This repository contains Camunda External Task Client written in Python3.
Implement your BPMN Service Task in Python3.
Python >= 3.7 is required
Add following line to requirements.txt
of your Python project.
git+https://github.com/trustfactors/camunda-external-task-client-python3.git/#egg=camunda-external-task-client-python3
Or use pip to install as shown below:
pip install camunda-external-task-client-python3
To run the examples provided in examples folder, you need to have Camunda running locally or somewhere.
To run Camunda locally with Postgres DB as backend, you can use docker-compose.yml file.
$> docker-compose -f docker-compose.yml up
To run the examples with Auth Basic provided in examples/examples_auth_basic folder, you need to have Camunda with AuthBasic, running locally or somewhere.
To run Camunda with AuthBasic locally with Postgres DB as backend, you can use docker-compose-auth.yml file.
$> docker-compose -f docker-compose-auth.yml up
- Make sure to have Camunda running.
- Create a simple process model with an External Service Task and define the topic as 'topicName'.
- Deploy the process to the Camunda BPM engine.
- In your Python code:
import time
from camunda.external_task.external_task import ExternalTask, TaskResult
from camunda.external_task.external_task_worker import ExternalTaskWorker
# configuration for the Client
default_config = {
"maxTasks": 1,
"lockDuration": 10000,
"asyncResponseTimeout": 5000,
"retries": 3,
"retryTimeout": 5000,
"sleepSeconds": 30
}
def handle_task(task: ExternalTask) -> TaskResult:
"""
This task handler you need to implement with your business logic.
After completion of business logic call either task.complete() or task.failure() or task.bpmn_error()
to report status of task to Camunda
"""
# add your business logic here
# ...
# mark task either complete/failure/bpmnError based on outcome of your business logic
failure, bpmn_error = random_true(), random_true() # this code simulate random failure
if failure:
# this marks task as failed in Camunda
return task.failure(error_message="task failed", error_details="failed task details",
max_retries=3, retry_timeout=5000)
elif bpmn_error:
return task.bpmn_error(error_code="BPMN_ERROR_CODE", error_message="BPMN Error occurred",
variables={"var1": "value1", "success": False})
# pass any output variables you may want to send to Camunda as dictionary to complete()
return task.complete({"var1": 1, "var2": "value"})
def random_true():
current_milli_time = int(round(time.time() * 1000))
return current_milli_time % 2 == 0
if __name__ == '__main__':
ExternalTaskWorker(worker_id="1", config=default_config).subscribe("topicName", handle_task)
External Tasks are service tasks whose execution differs particularly from the execution of other service tasks (e.g. Human Tasks). The execution works in a way that units of work are polled from the engine before being completed.
camunda-external-task-client-python allows you to create easily such client in Python3.
Camunda provides functionality to start a process instance for a given process definition.
To start a process instance, we can use start_process()
from engine_client.py
You can find a usage example here.
client = EngineClient()
resp_json = client.start_process(process_key="PARALLEL_STEPS_EXAMPLE", variables={"intVar": "1", "strVar": "hello"},
tenant_id="6172cdf0-7b32-4460-9da0-ded5107aa977", business_key=str(uuid.uuid1()))
ExternalTaskWorker(worker_id="1").subscribe("topicName", handle_task)
starts long polling of the Camunda engine for external tasks.
- Polling tasks from the engine works by performing a fetch & lock operation of tasks that have subscriptions. It then calls the handler function passed to
subscribe()
function. i.e.handle_task
in above example. - Long Polling is done periodically based on the
asyncResponseTimeout
configuration. Read more about Long Polling.
from camunda.external_task.external_task import ExternalTask, TaskResult
from camunda.external_task.external_task_worker import ExternalTaskWorker
def handle_task(task: ExternalTask) -> TaskResult:
# add your business logic here
# Complete the task
# pass any output variables you may want to send to Camunda as dictionary to complete()
return task.complete({"var1": 1, "var2": "value"})
ExternalTaskWorker(worker_id="1").subscribe("topicName", handle_task)
from camunda.external_task.external_task import ExternalTask, TaskResult
from camunda.external_task.external_task_worker import ExternalTaskWorker
def handle_task(task: ExternalTask) -> TaskResult:
# add your business logic here
# Handle task Failure
return task.failure(error_message="task failed", error_details="failed task details",
max_retries=3, retry_timeout=5000)
# This client/worker uses max_retries if no retries are previously set in the task
# if retries are previously set then it just decrements that count by one before reporting failure to Camunda
# when retries are zero, Camunda creates an incident which then manually needs to be looked into on Camunda Cockpit
ExternalTaskWorker(worker_id="1").subscribe("topicName", handle_task)
from camunda.external_task.external_task import ExternalTask, TaskResult
from camunda.external_task.external_task_worker import ExternalTaskWorker
def handle_task(task: ExternalTask) -> TaskResult:
# add your business logic here
# Handle a BPMN Failure
return task.bpmn_error(error_code="BPMN_ERROR", error_message="BPMN error occurred")
ExternalTaskWorker(worker_id="1" ).subscribe("topicName", handle_task)
from camunda.external_task.external_task import ExternalTask, TaskResult
from camunda.external_task.external_task_worker import ExternalTaskWorker
def handle_task(task: ExternalTask) -> TaskResult:
# add your business logic here
# get the process variable 'score'
score = task.get_variable("score")
if int(score) >= 100:
return task.complete(...)
else:
return task.failure(...)
ExternalTaskWorker().subscribe("topicName", handle_task)
Camunda provides functionality to send a message event to a running process instance.
You can read more about the message events here: https://docs.camunda.org/manual/7.13/reference/bpmn20/events/message-events/
In our to send a message event to a process instance, a new function called correlate_message()
is added to engine_client.py
We can correlate the message by:
- process_instance_id
- tenant_id
- business_key
- process_variables
You can find a usage example here.
client = EngineClient()
resp_json = client.correlate_message("CANCEL_MESSAGE", business_key="b4a6f392-12ab-11eb-80ef-acde48001122")
To create an EngineClient with AuthBasic simple
client = EngineClient(config={"auth_basic": {"username": "demo", "password": "demo"}})
resp_json = client.start_process(process_key="PARALLEL_STEPS_EXAMPLE", variables={"intVar": "1", "strVar": "hello"},
tenant_id="6172cdf0-7b32-4460-9da0-ded5107aa977", business_key=str(uuid.uuid1()))
To create an ExternalTaskWorker with AuthBasic simple
from camunda.external_task.external_task import ExternalTask, TaskResult
from camunda.external_task.external_task_worker import ExternalTaskWorker
config = {"auth_basic": {"username": "demo", "password": "demo"}}
def handle_task(task: ExternalTask) -> TaskResult:
# add your business logic here
# Complete the task
# pass any output variables you may want to send to Camunda as dictionary to complete()
return task.complete({"var1": 1, "var2": "value"})
ExternalTaskWorker(worker_id="1", config=config).subscribe("topicName", handle_task)
The source files in this repository are made available under the Apache License Version 2.0.