Structural-matching for JavaScript.
- validating user input / config files
- validating inbound HTTP request payloads
- writing expressive unit tests
- Getting started
- Syntactic sugar
- A more complex example
- Optional values
- Defining custom matchers
- Asserting on matchers
- Custom constraints
- A note on performance
npm install strummer
var s = require('strummer');
var person = s({
name: 'string',
age: 'number',
address: {
city: 'string',
postcode: 'number'
},
nicknames: ['string']
});
console.log(person.match(bob));
// [
// { path: 'name', value: null, message: 'should be a string' }
// { path: 'address.postcode', value: 'NY', message: 'should be a number' }
// ]The example above is actually syntactic sugar for:
var person = new s.object({
name: new s.string(),
age: new s.number(),
address: new s.object({
city: new s.string(),
postcode: new s.number()
}),
nicknames: new s.array({of: new s.string()})
});This means all matchers are actually instances of s.Matcher,
and can potentially take extra parameters.
new s.number({min:1, max:100})Built-in matchers include(all classes)
s.array({min, max, of, description})s.boolean({parse, description})s.duration({min, max, description})s.enum({name, values, verbose, description})s.func({arity})s.hashmap({keys, values})s.integer({min, max, description})s.ip({version: 4, description})s.isoDate({time, description})s.number({min, max, parse, description})s.object(fields, {description})s.objectWithOnly(fields, {description})s.regex(reg, {description})s.string({min, max, description})s.url({description})s.uuid({version, description})s.value(primitive, {description})s.email({domain, description})s.oneOf([matcher], {description})
They all come with several usage examples. Matchers usually support both simple / complex usages, with nice syntactic sugar.
Here's an example that mixes nested objects, arrays, and matches on different types with extra options.
var person = new s.object({
id: new s.uuid({version: 4}),
name: 'string',
age: new s.number({min: 1, max: 100}),
address: {
city: 'string',
postcode: 'number'
},
nicknames: [{max: 3, of: 'string'}],
phones: [{of: {
type: new s.enum({values: ['MOBILE', 'HOME']}),
number: 'number'
}}]
});You can of course extract matchers to reuse them, or to make the hierarchy more legible.
var age = new s.number({min: 1, max: 100})
var address = new s.object({
city: 'string',
postcode: 'number'
});
var person = new s.object({
name: 'string',
age: age,
home: address
});By default, all matchers expect the value to exist. In other words every field is required in your schema definition.
You can make a field optional by using the special {optional: true} argument.,
new s.number({optional: true, min: 1})To define a customer matcher, simply inherit the s.Matcher prototype
and implement the _match function.
var s = require('strummer');
function MyMatcher(opts) {
s.Matcher.call(this, opts);
}
util.inherits(MyMatcher, s.Matcher);
MyMatcher.prototype._match = function(path, value) {
// if this is a leaf matcher, we only care about the current value
return null;
return 'should be a string starting with ABC';
// if this matcher has children, we need to return an array of errors;
return [];
return [
{ path: path + '[0]', value: value[0], message: 'should be > 10' }
{ path: path + '[1]', value: value[1], message: 'should be > 20' }
]
};Or you can use the helper function to create it:
var MyMatcher = s.createMatcher({
initialize: function() {
// initialize here
// you can use "this" to store local data
},
match: function(path, value) {
// validate here
// you can also use "this"
}
});You can use these matchers like any of the built-in ones.
new s.object({
name: 'string',
id: new MyMatcher({max: 3})
})Matchers always return the following structure:
[
{ path: 'person.name', value: null, message: 'should be a string' }
]In some cases, you might just want to throw an error - for example in the context of a unit test.
Strummer provides the s.assert function for that purpose:
s.assert(name, 'string');
// name should be a string (but was null)
s.assert(nicknames, ['string']);
// name[2] should be a string (but was 123)
// name[3] should be a string (but was 456)
s.assert(person, {
name: 'string',
age: new s.number({max: 200})
});
// person.age should be a number <= 200 (but was 250)Custom constraints can be applied by passing a function as the second argument when creating the schema. This function will be run on match and you are able to return an array of errors.
Currently only objectWithOnly is supported.
An example use case is related optional fields
// AND relationship between two optional fields
var constraintFunc = function (path, value) {
if (value.street_number && !value.post_code) {
return [{
path: path,
value: value,
error: 'post_code is requried with a street_number'
}]
}
return []
}
var schema = new objectWithOnly({
email_address: new string(),
street_number: new number({optional: true}),
post_code: new number({optional: true}),
}, {
constraints: constraintFunc
});
var value = {
email_address: '[email protected]',
street_number: 12,
}
const errors = schema.match(value)
// will error with post_code is requried with a street_number
## JSON Schema Support
Strummer can generate some simple JSON Schema from strummer definition.
```js
var schema = s({
foo: 'string',
bar: s.string({ optional: true, description: 'Lorem Ipsum' }),
num: s.number({ max: 100, min: 0 })
});
console.log(schema.toJSONSchema());which will shows log like this:
{
type: 'object',
required: ['foo', 'num'],
properties: {
foo: {
type: 'string'
},
bar: {
type: 'string',
description: 'Lorem Ipsum'
},
num: {
type: 'number',
maximum: 100,
minimum: 0
}
}
}When you trying to create your own matcher which supports jsonSchema, then you needs to impement
the toJSONSchema option in the createMatcher, if toJSONSchema is not defined, when you call
matcher.toJSONSchema() it will return nothing.
The 2 main rules for performance are:
-
If you need to validate many objects of the same kind, you should declare matchers upfront and reuse them.
-
All syntactic sugar is processed at creation time. This means shorthand notations don't cause any performance overhead compared to their canonical equivalents.
Of course, actual performance depends on the complexity of your matchers / objects. If you're interested in figures, some stats are printed as part of the unit test suite:
new s.object({
id: new s.uuid({version: 4}),
name: 'string',
age: new s.number({optional: true, min: 1, max: 100}),
addresses: new s.array({of: {
type: 'string',
city: 'string',
postcode: 'number'
}}),
nicknames: [{max: 3, of: 'string'}],
phones: [{of: {
type: new s.enum({values: ['MOBILE', 'HOME']}),
number: /^[0-9]{10}$/
}}]
})
// ┌───────────────────────┬─────────────────┐
// │ Number of validations │ Total time (ms) │
// ├───────────────────────┼─────────────────┤
// │ 10,000 │ 85 │
// └───────────────────────┴─────────────────┘