config.py

"""Settings loaded from backend/.env (or real environment variables in prod)."""

from functools import cached_property, lru_cache
from pathlib import Path

from pydantic_settings import BaseSettings, SettingsConfigDict

BACKEND_DIR = Path(__file__).resolve().parent
# Repo root in dev (backend/ sits under finsight-v3/). In Docker the image root
# is /app, so parent would wrongly resolve to /.
ROOT = BACKEND_DIR.parent


def data_root() -> Path:
    """Directory that local_path values are relative to."""
    if (BACKEND_DIR / "data").is_dir():
        return BACKEND_DIR
    parent = BACKEND_DIR.parent
    if (parent / "data").is_dir():
        return parent
    return BACKEND_DIR


def resolve_data_path(local_path: str) -> Path:
    """Normalize DB paths (Windows backslashes) and resolve under data_root."""
    return data_root() / local_path.replace("\\", "/")


class Settings(BaseSettings):
    model_config = SettingsConfigDict(
        env_file=BACKEND_DIR / ".env", env_file_encoding="utf-8", extra="ignore"
    )

    openai_api_key: str = ""
    openai_model: str = "gpt-4o-mini"

    database_url: str = ""
    supabase_url: str = ""
    supabase_service_role_key: str = ""

    embedding_model: str = "sentence-transformers/all-MiniLM-L6-v2"
    retrieval_k: int = 6
    max_tool_rounds: int = 6

    # Comma-separated frontend origins. Plain string (not list) so pydantic
    # doesn't JSON-decode a bare URL at startup.
    cors_origins: str = "http://localhost:8080,http://127.0.0.1:8080,http://localhost:5173"

    sec_edgar_user_agent: str = ""

    @cached_property
    def cors_origin_list(self) -> list[str]:
        return [origin.strip() for origin in self.cors_origins.split(",") if origin.strip()]


@lru_cache
def get_settings() -> Settings:
    return Settings()