rag_server/applications/utils/pdf/book_extract.py

import requests
# -*- coding: utf-8 -*-


async def book_extract(book_path, book_id):
    with open(book_path, "rb") as f:
        files = {"files": (book_id, f, "application/pdf")}
        response = requests.post(
            "http://192.168.100.31:8003/file_parse",
            headers={"accept": "application/json"},
            data={
                "return_model_output": "false",
                "return_md": "false",
                "return_images": "false",
                "end_page_id": "99999",
                "parse_method": "auto",
                "start_page_id": "0",
                "lang_list": "ch",
                "output_dir": "./output",
                "server_url": "string",
                "return_content_list": "true",
                "backend": "pipeline",
                "table_enable": "true",
                "response_format_zip": "false",
                "formula_enable": "true",
            },
            files=files,
        )
    return response.json()