settings.py

"""
Static project configuration loaded from environment variables and .env.

Paths are anchored to the project root (the directory containing this package)
so the app works regardless of the current working directory.
"""
import os
from pathlib import Path
from dotenv import load_dotenv

load_dotenv()

# Project root: two levels up from this file (config/settings.py → project root)
_PROJECT_ROOT = Path(__file__).resolve().parent.parent


def _path(relative: str) -> str:
    """Return an absolute path anchored to the project root."""
    return str(_PROJECT_ROOT / relative)


# ---------------------------------------------------------------------------
# OpenRouter
# ---------------------------------------------------------------------------
OPENROUTER_API_KEY: str = os.getenv("OPENROUTER_API_KEY", "")
OPENROUTER_BASE_URL: str = os.getenv("OPENROUTER_BASE_URL", "https://openrouter.ai/api/v1").rstrip("/")
OPENROUTER_MODEL: str = os.getenv("OPENROUTER_MODEL", "openai/gpt-4o-mini")

# ---------------------------------------------------------------------------
# Ollama (embeddings — always local; LLM — fallback)
# ---------------------------------------------------------------------------
OLLAMA_BASE_URL: str = os.getenv("OLLAMA_BASE_URL", "http://127.0.0.1:11434").rstrip("/")
OLLAMA_LLM_MODEL: str = os.getenv("OLLAMA_LLM_MODEL", "llama3")
EMBEDDING_MODEL: str = os.getenv("OLLAMA_EMBEDDING_MODEL", "nomic-embed-text")

# ---------------------------------------------------------------------------
# ChromaDB
# ---------------------------------------------------------------------------
CHROMA_DB_PATH: str = _path("data/processed/chroma_db")

# ---------------------------------------------------------------------------
# Raw document directories
# ---------------------------------------------------------------------------
RAW_PDF_PATH: str = _path("data/raw/pdfs")
RAW_EXCEL_PATH: str = _path("data/raw/excel")
RAW_TEXT_PATH: str = _path("data/raw/text")
RAW_IMAGE_PATH: str = _path("data/raw/images")

# ---------------------------------------------------------------------------
# OCR
# ---------------------------------------------------------------------------
OCR_LANGUAGE: str = os.getenv("OCR_LANGUAGE", "eng")
TESSERACT_CMD: str = os.getenv("TESSERACT_CMD", "")
PDF_OCR_ENABLED: bool = os.getenv("PDF_OCR_ENABLED", "1") == "1"
PDF_OCR_MIN_CHARS: int = int(os.getenv("PDF_OCR_MIN_CHARS", "80"))

# ---------------------------------------------------------------------------
# Chunking
# ---------------------------------------------------------------------------
CHUNK_SIZE: int = 1000
CHUNK_OVERLAP: int = 200