Blocking user functionality - TrentoCrowdAI/crowdhub-api GitHub Wiki
The Blocking user functionality essentially checks, using some rules, given by the designer of a workflow, if a user can submit more job of the same workflow.
Context
The context is a sort of box which contains one or more do-blocks, a worker of a platform can submit only a job in a context. The contexts are defined in the workflow entity.
Worker ID
The id of the worker are retrieved or calculated in base of the platform. On Figure-eight we can retrieve the id of the worker from the html of the task. On Toloka we create a random number as an ID and then store it in the LocalStorage of the browser, next time we try to retrieve the id from there instead of recreating it.
The logic
When a user opens a task some custom JavaScript code is executed, this code:
- gets or creates the worker ID
- gets the job ID from the html of the page
- send a POST request to the servant-api endpoint
/worker-of-workflowsspecifying the worker ID, the job ID and the crowdsourcing platform - the endpoint returns:
- 'OK' message if the user can work on that task, so the user has not already worked on another job in the same context. The user is stored in the
workers_of_workflowtable. - error message if the user is not allowed to submit an answer to that specific task. Even a custom error message specified by the designer of the workflow is returned. The error message is showed to the worker.
- 'OK' message if the user can work on that task, so the user has not already worked on another job in the same context. The user is stored in the
The unique constraint of the workers_of_workflow table implements itself the decision if a worker can submit an answer for a specified task.