Skip to content

Value sharing

Info

Needs arta>=0.10.0.

Between conditions and actions

It is possible to share some informations between condition and action implementations.

It can be usefull when an action needs some data that was computed in a condition (e.g., sanity check use cases).

In the following example, a condition is computing the median of some input values and checking it. Then, the action retrieves this median value and uses it.

Two things have to be done for that:

  1. Add the **kwargs parameter in your functions' definition (validation and action functions) if not already there.
  2. Set some new subkeys in the input_data key.

Set the value (in a condition for example):

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
def is_median_above(
    values: list[float], 
    limit: float, 
    **kwargs: Any,  # (1)!
    ) -> bool:
    """Check if the median of some values is above limit."""
    median = statistics.median(values)
    # Store the value for later use by an action function
    kwargs["input_data"]["median"] = median  # (2)!
    kwargs["input_data"]["median_limit"] = limit
    return median > limit
  1. Add the **kwargs parameter.
  2. Set your value in the input_data.

Get the value (in an action for example):

1
2
3
4
5
6
def alert_on_median(**kwargs: Any) -> str:  # (1)!
    """Alert: "Median is too high: 13, limit is: 10."""
    return (
        f"Median is too high: {kwargs['input_data']['median']}, "  # (2)!
        f"limit is: {kwargs['input_data']['median_limit']}."
    )
  1. Add the **kwargs parameter.
  2. Get the value.

User extra arguments

The following code shows how to set custom user extra arguments (e.g., add_details) within the .apply_rules() method:

1
2
3
4
5
from arta import RulesEngine

eng = RulesEngine(config_path="/to/my/config/dir")

result = eng.apply_rules(input_data={...}, add_details=True)

Used inside action and/or condition functions:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
def my_action(**kwargs: Any) -> str:  # (1)!
    """A simple action function."""

    if kwargs["add_details"]:  # (2)!
        action_result = (
            f"Median is too high: {kwargs['input_data']['median']}, "
            f"limit is: {kwargs['input_data']['median_limit']}."
        )
    else:
        action_result = "Median is too high."

    return action_result
  1. Don't forget to add the **kwargs parameter.
  2. Straight use of the corresponding extra argument.