This project is deprecated and is no longer being actively maintained.
goforit is an experimental, quick-and-dirty client library for feature flags in Go.
Feature flags can be stored in any desired backend. goforit provides a several flatfile implementations out-of-the-box, so feature flags can be defined in a JSON or CSV file. See below for details.
Alternatively, flags can be stored in a key-value store like Consul or Redis.
Create a CSV file that defines the flag names and sampling rates:
go.sun.money,0
go.moon.mercury,1
go.stars.money,.5
func main() {
ctx := context.Background()
// flags.csv contains comma-separated flag names and sample rates.
// See: testdata/flags_example.csv
backend := goforit.BackendFromFile("flags.csv")
goforit.Init(30*time.Second, backend)
if goforit.Enabled(ctx, "go.sun.mercury", nil) {
fmt.Println("The go.sun.mercury feature is enabled for 100% of requests")
}
if goforit.Enabled(ctx, "go.stars.money", nil) {
fmt.Println("The go.stars.money feature is enabled for 50% of requests")
}
}Included flatfile backends are:
This is a very simple backend, where every row defines a flag name and a rate at which it should be enabled, between zero and one. Initialize this backend with BackendFromFile. See an example.
This backend allows each flag to have multiple rules, like a series of if-statements. Each call to .Enabled() takes a map of properties, which rules can match against. Each rule's matching or non-matching can cause the overall flag to be on or off, or can fallthrough to the next rule. See the proposal for this system or an example JSON file. It's a bit confusing to understand.
In this format, each flag can have a number of rules, and each rule can contain a number of predicates for matching properties. When a flag is evaluated, it uses the first rule whose predicates match the given properties. See an example JSON file, that also includes test cases.
goforit is in an experimental state and may introduce breaking changes without notice.
goforit is available under the MIT license.