"""
集中读取 .env / 环境变量。

使用方式：
    from config import settings
    print(settings.llm_model)
    print(settings.has_openai_key)
"""

import os
from dataclasses import dataclass
from pathlib import Path


def _load_dotenv(env_path: Path) -> None:
    """极简 .env 解析，避免引入 python-dotenv 依赖。"""
    if not env_path.is_file():
        return
    for raw_line in env_path.read_text(encoding="utf-8").splitlines():
        line = raw_line.strip()
        if not line or line.startswith("#"):
            continue
        if "=" not in line:
            continue
        key, _, value = line.partition("=")
        key = key.strip()
        value = value.strip().strip('"').strip("'")
        # 已存在则不覆盖（让真实环境变量优先）
        os.environ.setdefault(key, value)


_ROOT = Path(__file__).resolve().parent
_load_dotenv(_ROOT / ".env")


@dataclass(frozen=True)
class Settings:
    openai_api_key: str
    anthropic_api_key: str
    llm_provider: str
    llm_model: str
    llm_max_tokens: int
    rag_chroma_path: str
    rag_collection_name: str
    rag_embed_model: str
    validation_service_url: str
    log_level: str

    @property
    def has_openai_key(self) -> bool:
        return bool(self.openai_api_key) and self.openai_api_key != "your_openai_api_key_here"

    @property
    def has_anthropic_key(self) -> bool:
        return bool(self.anthropic_api_key) and self.anthropic_api_key != "your_anthropic_api_key_here"


def _int(name: str, default: int) -> int:
    raw = os.environ.get(name)
    try:
        return int(raw) if raw else default
    except ValueError:
        return default


settings = Settings(
    openai_api_key=os.environ.get("OPENAI_API_KEY", ""),
    anthropic_api_key=os.environ.get("ANTHROPIC_API_KEY", ""),
    llm_provider=os.environ.get("LLM_PROVIDER", "openai"),
    llm_model=os.environ.get("LLM_MODEL", "gpt-4o-mini"),
    llm_max_tokens=_int("LLM_MAX_TOKENS", 4096),
    rag_chroma_path=os.environ.get("RAG_CHROMA_PATH", "./db/chroma"),
    rag_collection_name=os.environ.get("RAG_COLLECTION_NAME", "jrxml_chunks"),
    rag_embed_model=os.environ.get("RAG_EMBED_MODEL", "sentence-transformers/all-MiniLM-L6-v2"),
    validation_service_url=os.environ.get("VALIDATION_SERVICE_URL", "http://localhost:8001"),
    log_level=os.environ.get("LOG_LEVEL", "INFO"),
)


if __name__ == "__main__":
    s = settings
    print(f"provider  = {s.llm_provider}")
    print(f"model     = {s.llm_model}")
    print(f"max_tok   = {s.llm_max_tokens}")
    print(f"openai?   = {s.has_openai_key}")
    print(f"anthro?   = {s.has_anthropic_key}")
    print(f"rag_path  = {s.rag_chroma_path}")