Convict expands on the standard pattern of configuring node.js applications in a way that is more robust and accessible to collaborators, who may have less interest in digging through imperative code in order to inspect or modify settings. By introducing a configuration schema, convict gives project collaborators more context on each setting and enables validation and early failures for when configuration goes wrong.
- Loading and merging: configurations are loaded from disk or inline and merged
- Nested structure: keys and values can be organized in a tree structure
- Environmental variables: values can be derived from environmental variables
- Command-line arguments: values can also be derived from command-line arguments
- Validation: configurations are validated against your schema (presence checking, type checking, custom checking), generating an error report with all errors that are found
- Comments allowed: JSON files are loaded with the
cjson
module, so comments are welcome
npm install convict
An example config.js
:
var convict = require('convict');
// Define a schema
var conf = convict({
env: {
doc: "The applicaton environment.",
format: ["production", "development", "test"],
default: "development",
env: "NODE_ENV"
},
ip: {
doc: "The IP address to bind.",
format: "ipaddress",
default: "127.0.0.1",
env: "IP_ADDRESS",
},
port: {
doc: "The port to bind.",
format: "port",
default: 0,
env: "PORT"
}
});
// Load environment dependent configuration
var env = conf.get('env');
conf.loadFile('./config/' + env + '.json');
// Perform validation
conf.validate({strict: true});
module.exports = conf;
var http = require('http');
var conf = require('./config.js');
var server = http.createServer(function (req, res) {
res.writeHead(200, {'Content-Type': 'text/plain'});
res.end('Hello World\n');
});
// Consume
server.listen(conf.get('port'), conf.get('ip'), function(x) {
var addy = server.address();
console.log('running on http://' + addy.address + ":" + addy.port);
});
A configuration module could look like this:
config.js:
var config = module.exports = convict({
env: {
doc: "The application environment.",
format: ["production", "development", "test"],
default: "development",
env: "NODE_ENV",
arg: "node-env",
}
});
config.loadFile(['./prod.json', './config.json']);
Each setting in the schema has four possible properties, each aiding in convict's goal of being more robust and collaborator friendly.
- Type information: the
format
property specifies either a built-in convict format (ipaddress
,port
,int
, etc.), or it can be a function to check a custom format. During validation, if a format check fails it will be added to the error report. - Default values: Every setting must have a default value.
- Environmental variables: If the variable specified by
env
has a value, it will overwrite the setting's default value. - Command-line arguments: If the command-line argument specified by
arg
is supplied, it will overwrite the setting's default value or the value derived fromenv
. - Documentation: The
doc
property is pretty self-explanatory. The nice part about having it in the schema rather than as a comment is that we can callconf.toSchemaString()
and have it displayed in the output.
Nested configuration settings are also supported:
var config = convict({
server: {
ip: {
doc: "IP address to bind",
format: 'ipaddress',
default: '0.0.0.0'
},
port: {
doc: "port to bind",
format: 'port',
default: 8080
}
},
database: {
host: {
doc: "Database host name/IP",
format: String,
default: 'testing'
},
name: {
doc: "Database name",
format: String,
default: 'users'
}
}
});
Note: Search for the word "nested" throughout this documentation to find out more about nested configuration settings.
In order to help detect misconfigurations, convict allows you to define a format for each setting. By default, convict checks if the value of the property has the same type (according to Object.prototype.toString.call
) as the default value specified in the schema. You can define a custom format checking function in the schema by setting the format
property.
convict provides several predefined formats for validation that you can use (using node-validator and moment.js). Most of them are self-explanatory:
*
- any value is validint
port
url
email
ipaddress
- IPv4 and IPv6 addressesduration
- milliseconds or a human readable string (e.g. 3000, "5 days")timestamp
- Unix timestamps or date strings recognized by moment.jsnat
- positive integer (natural number)
If format
is set to one of the built-in JavaScript constructors, Object
, Array
, String
, Number
, or Boolean
, validation will use Object.prototype.toString.call to check that the setting is the proper type.
You can specify a custom format checking method on a property basis.
For example:
var conf = convict({
key: {
doc: "API key",
format: function check (val) {
if (!/^[a-fA-F0-9]{64}$/.test(val)) {
throw new Error('must be a 64 character hex key')
}
},
default: '3cec609c9bc601c047af917a544645c50caf8cd606806b4e0a23312441014deb'
}
});
Or, you can use convict.addFormat()
to register a custom format checking
method that can be reused for many different properties:
convict.addFormat({
name: 'float-percent',
validate: function(val) {
if (val !== 0 && (!val || val > 1 || val < 0)) {
throw new Error('must be a float between 0 and 1, inclusive');
}
},
coerce: function(val) {
return parseFloat(val, 10);
}
});
var conf = convict({
space_used: {
format: 'float-percent',
default: 0.5
},
success_rate: {
format: 'float-percent',
default: 60.0
}
});
The coerce
function is optional.
Convict will automatically coerce environmental variables from strings to their proper types when importing them. For instance, values with the format int
, nat
, port
, or Number
will become numbers after a straight forward parseInt
or parseFloat
. duration
and timestamp
are also parse and converted into numbers, though they utilize moment.js for date parsing.
The second argument to coerce
is the config
object, this can be used to implement placeholders and other advanced functionality:
convict.addFormat({
name: "placeholder",
validate: function(val) {
/* validate proper path here */
},
coerce: function(val, config) {
return val.replace(/\$\{([\w\.]+)}/g, function(v,m) { return config.get(m); });
}
});
var conf = convict({
env: {
format: ['production', 'development'],
default: 'development',
env: 'NODE_ENV',
doc: 'The environment that we\'re running in.'
},
configPath: {
format: 'placeholder',
default: '/path/to/config',
doc: 'Path to configuration files.'
},
config: {
format: 'placeholder',
default: '${configPath}/${env}.json',
doc: 'Path to configuration file.'
}
});
conf.get('config'); /* "/path/to/config/development.json" */
When merging configuration values from different sources, Convict follows precedence rules. The order, from lowest to highest, is:
- Default value
- File (
config.loadFile()
) - Environment variables
- Command line arguments
- Set and load calls (
config.set()
andconfig.load()
)
convict()
takes a schema object and returns a convict configuration object. The configuration object has an API for getting and setting values, described below.
Adds a new custom format.
Adds new custom formats.
Returns the current value of the name
property. name
can use dot notation to reference nested values. E.g.:
config.get('database.host');
// or
config.get('database').host;
Returns the default value of the name
property. name
can use dot notation to reference nested values. E.g.:
config.default('server.port');
Returns true
if the property name
is defined, or false
otherwise. E.g.:
if (config.has('some.property')) {
// Do something
}
Sets the value of name
to value. name
can use dot notation to reference nested values, e.g. "database.port"
. If objects in the chain don't yet exist, they will be initialized to empty objects. E.g.:
config.set('property.that.may.not.exist.yet', 'some value');
config.get('property.that.may.not.exist.yet');
// Returns "some value"
This will load and merge a JavaScript object into config
. E.g.:
config.load({
"env": "test",
"ip": "127.0.0.1",
"port": 80
});
This will load and merge one or multiple JSON configuration files into config
. JSON files are loaded using cjson
, so they can contain comments. E.g.:
conf.loadFile('./config/' + conf.get('env') + '.json');
Or, loading multiple files at once:
// CONFIG_FILES=/path/to/production.json,/path/to/secrets.json,/path/to/sitespecific.json
conf.loadFile(process.env.CONFIG_FILES.split(','));
Validates config
against the schema used to initialize it. All errors are
collected and thrown at once.
Options: At the moment strict
is the only available option.
If the strict
option is passed (that is {strict: true}
is passed), any
properties specified in config files that are not declared in the schema will
result in errors. This is to ensure that the schema and the config files are in
sync. By default the strict mode is set to false.
Exports all the properties (that is the keys and their current values) as JSON.
Exports all the properties (that is the keys and their current values) as a JSON string.
Exports the schema as JSON.
Exports the schema as a JSON string.
The philosophy was to have production values be the default values. Usually you only want to change defaults for deploy or instance (in aws speak) specific tweaks. However, you can set a default value to null
and if your format doesn't accept null
it will throw an error.
Thanks to browserify, convict
can be used for web applications too. To do so,
- Use
brfs
to ensure thefs.loadFileSync
schema-loading calls are inlined at build time rather than resolved at runtime (in Gulp, add.transform(brfs)
to your browserify pipe). - To support "loading configuration from a
http://foo.bar/some.json
URL", build a thin wrapper around convict using your favorite http package (e.g.superagent
). Typically, in the success callback, call convict'sload()
on the body of the response.
Read the Contributing doc.