Flagd Evaluation Engine

A JSONLogic-based evaluation engine for flagd, providing local flag evaluation with support for targeting rules, fractional rollouts, and semantic version comparisons.

Features

• JSONLogic-based Targeting: Support for complex targeting rules using JSONLogic
• Fractional Rollouts: Consistent hashing (murmurhash3) for percentage-based rollouts
• Semantic Versioning: Compare values using semver operators (^, ~, =, !=, <, <=, >, >=)
• String Operations: Custom operators for "starts_with" and "ends_with" comparisons

Installation

Add the dependency in your Cargo.toml:

cargo add flagd-evaluation-engine

Usage

use flagd_evaluation_engine::{FlagdEvaluationError, FeatureFlag, FlagParser, Operator};
use open_feature::EvaluationContext;

fn main() {
    // Create the operator for evaluating targeting rules
    let operator = Operator::new();
    // Create evaluation context
    let context = EvaluationContext::default()
        .with_targeting_key("user-123")
        .with_custom_field("tier", "premium");

    // Apply a targeting rule (synchronous)
    let result = operator.apply(
        "my-flag",
        r#"{"if": [{"==": [{"var": "tier"}, "premium"]}, "gold", "silver"]}"#,
        &context,
    );
    println!("Result: {:?}", result);
}

License

Apache 2.0 - See LICENSE for more information.