from __future__ import annotations import asyncio from contextlib import asynccontextmanager from fastapi import FastAPI from fastapi.middleware.cors import CORSMiddleware from fastapi.middleware.gzip import GZipMiddleware from app.config import get_settings from app.core.database import pool_manager from app.core.logger import get_logger from app.services.embeddings_service import EmbeddingService from app.api.v1.router import api_v1_router _logger = get_logger(__name__) _settings = get_settings() _embedding_service: EmbeddingService = EmbeddingService() async def _self_ping(): import httpx health_url = _settings.self_ping_url while True: try: async with httpx.AsyncClient(timeout=30.0) as client: response = await client.get(health_url) if response.status_code == 200: _logger.info("Self-ping successful: %s", health_url) else: _logger.warning("Self-ping returned: %s - %s", health_url, response.status_code) except Exception as exc: _logger.error("Self-ping error: %s", exc) await asyncio.sleep(900) @asynccontextmanager async def lifespan(app: FastAPI): _logger.info("Initializing embedding service (loading all models)...") loop = asyncio.get_running_loop() await loop.run_in_executor(None, _embedding_service.load_all_models) await loop.run_in_executor(None, _embedding_service.load_vision_model) _logger.info("Embedding service initialized with dims: %s, vision=%s", _embedding_service.loaded_dimensions, _embedding_service._vision_loaded) asyncio.create_task(_self_ping()) yield _logger.info("Shutting down database connection pools...") await pool_manager.close_all() def create_application() -> FastAPI: app = FastAPI( title=_settings.app_name, description="All API Collection - Document extraction, conversion, and database query API.", version=_settings.app_version, docs_url="/docs", redoc_url="/redoc", openapi_tags=[ {"name": "Convert", "description": "Single-file and single-URL conversion"}, {"name": "Batch", "description": "Bulk conversion of files and URLs"}, {"name": "System", "description": "Health, info, and supported formats"}, {"name": "Embeddings", "description": "Text embedding generation using transformer models"}, {"name": "Verify", "description": "Phone number and identity verification"}, ], lifespan=lifespan, ) app.add_middleware(GZipMiddleware, minimum_size=1000) app.add_middleware( CORSMiddleware, allow_origins=["*"], allow_methods=["*"], allow_headers=["*"], ) app.include_router(api_v1_router, prefix="/api/v1") @app.get("/", include_in_schema=False) async def root(): return {"message": f"{_settings.app_name} v{_settings.app_version} is running"} @app.get("/health", include_in_schema=False) async def root_health(): return {"status": "ok", "version": _settings.app_version} @app.get("/ping", include_in_schema=False) async def ping(): return {"name": f"{_settings.app_name}", "version": _settings.app_version} return app app = create_application()