kami_spider_monorepo/core/config.py

"""
Core configuration management using Pydantic Settings.
All configuration loaded from environment variables with validation.
"""

from enum import StrEnum
from typing import Literal
from pydantic import Field, field_validator
from pydantic_settings import BaseSettings, SettingsConfigDict


class ProxyPoolType(StrEnum):
    """代理池类型枚举"""

    DEFAULT = "default"  # 默认代理池
    EXPIRING = "expiring"  # 带有效期的代理池


class Settings(BaseSettings):
    """
    Application settings with environment variable support.

    Configuration is loaded from environment variables and .env files.
    All settings are validated at startup using Pydantic validators.
    """

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

    # Application Settings
    app_name: str = Field(default="kami_spider", description="Application name")
    environment: Literal["development", "staging", "production"] = Field(
        default="development", description="Runtime environment"
    )
    debug: bool = Field(default=False, description="Debug mode")
    host: str = Field(default="0.0.0.0", description="Server host")
    port: int = Field(default=8000, description="Server port")
    workers: int = Field(default=1, description="Number of worker processes")
    log_level: Literal["DEBUG", "INFO", "WARNING", "ERROR", "CRITICAL"] = Field(
        default="INFO", description="Logging level"
    )

    # Database Settings
    db_host: str = Field(default="localhost", description="MySQL host")
    db_port: int = Field(default=3306, description="MySQL port")
    db_name: str = Field(default="kami_spider", description="Database name")
    db_user: str = Field(default="root", description="Database user")
    db_password: str = Field(default="", description="Database password")
    db_pool_size: int = Field(default=10, description="Database connection pool size")
    db_max_overflow: int = Field(
        default=20, description="Database max overflow connections"
    )
    db_pool_recycle: int = Field(
        default=3600, description="Database pool recycle time in seconds"
    )
    db_pool_pre_ping: bool = Field(
        default=True, description="Test connections before using"
    )
    db_echo: bool = Field(default=False, description="Echo SQL statements")

    # Redis Settings
    redis_host: str = Field(default="localhost", description="Redis host")
    redis_port: int = Field(default=6379, description="Redis port")
    redis_db: int = Field(default=0, description="Redis database number")
    redis_password: str = Field(default="", description="Redis password")
    redis_max_connections: int = Field(
        default=50, description="Redis connection pool max connections"
    )
    redis_decode_responses: bool = Field(
        default=True, description="Decode Redis responses to strings"
    )

    # OpenTelemetry Settings
    otel_enabled: bool = Field(default=True, description="Enable OpenTelemetry")
    otel_service_name: str = Field(
        default="kami_spider", description="Service name for traces"
    )
    otel_exporter_endpoint: str = Field(
        default="38.38.251.113:31547",
        description="OpenTelemetry collector gRPC endpoint",
    )
    otel_exporter_insecure: bool = Field(
        default=True, description="Use insecure gRPC connection"
    )
    otel_sample_rate: float = Field(
        default=1.0, description="Trace sampling rate (0.0 to 1.0)"
    )

    # CORS Settings
    cors_enabled: bool = Field(default=True, description="Enable CORS")
    cors_allow_origins: list[str] = Field(
        default=["*"], description="Allowed CORS origins"
    )
    cors_allow_credentials: bool = Field(
        default=True, description="Allow credentials in CORS"
    )
    cors_allow_methods: list[str] = Field(
        default=["*"], description="Allowed HTTP methods"
    )
    cors_allow_headers: list[str] = Field(
        default=["*"], description="Allowed HTTP headers"
    )

    # 代理设置
    proxy_enable: bool = Field(default=True, description="是否启用代理")
    proxy_url: str = Field(
        default="https://share.proxy.qg.net/get?key=7ASQH2BI&num=1&area=&isp=0&format=txt&seq=\r\n&distinct=false&area=510100",
        description="代理服务器地址",
    )
    proxy_type: ProxyPoolType = Field(
        default=ProxyPoolType.DEFAULT, description="代理服务器类型"
    )
    proxy_username: str = Field(default="7ASQH2BI", description="代理服务器用户名")
    proxy_password: str = Field(default="34D6652FE7B6", description="代理服务器密码")

    # Security Settings
    secret_key: str = Field(
        default="change-me-in-production", description="Secret key for signing tokens"
    )

    @field_validator("workers")
    @classmethod
    def validate_workers(cls, v: int) -> int:
        """Ensure workers is at least 1."""
        if v < 1:
            raise ValueError("workers must be at least 1")
        return v

    @field_validator("otel_sample_rate")
    @classmethod
    def validate_sample_rate(cls, v: float) -> float:
        """Ensure sample rate is between 0.0 and 1.0."""
        if not 0.0 <= v <= 1.0:
            raise ValueError("otel_sample_rate must be between 0.0 and 1.0")
        return v

    @property
    def database_url(self) -> str:
        """Generate async database URL for SQLModel/SQLAlchemy."""
        password_part = f":{self.db_password}" if self.db_password else ""
        return (
            f"mysql+aiomysql://{self.db_user}{password_part}"
            f"@{self.db_host}:{self.db_port}/{self.db_name}"
        )

    @property
    def sync_database_url(self) -> str:
        """Generate sync database URL for Alembic migrations."""
        password_part = f":{self.db_password}" if self.db_password else ""
        return (
            f"mysql+pymysql://{self.db_user}{password_part}"
            f"@{self.db_host}:{self.db_port}/{self.db_name}"
        )

    @property
    def redis_url(self) -> str:
        """Generate Redis URL."""
        password_part = f":{self.redis_password}@" if self.redis_password else ""
        return f"redis://{password_part}{self.redis_host}:{self.redis_port}/{self.redis_db}"

    @property
    def is_production(self) -> bool:
        """Check if running in production environment."""
        return self.environment == "production"

    @property
    def is_development(self) -> bool:
        """Check if running in development environment."""
        return self.environment == "development"


# Global settings instance
settings = Settings()