Asynchronous pdf extractor API made with django-rest framework and celery/rabbitmq worker, protected with JWT authentication
When we upload a PDF file to the /api/v1/create endpoint, this file is sent to the Celery in the backend and the record id that is linked to the authorized user id is returned and processing starts asynchronously. The file's content is extracted and saved to database. Celery storage is RabbitMQ. Also, you can use /api/v1/check/:id endpoint to track the status. If processing has been finalized then it returns the content that was extracted from the document.
List of endpoints:
- /api/v1/auth → JWT authorization
- /api/v1/refresh → JWT token refreshment
- /api/v1/create → as a POST payload this endpoint will receive the .pdf file and document name
- /api/v1/check/:id → get the status of processed file
Install dependencies
pip install -r requirements.txt
Set database
python manage.py makemigrations
python manage.py migrate
Note: I used MySql in this projects, you can change back to default Sqlite db in settings.py file before migrating
Create SuperUser
python manage.py createsuperuser
Run celery worker
celery -A pdfextract worker -l info --pool=solo
Note: This command should be different depending on your operating system
Run Django server
python manage.py runserver