lint_flag_programmatically

Lint Feature Flags Programmatically

This example demonstrates how to lint/validate a GO Feature Flag configuration programmatically without using the CLI.

Overview

This example shows how to:

1. Retrieve flag configurations from your storage
2. Unmarshal the configuration into DTOs
3. Validate each flag configuration programmatically

Prerequisites

• Access to the GO Feature Flag core module

Installation

Step 1: Install Dependencies

Install the required dependencies:

go get github.com/thomaspoignant/go-feature-flag/modules/core

Step 2: Retrieve Flag Configuration from Storage

myFlags := getMyFlagsFromStorage()

This simulates retrieving your flag configuration from a storage system (database, file system, API, etc.). In this example, it returns a YAML string, but you can adapt this to your specific storage mechanism.

Replace getMyFlagsFromStorage() with your actual retrieval logic.

Step 3: Unmarshal Flag Configuration

var flags map[string]dto.DTO
err := yaml.Unmarshal([]byte(myFlags), &flags)
if err != nil {
    panic(err) // Don't forget to handle errors appropriately in production code.
}

The configuration is unmarshaled from YAML format into a map where:

• Key: The flag key (identifier)
• Value: A dto.DTO object representing the flag configuration

Note: You can also use JSON or TOML formats depending on your storage format. Simply use the appropriate unmarshaler (json.Unmarshal or toml.Unmarshal).

Step 4: Validate Each Flag Configuration

for key, flagDto := range flags {
    convertedFlag := flagDto.Convert()
    if err := convertedFlag.IsValid(); err != nil {
        panic(fmt.Sprintf("Invalid flag %s: %s", key, err.Error()))
    }
}

For each flag:

1. Convert the DTO to a Flag object using flagDto.Convert()
2. Validate the flag using IsValid()
3. Handle any validation errors appropriately

Handle validation errors according to your application's error handling strategy.

Additional Resources

• GO Feature Flag Documentation
• YAML Package Documentation