feat(aci): Track DataCondition errors and propagate such results as tainted

[kcons]

DataCondition failures are currently reported to Sentry/logs individually, then treated as a non-triggering.
Here we return a special error type, so that the caller knows the result isn't necessarily reliable, and in DataConditionGroup logic we aggregate these results in such a way that we know at the end if an untrustworthy result influenced the outcome.

These "tainted" results are reported for each Detector evaluation, so we can track how often our results were influenced by errors, giving us a better high level picture of user impact than we'd have purely by looking at individual reported error counts.

This pattern will be extended to cover over DataCondition evaluation cases.

[codecov]

❌ 2 Tests Failed:

Tests completed	Failed	Passed	Skipped
29552	2	29550	243

View the top 2 failed test(s) by shortest run time

tests.sentry.workflow_engine.models.test_data_condition.EvaluateValueTest::test_condition_result_comparison_fails

Stack Traces | 1.11s run time

#x1B[1m#x1B[.../workflow_engine/models/test_data_condition.py#x1B[0m:97: in test_condition_result_comparison_fails
    assert dc.evaluate_value(2) is None
#x1B[1m#x1B[31mE   AssertionError: assert ConditionError(msg='Invalid condition result') is None#x1B[0m
#x1B[1m#x1B[31mE    +  where ConditionError(msg='Invalid condition result') = <bound method DataCondition.evaluate_value of <DataCondition at 0x7f3fec3b9130: id=129, type=<Condition.GREATER: 'gt'>, comparison=1.0, condition_result='wrong', condition_group_id=86>>(2)#x1B[0m
#x1B[1m#x1B[31mE    +    where <bound method DataCondition.evaluate_value of <DataCondition at 0x7f3fec3b9130: id=129, type=<Condition.GREATER: 'gt'>, comparison=1.0, condition_result='wrong', condition_group_id=86>> = <DataCondition at 0x7f3fec3b9130: id=129, type=<Condition.GREATER: 'gt'>, comparison=1.0, condition_result='wrong', condition_group_id=86>.evaluate_value#x1B[0m

tests.sentry.workflow_engine.models.test_data_condition.GetConditionResultTest::test_str

Stack Traces | 1.15s run time

#x1B[1m#x1B[.../workflow_engine/models/test_data_condition.py#x1B[0m:21: in test_str
    assert dc.get_condition_result() is None
#x1B[1m#x1B[31mE   AssertionError: assert ConditionError(msg='Invalid condition result') is None#x1B[0m
#x1B[1m#x1B[31mE    +  where ConditionError(msg='Invalid condition result') = <bound method DataCondition.get_condition_result of <DataCondition at 0x7f1e28340cb0: id=144, type=<Condition.EQUAL: 'eq'>, comparison='10', condition_result='wrong', condition_group_id=96>>()#x1B[0m
#x1B[1m#x1B[31mE    +    where <bound method DataCondition.get_condition_result of <DataCondition at 0x7f1e28340cb0: id=144, type=<Condition.EQUAL: 'eq'>, comparison='10', condition_result='wrong', condition_group_id=96>> = <DataCondition at 0x7f1e28340cb0: id=144, type=<Condition.EQUAL: 'eq'>, comparison='10', condition_result='wrong', condition_group_id=96>.get_condition_result#x1B[0m

To view more test analytics, go to the Test Analytics Dashboard
_{📋 Got 3 mins? Take this short survey to help us improve Test Analytics.}

[kcons]

TODO: document how DataConditions should handle failure.