Declarative object validation
Install
$ npm install @dprovodnikov/validationUsage example
Here's a typical login form data structure:
const form = {
login: '',
password: '',
repeatPassword: '',
};Rules declaration
Let's say we have a set of constraints to apply to the fields of this form:
- Login cannot be empty
- Password should contain at least one uppercase letter
- Password should be at least 7 chars long
- Password should match repeatPassword
These are common rules pretty much every login form uses. Now we need to declare these rules.
import { notEmpty, hasUpperLetter, longerThen, matches } from '@dprovodnikov/validation';
const rules = {
login: [notEmpty('Login cannot be empty')],
password: [
hasUpperLetter('Password should contain at least one uppercase letter'),
longerThen(7, 'Password should be at least 7 chars long'),
],
repeatPassword: [
matches('password', 'Password should match repeatPassword'),
],
};The library provides a couple of basic rules that are fairly common:
| Rule | Description |
|---|---|
| notEmpty | Makes sure the given value is not empty |
| hasUpperLetter | Makes sure the given value contains an uppercase letter |
| longerThen | Makes sure the given value has a length of more than specified |
| matches | Makes sure the given value matches the specified field of the form |
| emailFormat | Makes sure the given value matches the regexp (integrated in the lib) |
| urlFormat | Makes sure the given value matches the regexp (integrated in the lib) |
| shorterThan | Makes sure the given value is shorter than the specified value |
If you need any other rule you can easily implement one using simple api. Here's what a rule looks like:
const rule = message => (value, form) => {
// You need to return a string.
// This string represents the validation error message.
// If the output is empty - the value satisfies the rule and therefore is considered valid.
let output = '';
// here you have access to the value this rule was assigned to.
// the entire form object is also accessible from here for more complex and flexible validation logic
return output;
};Here's the implementation of the matches rule that is provided by the library:
export const matches = (comparisonField, message) => (value, fields) => {
if (value !== fields[comparisonField]) {
return message;
}
return '';
};As simple as that.
Rules application
In order to create a validator you need to apply the rules you declared earlier.
import { applyRules } from '@dprovodnikov/validation';
const rules = {
field: [notEmpty('Field should not be empty')],
};
const validate = applyRules(rules);
const form = {
field: '',
};
validate('all', form); // { field: 'Field should not be empty' }
validate('field', form); // 'Field should not be empty'The output of the validator is a structure where keys are field names and values are error messages. If a field has multiple rules applied to it and more than one of them fail - the output will contain the first one that failed.
There is a tool to easily check if the report contains any error messages:
import { hasError } from '@dprovodnikov/validation';
const report = validate(form);
if (hasError(report)) {
...
}There's also a method to extract the first error in the output;
import { hasError, getFirstError } from '@dprovodnikov/validation';
const report = validate(form);
if (hasError(report)) {
const errorMessage = getFirstError(report);
...
}The list of methods:
| Method | Input | Output | Description |
|---|---|---|---|
| applyRules | rules | validator function | Takes in rules and returns a validator |
| hasError | validation report | boolean | Takes in a validation report and returns wheather it contains an error on not |
| getFirstError | validation report | string | Takes in a validation report and returns the first error |
Build
Make sure tests pass
$ npm run testInstall dependencies
$ npm installBuild from sources
$ npm run buildThe output will appear in the dist folder in the project's root.
License
MIT