app/config.py

from __future__ import annotations

import os
from dataclasses import dataclass
from dotenv import load_dotenv

load_dotenv()


def _clean_env_value(value: str) -> str:
    value = (value or "").strip()
    if (value.startswith('"') and value.endswith('"')) or (value.startswith("'") and value.endswith("'")):
        value = value[1:-1].strip()
    return value


def _to_bool(value: str, default: bool = False) -> bool:
    if value is None:
        return default
    value = _clean_env_value(value).lower()
    return value in {"1", "true", "yes", "y", "on"}


@dataclass
class Settings:
    api_keys: list[str]
    base_url: str
    model: str
    max_workers: int
    video_input_mode: str
    video_mime_type: str
    video_fps: int
    output_dir: str

    enable_rag_final: bool

    alibaba_cloud_access_key_id: str
    alibaba_cloud_access_key_secret: str
    bailian_workspace_id: str
    bailian_index_id: str
    bailian_retrieve_topn: int
    bailian_retrieve_enable_rerank: bool
    bailian_retrieve_dense_topk: int
    bailian_retrieve_sparse_topk: int
    bailian_retrieve_min_score: float

    enable_video_preprocess: bool
    video_preprocess_mode: str
    video_preprocess_remove_audio: bool
    preprocessed_video_dir: str

    enable_audio_agent: bool
    audio_asr_model: str
    audio_chunk_seconds: int
    extracted_audio_dir: str

    @staticmethod
    def load() -> "Settings":
        raw_keys = _clean_env_value(os.getenv("DASHSCOPE_API_KEYS", ""))
        api_keys = []
        for item in raw_keys.split(","):
            k = _clean_env_value(item)
            if k:
                api_keys.append(k)

        if not api_keys:
            raise ValueError("DASHSCOPE_API_KEYS 为空，请在 .env 中配置至少一个有效 API Key。")

        video_input_mode = _clean_env_value(os.getenv("VIDEO_INPUT_MODE", "base64")).lower()
        if video_input_mode not in {"base64", "remote_url"}:
            raise ValueError("VIDEO_INPUT_MODE 只能是 base64 或 remote_url。")

        enable_rag_final = _to_bool(os.getenv("ENABLE_RAG_FINAL", "false"))

        settings = Settings(
            api_keys=api_keys,
            base_url=_clean_env_value(
                os.getenv("DASHSCOPE_BASE_URL", "https://dashscope.aliyuncs.com/compatible-mode/v1")
            ),
            model=_clean_env_value(os.getenv("QWEN_MODEL", "qwen3.5-plus")),

            # 云端更稳：默认从 6 改成 2
            max_workers=int(_clean_env_value(os.getenv("MAX_WORKERS", "2"))),

            video_input_mode=video_input_mode,
            video_mime_type=_clean_env_value(os.getenv("VIDEO_MIME_TYPE", "video/mp4")),
            video_fps=int(_clean_env_value(os.getenv("VIDEO_FPS", "2"))),
            output_dir=_clean_env_value(os.getenv("OUTPUT_DIR", "outputs")),

            enable_rag_final=enable_rag_final,

            alibaba_cloud_access_key_id=_clean_env_value(os.getenv("ALIBABA_CLOUD_ACCESS_KEY_ID", "")),
            alibaba_cloud_access_key_secret=_clean_env_value(os.getenv("ALIBABA_CLOUD_ACCESS_KEY_SECRET", "")),
            bailian_workspace_id=_clean_env_value(os.getenv("BAILIAN_WORKSPACE_ID", "")),
            bailian_index_id=_clean_env_value(os.getenv("BAILIAN_INDEX_ID", "")),
            bailian_retrieve_topn=int(_clean_env_value(os.getenv("BAILIAN_RETRIEVE_TOPN", "6"))),
            bailian_retrieve_enable_rerank=_to_bool(os.getenv("BAILIAN_RETRIEVE_ENABLE_RERANK", "true")),
            bailian_retrieve_dense_topk=int(_clean_env_value(os.getenv("BAILIAN_RETRIEVE_DENSE_TOPK", "20"))),
            bailian_retrieve_sparse_topk=int(_clean_env_value(os.getenv("BAILIAN_RETRIEVE_SPARSE_TOPK", "20"))),
            bailian_retrieve_min_score=float(_clean_env_value(os.getenv("BAILIAN_RETRIEVE_MIN_SCORE", "0.15"))),

            enable_video_preprocess=_to_bool(os.getenv("ENABLE_VIDEO_PREPROCESS", "true")),

            # 云端更稳：默认从 analysis 改成 preview
            video_preprocess_mode=_clean_env_value(os.getenv("VIDEO_PREPROCESS_MODE", "preview")),

            video_preprocess_remove_audio=_to_bool(os.getenv("VIDEO_PREPROCESS_REMOVE_AUDIO", "false")),
            preprocessed_video_dir=_clean_env_value(os.getenv("PREPROCESSED_VIDEO_DIR", "preprocessed_videos")),

            enable_audio_agent=_to_bool(os.getenv("ENABLE_AUDIO_AGENT", "true")),
            audio_asr_model=_clean_env_value(os.getenv("AUDIO_ASR_MODEL", "qwen3-asr-flash")),
            audio_chunk_seconds=int(_clean_env_value(os.getenv("AUDIO_CHUNK_SECONDS", "290"))),
            extracted_audio_dir=_clean_env_value(os.getenv("EXTRACTED_AUDIO_DIR", "extracted_audio")),
        )

        if settings.enable_rag_final:
            missing = []
            if not settings.alibaba_cloud_access_key_id:
                missing.append("ALIBABA_CLOUD_ACCESS_KEY_ID")
            if not settings.alibaba_cloud_access_key_secret:
                missing.append("ALIBABA_CLOUD_ACCESS_KEY_SECRET")
            if not settings.bailian_workspace_id:
                missing.append("BAILIAN_WORKSPACE_ID")
            if not settings.bailian_index_id:
                missing.append("BAILIAN_INDEX_ID")
            if missing:
                raise ValueError(
                    "ENABLE_RAG_FINAL=true 时，以下环境变量必须配置：" + ", ".join(missing)
                )
        return settings