DotEnv with schema, validation, expiring keys and multi-input support.

Features:

• All standard DotEnv parser / reader
• Mongo like Schema support
• Entirely synchronous - config is available immediately, no waiting on promises or fixing up race conditions
• Easy config key rewriting via .map(fn) / .camelCase() / .startCase() etc. to make config compatible with other systems
• Support for destructing config - wipe values like passwords or API-keys after a set interval

import dotenv from '@momsfriendlydevco/dotenv';

let config = dotenv
    .parse(['.env.example', '.env']) // Import from one or more files in order of precidence
    .importEnv({prefix: 'MY_APP_'}) // Import from process.env
    .schema({ // Apply a schema
        port: Number, // Simple schemas (`required` is implied for these)
        host: {required: true, type: String}, // ...or be more explicit
        password: {
            type: 'string', // Custom support also for your own types
            cast: v => v.toLowerCase(), // Case incoming values
            destruct: '10s', // Variables not allowed to be read after this time from start (timestring)
        },
    })
    .value() //= {host: String, port: Number, password: String}

Returns an instance of the DotEnv class.

Read files from path (or relative to current directory) in presidence order. Source can be a path (string with no '='), paths (an array of strings) or raw blob data (string || buffer).

Later files with keys will overwrite earlier ones.

Options are:

Returns the chainable DotEnv instance.

Import environment variables (from process.env by default).

Returns the chainable DotEnv instance.

Valid options are:

Apply a schema to the existing internal config state.

A schema is a object mapping each key to a FieldSchema.

Returns the chainable DotEnv instance.

FieldSchemas are made up of any of the following properties. If a simple string or built-in type is provided instead it is assumed as the type subkey.

Schema options can contain:

The required boolean has the following logic:

1. If we have a value that isn't undefined or an empty string stop here
2. If any default / defaultRaw is specified - apply those (see next section)
3. If no value is still set - return undefined

Defaults are applied with the following logic:

1. If we have a value that isn't undefined or an empty string stop here
2. If defaultRaw exists - use that
3. If default exists and is a function - evaluate it first, if not contiue with the default value
4. If the value from step 3 is NOT a string - assume the return value should be used as is, stop any further checks (e.g. casting, validation)
5. If the value from step 3 is a string - assume this is the same as what would be set in the .env file and continue on to cast + validate etc.

Apply a schema to all config keys matching a glob, array of globs or a RegExp.

Export the current config to a string.

Options are:

Utility function to call .export() and dump the contents into a file.

Indicates that key mutation functions (see below) should operate on nested objects. Set to any falsy value to revert to only applying mutations to top level keys only. Returns the DotEnv instance.

Apply camelCase, startCase or envCase (with options) to all keys, returning the DotEnv instance.

Mutate the config keys by applying a common string mutator, returning the DotEnv instance.

Apply a replacement to all config keys, returning the DotEnv instance.

Remove all config keys that either dont have the string prefix or don't match a given RegEx, returning the DotEnv instance.

Apply a replacement to all config keys removing a given String prefix / RegExp match, returning the DotEnv instance.

Apply both a filter + trim to a config keys - removing all config that doesnt match the string prefix (or RegEx) whilst also removing the given prefix (or RegExp). Returns the DotEnv instance.

Applies a JavaScript string template (via @MomsFriendlyDevCo/Template) to all values with the given context. If no context is provided the current state is used. Returns the DotEnv instance.

Run a function on all state config keys, potencially mutating the key / value. Returns the DotEnv instance afterwards.

• If the function returns a tuple array its assumed to mutate the key+val of the input config key+val combo
• If the function returns a object, that object return (all keys) replace the state for that config key
• If the function returns boolean false the key is removed completely
• If the funciton returns boolean true OR undefined, no action or mutation is taken

Run an arbitrary function passing in this DotEnv instance as the first argument and context. Returns the chainable DotEnv instance.

new DotEnv()
    .parse(...)
    .tap(dotEnv => console.log('Raw config:', dotEnv.value())
    .toTree(/_/)
    .tap(dotEnv => console.log('Config as a tree:', dotEnv.value())
    .value()

Note that if you intend to copy the state inside tap() it is advisable to use .value({clone: true}) as functions such as .toTree() mutate sub-keys which can change state.

Like DotEnv.tap(fn) only the return value is used as the new state. Returns the chainable DotEnv instance.

Transform a flat key/val config item a hierarchical object-of-objects based on rules. Returns the chainable DotEnv instance.

let result = new DotEnv()
    .parse([
        'FOO_BAR_FOO=Foo!',
        'FOO_BAR_BAR=123',
        'FOO_BAR_BAZ=true',
        'BAR_BAR_FOO=Foo2!',
        'BAR_BAR_BAR=456',
        'BAR_BAR_BAZ=false',
    ].join('\n'))
    .toTree(/_/)
    .value()

/**
// Will return
{
    FOO: {
        BAR: {
            FOO: 'Foo!',
            BAR: '123',
            BAZ: 'true',
        },
    },
    BAR: {
        BAR: {
            FOO: 'Foo2!',
            BAR: '456',
            BAZ: 'false',
        },
    },
}

Options are:

Return the final, computed config object.

Options are:

The following is a list of built in types which provide a handy shorthand for various functionality.