"""central configuration for RAG application"""

from __future__ import annotations

import os
from dataclasses import dataclass

from dotenv import load_dotenv

load_dotenv()


@dataclass(frozen=True)
class Settings:
    dataset_id: str = os.getenv("RAG_DATASET_ID", "rag-datasets/rag-mini-wikipedia")
    dataset_config: str = os.getenv("RAG_DATASET_CONFIG", "text-corpus")
    dataset_split: str = os.getenv("RAG_DATASET_SPLIT", "passages")

    generation_model: str = os.getenv(
        "GENERATION_MODEL", "Qwen/Qwen3-4B-Instruct-2507"
    )
    embedding_model: str = os.getenv(
        "EMBEDDING_MODEL", "sentence-transformers/all-MiniLM-L6-v2"
    )
    reranker_model: str = os.getenv(
        "RERANKER_MODEL", "cross-encoder/ms-marco-MiniLM-L6-v2"
    )

    cache_dir: str = os.getenv("RAG_CACHE_DIR", ".rag_cache")
    max_documents: int = int(os.getenv("MAX_DOCUMENTS", "5000"))
    chunk_size: int = int(os.getenv("CHUNK_SIZE", "900"))
    chunk_overlap: int = int(os.getenv("CHUNK_OVERLAP", "140"))
    candidate_count: int = int(os.getenv("CANDIDATE_COUNT", "18"))

    hf_token: str | None = os.getenv("HF_TOKEN") or os.getenv("HUGGING_FACE_HUB_TOKEN")


SETTINGS = Settings()