CLI utility for creating and managing development workspace

This repository contains the source code for the CLI utility to initialize and manage a development workspace.

Table of Contents

• Prerequisites
• Usage
• Commands
  • init command
  • run command
  • sync command
  • help command
• Workspace structure
  • Repository configuration
  • Server configuration
  • Secret templates
• Use cases
  • Available use cases
  • Creating use cases
• Template repository
• Development
  • Testing

Prerequisites

The CLI utility is based on Node.js and NPM. To install the utility, it is necessary to have Node.js and NPM installed.

Usage

The CLI utility can be executed via npx:

npx @cycrilabs/ws-ctrl@latest

This will pull the latest version of the utility from the NPM registry and execute it. The utility will then display the available commands and options.

Commands

A command is executed by appending it to the initial npx, for example for the init command:

npx @cycrilabs/ws-ctrl@latest init

The following table lists all available commands and options.

Command	Description
init	Initialize a new workspace
run	Run a specific use case in the workspace
sync	Syncs the templates
help	Displays help information

init command

The init command initializes a new workspace. After executing the command, the user is prompted to enter at least to location of new workspace. Alternatively, the location can be passed as an argument directly, e.g.

npx @cycrilabs/ws-ctrl@latest init ./platform

In addition, the user is prompted to specify whether a template repository should be used, and if so, which one.

Furthermore, the following options are available:

Option	Description
--templates-repository <templatesRepository>	use the specified template repository

The template repository is a Git repository that contains the templates for the workspace. If no template repository is specified, provided templates from the CLI utility package are used.

For more details, see the Workspace structure and Template repository sections.

run command

The run command executes a specific use case in the workspace. Use cases can only be executed if the workspace has been initialized before and only within the top-level workspace directory.

The following options are available:

Option	Description
-u, --use-case [use-case]	the use case to run
--debug	enable debug mode; allows to run use cases from any location, assuming the config structure matches

For example, to run the check-requirements use case, execute:

npx @cycrilabs/ws-ctrl@latest run -u check-requirements

For more details, see the Use cases section.

sync command

The sync command syncs the templates. This command is used to update the templates in the workspace. Templates are copied from the CLI utility package itself and, if a template repository is used, from the template repository.

The command is executed via:

npx @cycrilabs/ws-ctrl@latest sync

For more details, see the Template repository section.

help command

The help command displays help information. It can be executed via:

npx @cycrilabs/ws-ctrl@latest help

Workspace structure

The workspace is structured as follows, assuming the workspace is located in ./platform and a template repository is used:

├── platform
│   ├── config
│   │   ├── development
│   |   |    ├── ...
│   │   ├── docker
│   |   |    ├── ...
│   │   ├── git-templates
│   |   |    ├── ...
│   │   ├── repositories
│   |   |    ├── ...
│   │   ├── secret-templates
│   |   |    ├── ...
│   │   ├── servers
│   |   |    ├── ...
│   │   ├── services-config
│   |   |    ├── ...
│   │   ├── use-cases
│   |   |    ├── ...
│   │   ├── cli.sh
│   ├── development
│   |   ├── ...
│   ├── adbb1a3d55.json
│   ├── cli.sh

The following table lists the directories and files in the workspace and their purpose:

Directory/File	Purpose
config	configuration files for the workspace; all files in here are the basis for the workspace
config/development	development files for the workspace; this directory is copied to the root directory and will contain all cloned repositories as well as IDE configuration files
config/docker	Docker configuration files for the workspace; the files are copied to the root directory and will be used to start services
config/git-templates	Git template repository; contains the cloned template repository, if configured; all files from it will be copied over recursively to the config directory
config/repositories	repository configuration files for the workspace; see Repository Configuration
config/secret-templates	secret templates for the workspace; see Secret Templates
config/servers	server configuration files for the workspace; see Server Configuration
config/services-config	service configuration files for the workspace; the file are automatically generated by the use case generate-service-configuration
config/use-cases	use case configuration files for the workspace; see Use Cases

Repository configuration

The repository configuration files are used to configure the repositories that can be cloned into the workspace. The configuration files are located in the config/repositories directory. All files are automatically available for prompts.

This JSON Schema defines the structure of the repository configuration.

Server configuration

The server configuration files are used to configure the servers that can be used to run the platform against. The configuration files are located in the config/servers directory. All files are automatically available for prompts.

This JSON Schema defines the structure of the server configuration.

Secret templates

The secret templates are used to configure the secrets that are necessary for all services running locally to connect to the remote Keycloak and other services. In addition, they container environment variables specific to each service. The secret templates are located in the config/secret-templates directory.

The secret templates are based on the Keycloak Configurator. For more details, check the documentation.

Use cases

Use cases are scripts based on JSON that can be executed in the workspace. They are located in the config/use-cases directory. Use cases can be executed via the run command.

Use cases provide a special syntax to define the steps that are executed in order to modify the configuration of a workspace, execute commands, or perform other tasks.

In the following, available use cases are listed and it is described how to create new use cases.

Available use cases

The following use cases are available out of the box and provided by the CLI utility package:

Use case name	Description
check-requirements	Checks the requirements for the workspace
clone-repositories	Clones selected repositories into the workspace
list-containers	Lists all running docker containers in the workspace
pull-repositories	Pulls the latest changes from the repositories in the workspace

The use cases are available by default and can be disabled via a template repository.

In addition, there are two use cases that are executed right after the commands init and sync are executed:

Use case name	Description
init	Executed directly after the init command did it's hard-coded work; only executed once on creation
sync	Executed directly after the sync command did it's hard-coded work; executed after every sync

These use cases have the state HIDDEN (see Creating use cases) and are not displayed in the list of available use cases. They can't be executed directly. However, via template repositories, it is possible to overwrite these use cases and easily extend what should be done after the init or sync command per project basis.

Creating use cases

Although there are a lot of use cases available, it is possible to create new use cases. Use cases are simply JSON files that expect some specific structure and keywords to be correctly interpreted by the CLI utility. Like this, it is possible to easily configure and execute pretty much everything on the workspace configuration files.

This JSON schema describes the structure of a use case. In the following, each keyword is described and potential end-2-end usage examples are provided.

A use case consists of the follow properties:

• id: a unique identifier for the use case; can be used to reference the use case or call it via the -u option
• name: the name of the use case; mainly used for display purposes
• description (optional): a description of the use case; mainly used for display purposes
• state (optional): the state of the use case; can be one of the following
  • ENABLED: the use case is generally usable; this is the default state
  • DISABLED: the use case is not usable at all
  • INITIAL: the use case is only usable with commands
  • HIDDEN: the use case is only callable by id, either by -u or via reference from another use case
• steps: an array of steps that are executed in order

Each step consists of the following properties:

• type: the type of the step; can be one of the following
  • FORMULA: a JavaScript formula that is executed in this step; executed via eval
  • COMMAND: execute a command; the command is executed via NodeJS child_process.execSync
  • EXECUTOR: a special step that is connected to some build-in executor; see Available executors
  • USE_CASE: call another use case
  • PROMPT: trigger a prompt for user input
• description (optional): a description of the step; mainly used for display purposes
• loop (optional): a loop that is executed for each element in the array; the loop is an object with the following properties
  • list: the array that is looped over
  • name: the variable that is used in the loop
• inputFile (optional): the input file for the step; the file is read and the content is passed to the step; can be used to write back content
• outputFile (optional): the output file for the command; the output of the step is written to the file
• abortIf (optional): a formula that is evaluated; if the formula is true, the use case is stopped
• skipIf (optional): a formula that is evaluated; if the formula is true, the step is skipped
• catchErrors (optional): a flag to catch errors; if set to true, errors are caught and the use case is not stopped; defaults to false
• useCase (optional): the use case that is called; only used for USE_CASE steps
• formula (optional): the formula that is executed; only used for FORMULA steps; a JavaScript formula that is executed in this step; executed via eval
• command (optional): the command that is executed; only used for COMMAND steps; the command is executed via NodeJS child_process.execSync
• executor (optional): the executor that is used; only used for EXECUTOR steps; see Available executors
• resultVariable (optional): the variable that is used to store the result of the step; the variable is available in the next steps; if not provided, results are stored as STEP_RESULT
• context (optional): the context that is used for the step; the context is an object that is passed to the step; can be used to store variables

Available executors

The following executors are available:

• custom-prompt: trigger a custom prompt on the CLI; the following context is expected:
  • name: the name of the variable that is used to store the result of the prompt
  • message: the message that is displayed to the user
  • type: the type of the prompt; can be one of the following (see prompts)
    • text: a simple text input
    • number: a simple number input
    • confirm: a yes or no questions
    • select: a single select prompt
    • multiselect: a multi select prompt
  • entities: the entities that are used for the select or multiselect prompt
  • entityKey: a key that is used to display entities from the context
• generate-service-configuration: generates .env files for services based on Keycloak Configurator; the following context is expected:
  • kcVersion: the version of the Keycloak Configurator to use
  • authServerUrl: the URL of the Keycloak server
  • authUser: the user to authenticate with
  • authPassword: the password to authenticate with
  • authTenant: the tenant to authenticate with
• change-env-var-value: changes the value of an environment variable in a given inputFile; the following context is expected:
  • name: the variable to change the value of
  • value: the new value of the variable

All steps are working with a context object that is empty initially. Whenever a use case starts, the context is filled with the following properties. Throughout execution, steps can add or modify properties in the context object.

• OS: the operating system where the CLI is executed on; represents the result of process.platform
• WORKSPACE_PATH: the absolute path to the workspace
• WORKING_DIR: the absolute path to the working directory, see Workspace structure
• SERVERS: the list of all servers, see Server configuration
• REPOSITORIES: the list of all repositories, see Repository configuration

The context can be accessed throughout different step types within a formula, command, or executor via the properties in context. Whenever input is evaluated based on the Javascript eval function, the context is available as a flat list of properties. For example, assume the context looks as follows:

const context = {
  server: 'local',
};

A formula can access the server property via:

server === 'local' : 'yes' : 'no'

Template repository

The template repository is a Git repository. The structure of the repository is the same as the config directory in the workspace, described in the Workspace structure section.

If a template repository is used, the repository is cloned into the workspace, to the config/git-templates directory. The files from the template repository are then copied over recursively to the config directory, overwriting all existing files.

This allows to provide different servers, use cases, and other configurations that are specific to a project. A common use case could be to provide additional servers or repositories that are specific to a project. Furthermore, often the docker-compose-cxc.yml is overwritten to use cxc specific services.

CLI Development

To develop the CLI utility, it is necessary to install the dependencies:

npm install

Afterward, the tool can be executed via:

npm run dev

Commands can be executed as described above. They must be simply appended to the npm run dev command:

npm run dev -- help init

Testing

The tool can be tested end-2-end by executing the following command:

npm link

This adds a symlink to the global NPM packages, so that the tool can be executed from any directory. The tool can be executed via:

ws-ctrl help

After testing, the symlink can be removed via:

npm uninstall ws-ctrl -g

Use cases

It is possible to test use cases directly in the workspace. To do so, the use case should be placed in the assets/templates/config/use-cases directory. The use case can then be executed via:

npm run dev -- run ./assets/templates -u <use-case-name> --debug