rule-engine.md

Rule Engine

pyfly.rule_engine is a YAML-based business rules engine: parse rules into an AST, evaluate them against a context dict, take actions.

Defining rules in YAML

id: order-rules
name: Order processing rules
rules:
  - id: high-value
    priority: 10
    when:
      op: ge
      field: order.amount
      value: 1000
    then:
      - { type: set, target: flags.high_value, value: true }
      - { type: log, value: "marking order as high-value" }
  - id: blocked-region
    when:
      op: in
      field: order.shipping_country
      value: ["XX", "YY"]
    then:
      - { type: set, target: flags.blocked, value: true }

Evaluating

from pyfly.rule_engine import RuleSetLoader, RuleSetEvaluator

ruleset = RuleSetLoader.from_yaml(yaml_text)
evaluator = RuleSetEvaluator()
ctx = {"order": {"amount": 5000, "shipping_country": "US"}, "flags": {}}
results = evaluator.evaluate(ruleset, ctx)
print(ctx["flags"])  # {"high_value": True}

Operators

Comparison: eq, ne, gt, ge, lt, le, in, not_in, regex. Logical: and, or, not (with conditions: [...]).

Action types: set (write context path), increment, log. Subclass RuleEvaluator._execute_action to support call, calculate, etc.