Update app.py

app.py

@@ -9,6 +9,25 @@ from pathlib import Path
 from fastapi import FastAPI, UploadFile, File, HTTPException
 from docling.datamodel.base_models import InputFormat
 from docling.document_converter import DocumentConverter
+from docling.datamodel.pipeline_options import PdfPipelineOptions, EasyOcrOptions
+from docling.document_converter import PdfFormatOption
+from docling.datamodel.base_models import InputFormat
+
+md = MarkItDown()
+
+UPLOAD_DIR = Path("temp_uploads")
+UPLOAD_DIR.mkdir(exist_ok=True)
+
+pipeline_options = PdfPipelineOptions()
+pipeline_options.ocr_options = EasyOcrOptions()
+
+converter = DocumentConverter(
+    format_options={
+        InputFormat.PDF: PdfFormatOption(pipeline_options=pipeline_options)
+    }
+)
+
+print("Converter initialized successfully with EasyOCR!")
 
 app = FastAPI()
 
@@ -103,11 +122,7 @@ async def extract(file: UploadFile = File(...)):
         "stream": stream
     }
 
-md = MarkItDown()
 
-converter = DocumentConverter()
-UPLOAD_DIR = Path("temp_uploads")
-UPLOAD_DIR.mkdir(exist_ok=True)
 
 @app.post("/docling")
 async def convert_document(file: UploadFile = File(...)):