src/gistillery/config.py

import os
from pathlib import Path

from pydantic import BaseSettings
from pydantic.types import PositiveInt


class Config(BaseSettings):
    hf_hub_token: str = "missing"
    hf_agent: str = "https://api-inference.huggingface.co/models/bigcode/starcoder"
    db_file_name: Path = Path("sqlite-data.db")
    processing_max_length: PositiveInt = 10000  # in characters
    sampling_rate: PositiveInt = 16_000  # audio transcription
    max_yt_length: PositiveInt = 1800  # in minutes
    pdf_stop_words: list[str] = [
        "ACKNOWLEDGMENTS",
        "Acknowledgments",
        "acknowledgments",
        "REFERENCES",
        "References",
        "references",
    ]

    class Config:
        # load .env file by default, with provisio to use other .env files if set
        env_file = os.getenv('ENV_FILE', '.env')


_config = None


def get_config() -> Config:
    global _config
    if _config is None:
        _config = Config()
    return _config