Project Setup

It's considered good practice to define any sensitive information (e.g. passwords) in a .env environment file along with any other information that is specific to the machine it's running on or the deployment environment (e.g. development, staging, production, etc).

The setup() function can help you automate this process. It allows you to create a setup script that users can run when they first check out a project to a machine, or when configuration options change. It uses a configuration file to define the command line arguments that the script will accept, and will prompt the user to enter any that haven't been specified. It will then generate the .env file.

Setup Script

You should define a setup script for your project. A good place for this is bin/setup.js. It should look something like this:

#!/usr/bin/env node
import { bin, setup } from '@abw/badger'

await setup({
  rootDir: bin().up(),
  writeEnv: true,
});

The rootDir option specifies where the root directory for your project is. It's usually not necessary as it defaults to the current working directory. If you run the script from the project root then it will work as expected. But it doesn't hurt to be explicit. Here we're using the bin() function from badger-filesystem which returns a directory object for the directory in which the script is located (appropriately called bin).

The up() method returns the directory one level up. If your script is located in a lower level directory (e.g. bin/project/setup.js) then you should go up another level (e.g. bin().up(2)). If your script is already in the top-level directory then bin() will suffice.

The benefit of being explicit is that you can then run this script from any directory and it will work correctly.

The writeEnv option tells it to write a .env file when it's done. The envFile option can be used to specify the name for the file, relative to the rootDir directory. The default is .env.

You'll need to define a setup configuration file to make it work. We'll go through that in the next section. But first here's a sneak preview of what happens when you run the script with an example configuration. This example is in examples/setup if you want to play along at home.

bash

$ bin/setup.js

Project Configuration
---------------------

Please enter the following questions to setup the project.

✔ Where is the project root directory on this machine? … /home/abw/my-project
✔ What kind of deployment is this? › development

Database Configuration
----------------------

Enter the connection details for the database.

✔ Enter the name of the database … badger
✔ Enter the name of the database host … localhost
✔ Enter the port on which the database is running … 3306
✔ Enter the username for connecting to the database … badger
✔ Enter the password for connecting to the database … *********

Programs
--------

We need to know the location of some programs on this machine.

✔ Enter the full path to the gzip program … /usr/bin/gzip
✔ Enter the full path to the gzcat (or zcat) program … /usr/bin/gzcat

👍  All configuration options have been set

The script will prompt you to confirm or enter values for each of the options in your configuration file. It will then write a nicely formatted .env file and also a .env.yaml file containing the raw data.

Here's what the .env file might look like:

bash

#=============================================================================
# WARNING: This file is generated automatically when the undefined
# script is run.  Any changes made here may be lost.
#
# Generated: 2023-03-10 11:02:59
#=============================================================================

#-----------------------------------------------------------------------------
# Project Configuration
#-----------------------------------------------------------------------------

# Project directory
MYAPP_ROOT_DIR=/home/abw/my-project

# Deployment type
DEPLOYMENT=development


#-----------------------------------------------------------------------------
# Database Configuration
#-----------------------------------------------------------------------------

# Database name
DB_NAME=badger

# Database host
DB_HOST=localhost

# Database port
DB_PORT=3306

# Database user
DB_USER=badger

# Database password
DB_PASS=TOP_SECRET


#-----------------------------------------------------------------------------
# Programs
#-----------------------------------------------------------------------------

# Path to gzip
GZIP=/usr/bin/gzip

# Path to gzcat (or zcat)
GZCAT=/usr/bin/gzcat

The -v or --verbose options can be added to get some extra messages generated. The -h or --help options can be used to view the help for the script:

bash

$ bin/setup.js -h
Usage: bin/setup.js [options]

Project setup script.

Options:
  -y, --yes              Accept default answers
  -v, --verbose          Verbose output
  -q, --quiet            Quiet output
  --root <dir>           Project directory (default: "/home/abw/my-project")
  --deployment <type>    Deployment type (default: "development")
  --db_name <dbname>     Database name (default: "badger")
  --db_host <hostname>   Database host (default: "localhost")
  --db_port <port>       Database port (default: 3306)
  --db_user <username>   Database user (default: "badger")
  --db_pass <password>   Database password
  --gzip <path>          Path to gzip (default: "/usr/bin/gzip")
  --gzcat <path>         Path to gzcat (or zcat) (default: "/usr/bin/gzcat")
  -h, --help             display help for command

In this example the options from --deployment down to --gzcat are specific to the configuration file and everything else comes as standard.

If you need to change an option you can go through the questions over again (it will remember the answers you previously gave by looking at the generated .env and/or .env.yaml files), or you can use a command line option to set the relevant value along with -y or --yes to automatically accept all the other value. The -q or --quiet option can be used to supress any other output.

For example, to update the database password:

bash

$ bin/setup.js --db_pass NEWPASSWORD -y -q

The .env file will then be re-generated with the new value for DB_PASS.

Configuration file

You need to define a configuration file to tell the script what data you need to capture. It can be in any one of the standard locations, relative to your project root:

config/setup.yaml
config/setup.json
setup.yaml
setup.json
.setup.yaml
.setup.json

You can use the configFile option to specify a different location if you don't like the look of any of those.

Here's the configuration file used in the earlier example.

yaml

# standard options
yes:            true
verbose:        true
quiet:          true

# custom project options
options:
  #---------------------------------------------------------------------------
  # Project
  #---------------------------------------------------------------------------
  -
    title:      Project Configuration
    info:       Please answer the following questions to setup the project.
  -
    name:       root
    about:      Project directory
    prompt:     Where is the project root directory on this machine?
    pattern:    <dir>
    required:   true
    envvar:     MYAPP_ROOT_DIR
  -
    name:       deployment
    about:      Deployment type
    prompt:     What kind of deployment is this?
    type:       select
    pattern:    <type>
    required:   true
    default:    development
    choices:
      -
        title:  development
        value:  development
      -
        title:  staging
        value:  staging
      -
        title:  production
        value:  production

  #---------------------------------------------------------------------------
  # Database
  #---------------------------------------------------------------------------
  -
    title:      Database Configuration
    info:       Enter the connection details for the database.
  -
    name:       db_name
    about:      Database name
    prompt:     Enter the name of the database
    pattern:    <dbname>
    required:   true
    default:    badger
  -
    name:       db_host
    about:      Database host
    prompt:     Enter the name of the database host
    pattern:    <hostname>
    required:   true
    default:    localhost
  -
    name:       db_port
    about:      Database port
    prompt:     Enter the port on which the database is running
    pattern:    <port>
    required:   true
    default:    3306
  -
    name:       db_user
    about:      Database user
    prompt:     Enter the username for connecting to the database
    pattern:    <username>
    required:   true
    default:    badger
  -
    name:       db_pass
    about:      Database password
    prompt:     Enter the password for connecting to the database
    pattern:    <password>
    type:       password
    required:   true

  #---------------------------------------------------------------------------
  # Programs
  #---------------------------------------------------------------------------
  -
    title:      Programs
    info:       We need to know the location of some programs on this machine.
  -
    name:       gzip
    about:      Path to gzip
    program:    gzip
    prompt:     Enter the full path to the gzip program
    mandatory:  true
  -
    name:       gzcat
    about:      Path to gzcat (or zcat)
    program:    gzcat zcat
    prompt:     Enter the full path to the gzcat (or zcat) program
    mandatory:  true

The first section contains the standard options that enable the -y / --yes, -v / --verbose and -q / --quiet options.

yaml

# standard options
yes:            true
verbose:        true
quiet:          true

Following that we have the options section. Each option must contain a name and can have various other parameters.

yaml

options:
  -
    name:       root
    envvar:     MYAPP_ROOT_DIR
    about:      Project directory
    prompt:     Where is the project root directory on this machine?
    required:   true
    pattern:    <dir>
  # ...etc.

The name is used as the command line option for the script, e.g. --root. If an envvar is defined then this will be the name used for the environment variable in the .env file (MYAPP_ROOT_DIR in this example). Otherwise it defaults to an upper can version of the name (e.g. ROOT).

If you don't want an answer to be saved either in the .env file or in the cached answers file then set the save property to false.

yaml

    save:       false

The about property is displayed in the help text. It will also be added as a comment to the .env file. The prompt is the prompt used to request user input. If this is not defined then the user will not be prompted to enter a value for it. The pattern is displayed against the option in the help text, e.g.

bash

   --root <dir>           Project directory (default: ...)

If you don't specify it then it defaults to <text> (or <path> for program options)

The required flag indicates that a value must be provided.

The default input type is text, but you can explicitly set it to password where appropriate:

yaml

  -
    name:       db_pass
    about:      Database password
    prompt:     Enter the password for connecting to the database
    pattern:    <password>
    type:       password
    required:   true

The select type can be be used to provide a list of values to select from:

yaml

  -
    name:       deployment
    about:      Deployment type
    prompt:     What kind of deployment is this?
    type:       select
    pattern:    <type>
    required:   true
    default:    development
    choices:
      -
        title:  development
        value:  development
      -
        title:  staging
        value:  staging
      -
        title:  production
        value:  production

The program option provides some extra magic. It will look for the named program in your PATH and set the default value to be the first location it finds.

yaml

  -
    name:       gzip
    about:      Path to gzip
    program:    gzip
    prompt:     Enter the full path to the gzip program
    mandatory:  true

If you don't set the pattern for a program option then it defaults to <path>. If there are different names for acceptable programs then you can specify multiple values to the program option. For example, gzcat and zcat both do the same thing.

yaml

    program:    gzcat zcat

Lots of questions can be a bit overwhelming to a user. You can break them up into sections to relieve the monotony by defining a title option.

yaml

  -
    title:      Programs
    info:       We need to know the location of some programs on this machine.

Setup Options

Here's a summary of the options that you can pass to the setup() function.

Option	Description	Default
rootDir	Your project root directory	Current working directory
description	A description for the help text	Project setup script.
version	The version number for the help text
configFile	The path to your setup configuration file, relative to `rootDir`	`config/setup.yaml`, `config/setup.json`, `setup.yaml`, `setup.json`, `.setup.yaml` or `.setup.json`
values	Default answers for values
preprocess	A function to pre-process the setup configuration file after loading	`false`
envFile	The name of your environment file, relative to `rootDir`	`.env`
writeEnv	Should it automatically write the `envFile` file?	`false`
envExtra	Additional file to append to `envFile`, e.g. `.env.extra`
envSections	Add section block comments to `envFile`	`true`
envComments	Add item comments to `envFile`	`true`
envPrefix	Add a prefix to all environment variable written to the `envFile`
compact	Don't add blank spacing lines in `envFile`	`false`
dataFile	JSON or YAML file for caching answers	`.env.yaml`
writeData	Should it automatically write the `dataFile` file?	`true`
cancelled	Text to display when script is cancelled (by hitting escape or Ctrl-C)	`Setup cancelled`
allDone	Text to display when all questions have been answered	`All configuration options have been set`
warning	Warning text to add to top of `envFile`	(WARNING text block)

Project Setup ​

Setup Script ​

Configuration file ​

Setup Options ​

Project Setup

Setup Script

Configuration file

Setup Options