How to Implement Delayed Retry Pattern on Taipy Core Task

I want to implement a retry pattern on my task API.

Can I do this without doing a try/catch myself? Can the next task wait few seconds before retrying if there is an error?

import requests
from taipy import Config
import taipy as tp


def api():
    r = requests.get('<url>')
    return r.json()


scc = Config.configure_scenario_from_tasks("sc",
                                          [Config.configure_task("api_call",
                                                                 api,
                                                                 [],
                                                                 [Config.configure_data_node("api_result")])
                                          ])

if __name__ == '__main__':
    tp.Core().run()
    scenario = tp.create_scenario(scc)
    tp.submit(scenario)

    print(scenario.api_result.read())

At the moment, there are no built-in retry patterns in Taipy Core.

However, a scenario can be subscribed to a callback. At each change of a status job (the execution of a task), this callback is called. If the job status is failed, you can execute some code. For example, you can re-submit the scenario:

from taipy.core import Scenario, Job
import taipy as tp

MAX_RETRY = 5

def subscribe_to_retry(scenario: Scenario, job: Job):
    print(scenario.name, job.status.name)
    if 'retry' not in scenario.properties.keys():
        scenario.properties['retry'] = 0
    if job.status.name == 'FAILED' and scenario.properties['retry'] < MAX_RETRY:
        scenario.properties['retry'] += 1
        scenario.submit()

# this mechanism will be applied to all the scenarios
tp.subscribe_scenario(subscribe_to_retry)

You could also directly put a retry pattern in your Taipy task's function.

def funct(retry=3):
    try:
       ...
        return ...
    except e:
        if retry > 0:
            return funct(retry-1)
        raise e