Apppublic

Ibrahim5020/lisanify-docling

sourceHugging Faceapache-2.0updated 8mo agoView on Hugging Face

0likes

App README

Lisanify Docling API

Document conversion API with OCR support for the Lisanify translation app.

Endpoints

`GET /health`

Health check endpoint.

`POST /convert`

Convert documents (PDF, DOCX, PPTX) to structured text blocks.

Parameters:

—file: Document file (multipart/form-data)
—ocr: OCR mode - off, auto, or always (default: auto)
—lang: Tesseract language code for OCR (default: eng)
—source_lang: Source language of document (default: eng)
—segment: Sentence segmentation - off or sentences (default: off)
—segment_lang: Language for segmentation (optional)

Supported Languages for OCR:

—English (eng), German (deu), Turkish (tur), French (fra)
—Spanish (spa), Italian (ita), Portuguese (por), Dutch (nld)
—Polish (pol), Russian (rus), Arabic (ara)
—Japanese (jpn), Korean (kor), Chinese Simplified (chisim), Chinese Traditional (chitra)

Example Usage

python

import requests

url = "https://ibrahim5020-lisanify-docling.hf.space/convert"

files = {"file": open("document.pdf", "rb")}
data = {"ocr": "always", "lang": "deu", "source_lang": "deu"}

response = requests.post(url, files=files, data=data)
result = response.json()

print(f"Extracted {len(result['blocks'])} text blocks")