,market_bars.stock.1d.raw,market_bars.etf.1d.raw,"
                         "market_bars.stock.5m.raw,market_bars.etf.5m.raw")
    bulk_workers: int = 4
    backup_target: str = ""
    # vendor + optional providers (no FINDATA_ prefix: read explicitly)
    frd_userid: SecretStr = Field(default=SecretStr(""), validation_alias="FRD_USERID")
    sec_edgar_user_agent: str = Field(default="", validation_alias="SEC_EDGAR_USER_AGENT")
    anthropic_api_key: SecretStr = Field(default=SecretStr(""), validation_alias="ANTHROPIC_API_KEY")
    openai_api_key: SecretStr = Field(default=SecretStr(""), validation_alias="OPENAI_API_KEY")
    # web static bundle
    web_dist: Path | None = None
    # rate limiting defaults (per API key / session)
    rate_limit_per_minute: int = 600
    query_timeout_seconds: int = 30
    query_max_rows: int = 100_000
    bars_max_rows_json: int = 50_000
    export_ttl_hours: int = 72

    @property
    def is_production(self) -> bool:
        return self.env == "production"

    @property
    def hot_dataset_ids(self) -> set[str]:
        return {x.strip() for x in self.hot_datasets.split(",") if x.strip()}

    @property
    def raw_dir(self) -> Path:
        return (self.lake_cold or self.lake) / "raw"

    @property
    def curated_dir(self) -> Path:
        return self.lake / "curated"

    @property
    def exports_dir(self) -> Path:
        return self.lake / "exports"

    @property
    def tmp_dir(self) -> Path:
        return self.lake / "_tmp"

    def secret_values(self) -> list[str]:
        """Every secret string that must never appear in logs or error messages."""
        vals = [self.frd_userid.get_secret_value(), self.anthropic_api_key.get_secret_value(),
                self.openai_api_key.get_secret_value()]
        pw = self.database_url.get_secret_value()
        if "@" in pw and ":" in pw.split("@")[0]:
            vals.append(pw.split("@")[0].rsplit(":", 1)[-1])
        return [v for v in vals if v and len(v) >= 6]


@lru_cache
def get_settings() -> Settings:
    return Settings()