Spaces:
Sleeping
Sleeping
from fastapi import FastAPI, UploadFile, File, HTTPException | |
from fastapi.responses import JSONResponse | |
from fastapi.middleware.cors import CORSMiddleware | |
import shutil | |
import os | |
from uuid import uuid4 | |
from docling.document_converter import DocumentConverter | |
from threading import Lock | |
from concurrent.futures import ThreadPoolExecutor | |
import asyncio | |
app = FastAPI() | |
# CORS for all | |
app.add_middleware( | |
CORSMiddleware, | |
allow_origins=["*"], | |
allow_credentials=True, | |
allow_methods=["*"], | |
allow_headers=["*"], | |
) | |
# Thread pool sized according to vCPU | |
MAX_WORKERS = os.cpu_count() or 2 # fallback to 2 if os.cpu_count() returns None | |
thread_pool = ThreadPoolExecutor(max_workers=MAX_WORKERS) | |
# Singleton class for DocumentConverter | |
class PdfConverterSingleton: | |
_instance = None | |
_lock = Lock() | |
def __new__(cls): | |
if cls._instance is None: | |
with cls._lock: | |
if cls._instance is None: | |
instance = super().__new__(cls) | |
instance._initialize() | |
cls._instance = instance | |
return cls._instance | |
def _initialize(self): | |
self.converter = DocumentConverter() | |
def get_text(self, pdf_path: str) -> str: | |
result = self.converter.convert(pdf_path) | |
return result.document.export_to_markdown() | |
# Run sync function in threadpool | |
def sync_extract_text(pdf_path: str) -> str: | |
return PdfConverterSingleton().get_text(pdf_path) | |
# Async wrapper for thread pool | |
async def async_extract_text(pdf_path: str) -> str: | |
loop = asyncio.get_event_loop() | |
return await loop.run_in_executor(thread_pool, sync_extract_text, pdf_path) | |
# Main endpoint | |
async def extract_pdf_text(file: UploadFile = File(...)): | |
if file.content_type != "application/pdf": | |
raise HTTPException(status_code=400, detail="Only PDF files are supported.") | |
temp_filename = f"/tmp/{uuid4().hex}.pdf" | |
try: | |
with open(temp_filename, "wb") as buffer: | |
shutil.copyfileobj(file.file, buffer) | |
text = await async_extract_text(temp_filename) | |
return JSONResponse(content={"markdown_text": text}) | |
except Exception as e: | |
raise HTTPException(status_code=500, detail=str(e)) | |
finally: | |
if os.path.exists(temp_filename): | |
os.remove(temp_filename) | |