rulestone

Rulestone

Lightweight and fast rule engine written in Go, with API for other languages:

• Go (this!)
• Java

With Rulestone you can define thousands of rules and then process tens of thousands events/objects per second getting the matching rules for each object.

Installation

Install the package:

go get github.com/atlasgurus/rulestone

Usage

Go

The following Go example shows how to load a rule from file and match an object against this rule:

    package main

    import (
        "fmt"
        "github.com/atlasgurus/rulestone/utils"
        "github.com/atlasgurus/rulestone/engine"
    )

    func Match() {
        repo := engine.NewRuleEngineRepo()
        _, err := repo.RegisterRuleFromFile("rule.json")
        if err != nil {
            return
        }

        ruleEngine, err := engine.NewRuleEngine(repo)
        if err != nil {
            return
        }

        event, err := utils.ReadEvent("object.json")
        if err != nil {
            return
        }

        matches := ruleEngine.MatchEvent(event)

        for _, ruleId := range matches {
			// Optionally get matching rules metadata
            ruleDefinition := ruleEngine.GetRuleDefinition(ruleId)
            if ruleIdStr, ok := ruleDefinition.Metadata["rule_id"].(string); ok {
                fmt.Println("Rule matched: ", ruleIdStr)
            }
        }
        // Report all the errors if any
        if repo.GetAppCtx().NumErrors() > 0 {
            repo.GetAppCtx().PrintErrors()
        }
    }

    func main() {
        Match()
    }

The example assumes rule contains the metadata field called rule_id. See for more Go usage examples in tests/rule_api_test.go.

Rules

Simple rule definition

metadata:
  created: "2023-03-29"
  priority: 10
  rule_id: "BUSINESS_RULE_1"
expression: name == "Frank" && age == 20

The rule match if the JSON object has a name field with value Frank and an age field with value 20. The metadata section may store any information linked to the rule, including the rule ID. The condition section contains the expression that will be evaluated against the JSON object.

See examples/rules for more rules examples.

Operators and functions

Rulestone expressions supports:

• Comparison and negation operators like ==, >, >=, <, <=

• Arithmetic operations: +, -, *, /

• Logical operators: &&, ||, !

• Parentheses: (, )

• String literals: "string"

• Numeric literals: 1, 2.3

• Field access: field1, field1.field2

• Functions: hasValue, isEqualToAny, regexpMatch, date, forAll, forSome

• Date literals: date("11/29/1968")

• hasValue - check that object has specified field, for example hasValue(field1)

• isEqualToAny - check that object field is equal to any specified value, for example isEqualToAny(field1, 1, 2, 3, '4')

• regexpMatch - match the Go regexp, for example regexpMatch("^\\d{4}/\\d{2}/\\d{2}$", child.dob)

• forAll - test that a logical expression is true for all members of the list, for example forAll('children', 'child', child.age > 10)

• forSome - test that a logical expression is true for at least one member of the list, for example forSome('children', 'child', child.age > 10)

Dates

Rulestone handles dates and comparison operators on them, but since JSON doesn't provide field type information, need to use date() function. The function can parse date string in different formats:

metadata:
  created: "2023-03-29"
  priority: 10
expression: 'name == "Frank" && date(dob) < date(child.dob) && date("11/29/1968") > date(dob) && date(dob) == date("11/28/1968")'

Contributing

We love contributions! If you have any suggestions, bug reports, or feature requests, please open an issue in our tracker.

License

This project is licensed under the MIT License - see the LICENSE file for details.