ABBYY synchronous query

abbyy.py

import urllib.parse
import requests
import time
import xml.dom.minidom

def get_abbyy_transcription(doc, app, password):

    proxies = {}

    url_params = {
        'language': 'French',
        'profile': 'textExtraction',
        'exportFormat': 'txtUnstructured',
        'txtUnstructured:paragraphAsOneLine': True
    }
    server_url = 'https://cloud-eu.ocrsdk.com/v2'
    request_url = urllib.parse.urljoin(server_url, 'processImage')

    path = f'/Users/max.halford/alan_data/osteo_invoices/{os.path.basename(doc.doc_uri)}'
    with open(path, 'rb') as image_file:
        image_data = image_file.read()

    import xml.dom.minidom

    def parse_task_response(response):
        dom = xml.dom.minidom.parseString(response.text)
        task_node = dom.getElementsByTagName('task')[0]
        task = {
            'taskId': task_node.getAttribute('id'),
            'status': task_node.getAttribute('status')
        }
        if task['status'] == 'Completed':
            task['downloadUrl'] = task_node.getAttribute('resultUrl')
        return task

    response = requests.post(
        request_url,
        data=image_data,
        params=url_params,
        auth=(app, password),
        proxies=proxies
    )
    response.raise_for_status()

    task = parse_task_response(response)
    
    while True:
        url_params = {'taskId': task['taskId']}
        status_url = urllib.parse.urljoin(server_url, 'getTaskStatus')
        response = requests.get(
            status_url,
            params=url_params,
            auth=(app, password)
        )
        task = parse_task_response(response)
        if task['status'] == 'Completed':
            break
        time.sleep(1)
        
    file_response = requests.get(task['downloadUrl'], stream=True, proxies=proxies)
    buffer = io.BytesIO()
    shutil.copyfileobj(file_response.raw, buffer)
    text = buffer.getvalue().decode('utf-8')
    words = [w for w in text.splitlines() if w]
    
    return words