docling-studio/document-parser/infra/settings.py
Pier-Jean Malandrino 6327b13614 fix: pass max_num_pages and max_file_size to Docling convert (C3)
Defense-in-depth: even if upload validation passes, Docling itself
now enforces page count and file size limits. Configurable via
MAX_PAGE_COUNT and MAX_FILE_SIZE env vars (0 = unlimited).

Ref #57 (C3)
2026-04-07 15:32:38 +02:00

49 lines
2.2 KiB
Python

"""Centralized application settings — loaded from environment variables."""
from __future__ import annotations
import os
from dataclasses import dataclass, field
@dataclass(frozen=True)
class Settings:
app_version: str = "dev"
conversion_engine: str = "local" # "local" or "remote"
deployment_mode: str = "self-hosted" # "self-hosted" or "huggingface"
docling_serve_url: str = "http://localhost:5001"
docling_serve_api_key: str | None = None
conversion_timeout: int = 900
document_timeout: float = 120.0 # Docling-level per-document timeout (seconds)
max_concurrent_analyses: int = 3
max_page_count: int = 0 # 0 = unlimited (upload validation)
max_file_size: int = 0 # 0 = unlimited (Docling-level, bytes)
upload_dir: str = "./uploads"
db_path: str = "./data/docling_studio.db"
cors_origins: list[str] = field(
default_factory=lambda: ["http://localhost:3000", "http://localhost:5173"]
)
@classmethod
def from_env(cls) -> Settings:
"""Build a Settings instance from environment variables."""
cors_raw = os.environ.get("CORS_ORIGINS", "http://localhost:3000,http://localhost:5173")
return cls(
app_version=os.environ.get("APP_VERSION", "dev"),
conversion_engine=os.environ.get("CONVERSION_ENGINE", "local"),
deployment_mode=os.environ.get("DEPLOYMENT_MODE", "self-hosted"),
docling_serve_url=os.environ.get("DOCLING_SERVE_URL", "http://localhost:5001"),
docling_serve_api_key=os.environ.get("DOCLING_SERVE_API_KEY"),
conversion_timeout=int(os.environ.get("CONVERSION_TIMEOUT", "900")),
document_timeout=float(os.environ.get("DOCUMENT_TIMEOUT", "120.0")),
max_concurrent_analyses=int(os.environ.get("MAX_CONCURRENT_ANALYSES", "3")),
max_page_count=int(os.environ.get("MAX_PAGE_COUNT", "0")),
max_file_size=int(os.environ.get("MAX_FILE_SIZE", "0")),
upload_dir=os.environ.get("UPLOAD_DIR", "./uploads"),
db_path=os.environ.get("DB_PATH", "./data/docling_studio.db"),
cors_origins=[o.strip() for o in cors_raw.split(",")],
)
# Module-level singleton — import this from other modules.
settings = Settings.from_env()