Configurator is a free, open-source tool used to configure CI/CD pipelines by injesting input arguments and producing custom environment configuration in an easy, reliable, testable and easily containerized way.
install-package Configurator.Core
install-package Configurator.AspNetCore
Following package installation create a implementation of IArguments that will define input arguments. This example there are 4 simple arguments.
- Build - boolean
- Test - boolean
- Branch - string
- Confirmation - boolean
public class Arguments : IArguments
{
public Arguments(...)
{
...
}
public bool Build { get; private set; }
public bool Test { get; private set; }
public string Branch { get; private set; }
public bool Confirmation { get; private set; }
}
Create an implementation of IArgumentExtractor<> to get the arguments from IConfiguration into the IArgument implementation. IArgumentExtractor defines one method and it is used to extract arguments from IConfiguration.
Extractis used, well, to extract...
Example:
public class ArgumentExtractor : IArgumentExtractor<Arguments>
{
public const string BUILD = "build";
public const string TEST = "test";
public const string BRANCH = "branch";
public const string CONFIRMATION = "confirmation";
public Arguments Extract(IConfiguration config)
{
var build = TryParseBool(config, BUILD);
var test = TryParseBool(config, TEST);
var branch = config[BRANCH];
var confirmation = TryParseBool(config, CONFIRMATION);
var args = new Arguments(build, test, branch, confirmation);
return args;
}
private bool TryParseBool(IConfiguration config, string key)
{
if (!bool.TryParse(config[key], out bool value))
{
throw new ArgumentException($"Could not parse boolean value for {key}: \"{config[key]}\".");
}
return value;
}
}
Create a class that will define your pipeline configuration output, what you will use to control your CI/CD.
public class Output : IOutput
{
public Output() { }
public bool Build { get; set; }
public bool Test { get; set; }
public Git Git { get; set; }
public bool Confirmation { get; set; }
}
Output Processors are what convert some or all of the IArguments implementation into sections of the IOutput implementation. A processor is a subclass of BaseOutputProcessor<TProperty, TOutput, TArguments> which requires two abstract members to be implemented.
OutputPropertySelectoris used to define which property ofIOutputwill be set using this processor.Processis to calculate that value using theIArgumentsimplementation.
Example:
public class BuildProcessor : BaseOutputProcessor<bool, Output, Arguments>
{
public override Expression<Func<Output, bool>> OutputPropertySelector => (x) => x.Build;
public override bool Process(Arguments arguments)
{
return arguments.Build || arguments.Test;
}
}
Add Configurator using PipelineConfigurationExtensions.AddConfigurator(). This will return a IPipelineBuilder which will allow chaining of configurations in a fluent manner.
var builder = PiplineConfigurationExtensions.AddConfigurator<Arguments, Output>()
// Add Environment Variable configuration
.AddEnvironmentVariableArguments()
// Add Json file configuration
.AddJsonFileArguments("appsettings.json")
// Add BuildProcessor to the pipeline
.AddProcessor<BuildProcessor>()
// Add some ways to display the output
// Console
.AddConsoleGenerator()
// .yaml file generator
.AddYamlGenerator();
// Build the pipeline executor
var executor = builder.Build();
// Execute the pipeline and run Configurator
executor.Execute();
Use the sample powershell scripts as a base for how to build and run the sample project in docker.
build.ps1Build the dockerfilerun-env.ps1Run the docker image, uses environment variables for configurationrun-appsettings.ps1Run the docker image, uses the appsettings.json file for configuration
Both scripts use -v to mount an output directory so that any output files are persisted to the local machine.
Use Visual Studio debugging to run the sample project, there are no environment variables configured, only appsettings.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent