LIVR Validator
LIVR.Validator - Lightweight JavaScript validator supporting Language Independent Validation Rules Specification (LIVR).
SYNOPSIS
There are 2 implementations:
- "Validator" supports only synchronous rules. All built-in rules are synchronous.
- "AsyncValidator" (experimental) supports both synchronous and asynchronous rules.
Common usage:
import LIVR from 'livr';
LIVR.Validator.defaultAutoTrim(true);
const validator = new LIVR.Validator({
name: 'required',
email: ['required', 'email'],
gender: { one_of: ['male', 'female'] },
phone: { max_length: 10 },
password: ['required', { min_length: 10 }],
password2: { equal_to_field: 'password' },
});
const validData = validator.validate(userData);
if (validData) {
saveUser(validData);
} else {
console.log('errors', validator.getErrors());
}
All standard rules are supported in camel case as well (Validator.registerDefaultRules always does autocamelization):
Camel case names are closer to JS naming conventions but underscore rule names are more compatible with LIVR spec.
import LIVR from 'livr';
LIVR.Validator.defaultAutoTrim(true);
const validator = new LIVR.Validator({
name: 'required',
email: ['required', 'email'],
gender: { oneOf: ['male', 'female'] },
phone: { maxLength: 10 },
password: ['required', { minLength: 10 }],
password2: { equalToField: 'password' },
});
const validData = validator.validate(userData);
if (validData) {
saveUser(validData);
} else {
console.log('errors', validator.getErrors());
}
Common usage of async version:
import LIVR from 'livr/async';
LIVR.AsyncValidator.defaultAutoTrim(true);
const validator = new LIVR.AsyncValidator({
name: 'required',
email: ['required', 'email'],
gender: { one_of: ['male', 'female'] },
phone: { max_length: 10 },
password: ['required', { min_length: 10 }],
password2: { equal_to_field: 'password' },
});
try {
const validData = await validator.validate(userData);
saveUser(validData);
} catch (errors) {
console.log('errors', errors);
}
You can use modifiers separately or can combine them with validation:
const validator = new LIVR.Validator({
email: ['required', 'trim', 'email', 'to_lc'],
});
Feel free to register your own rules:
You can use aliases(preferable, syntax covered by the specification) for a lot of cases:
const validator = new LIVR.Validator({
password: ['required', 'strong_password'],
});
validator.registerAliasedRule({
name: 'strong_password',
rules: { min_length: 6 },
error: 'WEAK_PASSWORD',
});
Or you can write more sophisticated rules directly:
const validator = new LIVR.Validator({
password: ['required', 'strong_password'],
});
validator.registerRules({
strong_password() {
return (value) => {
if (value === undefined || value === null || value === '') return;
if (value.length < 6) {
return 'WEAK_PASSWORD';
}
};
},
});
Or you can write more sophisticated async rules as well:
const validator = new LIVR.AsyncValidator({
userId: ['required', 'valid_user_id'],
});
validator.registerRules({
valid_user_id() {
return async (value) => {
if (value === undefined || value === null || value === '') return;
const user = await Users.findUserById(value);
if (!user) {
return 'WRONG_USER_ID';
}
};
},
});
If you use LIVR in browser, you can import only the rules you use (it can reduce budle size a little bit):
import Validator from 'livr/lib/Validator';
Validator.registerDefaultRules({
required: require('livr/lib/rules/common/required'),
email: require('livr/lib/rules/special/email'),
one_of: require('livr/lib/rules/string/one_of'),
min_length: require('livr/lib/rules/string/min_length'),
max_length: require('livr/lib/rules/string/max_length'),
equal_to_field: require('livr/lib/rules/special/equal_to_field'),
});
Validator.defaultAutoTrim(true);
import Validator from 'livr/lib/Validator';
const validator = new Validator({
name: 'required',
email: ['required', 'email'],
gender: { one_of: ['male', 'female'] },
phone: { max_length: 10 },
password: ['required', { min_length: 10 }],
password2: { equal_to_field: 'password' },
});
const validData = validator.validate(userData);
if (validData) {
saveUser(validData);
} else {
console.log('errors', validator.getErrors());
}
DESCRIPTION
See LIVR Specification and rules documentation for detailed documentation and list of supported rules.
Features:
- Rules are declarative and language independent
- Any number of rules for each field
- Return together errors for all fields
- Excludes all fields that do not have validation rules described
- Has possibility to validate complex hierarchical structures
- Easy to describe and understand rules
- Returns understandable error codes(not error messages)
- Easy to add own rules
- Rules are be able to change results output ("trim", "nested_object", for example)
- Multipurpose (user input validation, configs validation, contracts programming etc)
- Supports sync and async validation
JavaScript version extra features:
- Zero dependencies
- Works in NodeJs and in a browser
- Validator (without rules) less than 1KB (min+gzip)
- Validator with all rules 2.84KB (min+gzip)
- You can find more rules in livr-extra-rules
ASYNC VALIDATION (NEW)
LIVR supports async validation but it was added only in v2.5. So, it uses a little bit different API.
What you need to know about implementation:
- All simple sync rules are supported out of the box.
- Meta rules (rules that construct a new validator instance inside them) were rewritten to use AsyncValidator.
If you import "livr/async" they will be automatically used
- Fields validation is done in parallel but rules for one field are processed one after another.
Usage example:
import LIVR from 'livr/async';
LIVR.AsyncValidator.defaultAutoTrim(true);
const validator = new LIVR.AsyncValidator({
name: 'required',
email: ['required', 'email'],
});
try {
const validData = await validator.validate(userData);
saveUser(validData);
} catch (errors) {
console.log('errors', errors);
}
INSTALL
nodejs/npm
npm install livr
Browser (if you do not use npm)
You can find prebuilt browser versions in "dist" folder
- development/main.js - not minified development version with source maps
- production/main.js - minified production version. Possible you will need some polyfills ("isInteger" etc) for older browsers.
- development-async/main.js - not minified development version with source maps of "AsyncValidator"
- production-async/main.js - minified production version of "AsyncValidator"
CLASS METHODS
new LIVR.Validator(rules, options);
Constructor creates validator objects.
rules - validations rules. Rules description is available here - https://livr-spec.org/
Supported options:
- "autoTrim" (default false) - asks validator to trim all values before validation. Output will be also trimmed. If key is not passed then defaultAutoTrim value will be used.
Instead of "options" object "isAutoTrim" boolean value can be passed for compatibility with previous API.
if isAutoTrim is undefined(or null) then defaultAutoTrim value will be used.
LIVR.Validator.registerAliasedDefaultRule(alias)
alias - is a plain javascript object that contains: name, rules, error (optional).
LIVR.Validator.registerAliasedDefaultRule({
name: 'valid_address',
rules: {
nested_object: {
country: 'required',
city: 'required',
zip: 'positive_integer',
},
},
});
Then you can use "valid_address" for validation:
{
address: 'valid_address';
}
You can register aliases with own errors:
LIVR.Validator.registerAliasedDefaultRule({
name: 'adult_age'
rules: [ 'positive_integer', { min_number: 18 } ],
error: 'WRONG_AGE'
});
All rules/aliases for the validator are equal. The validator does not distinguish "required", "list_of_different_objects" and "trim" rules. So, you can extend validator with any rules/alias you like.
Note: Each rule which contains uderscore in name will be additionally registered using camel case name if there is no such rule name already.
LIVR.Validator.registerDefaultRules({"rule_name": ruleBuilder })
ruleBuilder - is a function reference which will be called for building single rule validator.
Note: Each rule which contains uderscore in name will be additionally registered using camel case name if there is no such rule name already.
LIVR.Validator.registerDefaultRules({
my_rule(arg1, arg2, arg3, ruleBuilders) {
return (value, allValues, outputArr) => {
if (notValid) {
return 'SOME_ERROR_CODE';
} else {
}
};
},
});
Then you can use "my_rule" for validation:
{
name1: 'my_rule'
name2: { 'my_rule': arg1 }
name3: { 'my_rule': [arg1] }
name4: { 'my_rule': [ arg1, arg2, arg3 ] }
}
Here is "max_number" implemenation:
function maxNumber(maxNumber) {
return (value) => {
if (value === undefined || value === null || value === '') return;
if (value > maxNumber) return 'TOO_HIGH';
};
}
LIVR.Validator.registerDefaultRules({ max_number: maxNumber });
All rules for the validator are equal. The validator does not distinguish "required", "list_of_different_objects" and "trim" rules. So, you can extend validator with any rules you like.
LIVR.Validator.getDefaultRules();
returns object containing all default ruleBuilders for the validator. You can register new rule or update existing one with "registerRules" method.
LIVR.Validator.defaultAutoTrim(isAutoTrim)
Enables or disables automatic trim for input data. If is on then every new validator instance will have auto trim option enabled
LIVR.util
List of useful utils for writing your rules (see source code)
OBJECT METHODS
validator.prepare()
Parses all validation rules to make subsequent calls faster. This step is always automatically called on first call of validator.validate(input) but you can call it manually if you want to warm up your validator object before validation was called. Usually, it is useful in several cases:
- Benchmarks. As first validation call will take more time without prepare
- Custom meta rules. Just to prepare nested validators when parent validator prepare method is called.
validator.validate(input)
Validates user input. On success returns validData (contains only data that has described validation rules). On error return false.
const validData = validator.validate(input);
if (validData) {
} else {
const errors = validator.getErrors();
}
for AsyncValidator
try {
const validData = await validator.validate(input);
} catch (errors) {
}
validator.getErrors() (only for sync version of validator)
Returns errors object.
{
"field1": "ERROR_CODE",
"field2": "ERROR_CODE",
...
}
For example:
{
"country": "NOT_ALLOWED_VALUE",
"zip": "NOT_POSITIVE_INTEGER",
"street": "REQUIRED",
"building": "NOT_POSITIVE_INTEGER"
}
validator.registerRules({"rule_name": ruleBuilder})
ruleBuilder - is a function reference which will be called for building single rule validator.
See "LIVR.Validator.registerDefaultRules" for rules examples.
validator.registerAliasedRule(alias)
alias - is a composite validation rule.
See "LIVR.Validator.registerAliasedDefaultRule" for rules examples.
validator.getRules()
returns object containing all ruleBuilders for the validator. You can register new rule or update existing one with "registerRules" method.
Performance
LIVR is fast but you should be aware about following:
Do not construct Validator for each validation call. Construct object once for each schema and reuse validators with different inputs. "validator.validate(input)" is very fast.
In some cases you need to construct object each time, it is slower but still ok. It still will be twice faster than "Joi". LIVR validator preparation (rules compile step) is 100 time faster than "fastest-validator" compile time.
AUTHOR
koorchik (Viktor Turskyi)
Contributors
eNdiD
BUGS
Please report any bugs or feature requests to Github https://github.com/koorchik/js-validator-livr