Top Related Projects
The fastest JSON schema Validator. Supports JSON Schema draft-04/06/07/2019-09/2020-12 and JSON Type Definition (RFC8927)
PHP implementation of JSON schema. Fork of the http://jsonschemaphpv.sourceforge.net/ project
Tiny Validator for JSON Schema v4
The JSON Schema specification
Quick Overview
tdegrunt/jsonschema is a JavaScript implementation of JSON Schema validation. It allows developers to validate JSON objects against a predefined schema, ensuring data consistency and integrity in JavaScript applications.
Pros
- Lightweight and easy to use
- Supports JSON Schema draft-04
- Can be used in both browser and Node.js environments
- Actively maintained with regular updates
Cons
- Limited support for newer JSON Schema versions
- May have performance issues with very large schemas or data sets
- Documentation could be more comprehensive
- Lacks some advanced features found in other JSON Schema validators
Code Examples
Validating a simple object:
const Validator = require('jsonschema').Validator;
const v = new Validator();
const schema = {
"type": "object",
"properties": {
"name": { "type": "string" },
"age": { "type": "number", "minimum": 0 }
},
"required": ["name", "age"]
};
const data = { "name": "John Doe", "age": 30 };
console.log(v.validate(data, schema).valid); // Output: true
Using custom formats:
const Validator = require('jsonschema').Validator;
const v = new Validator();
v.addFormat("uppercase", function(input) {
return input === input.toUpperCase();
});
const schema = {
"type": "string",
"format": "uppercase"
};
console.log(v.validate("HELLO", schema).valid); // Output: true
console.log(v.validate("Hello", schema).valid); // Output: false
Validating nested objects:
const Validator = require('jsonschema').Validator;
const v = new Validator();
const schema = {
"type": "object",
"properties": {
"user": {
"type": "object",
"properties": {
"name": { "type": "string" },
"email": { "type": "string", "format": "email" }
},
"required": ["name", "email"]
}
}
};
const data = {
"user": {
"name": "John Doe",
"email": "john@example.com"
}
};
console.log(v.validate(data, schema).valid); // Output: true
Getting Started
To use tdegrunt/jsonschema in your project, follow these steps:
-
Install the package:
npm install jsonschema
-
Import the Validator in your JavaScript file:
const Validator = require('jsonschema').Validator; const v = new Validator();
-
Define your schema and validate data:
const schema = { "type": "string" }; const data = "Hello, world!"; const result = v.validate(data, schema); console.log(result.valid); // true if valid, false otherwise
Competitor Comparisons
The fastest JSON schema Validator. Supports JSON Schema draft-04/06/07/2019-09/2020-12 and JSON Type Definition (RFC8927)
Pros of ajv
- Significantly faster performance, especially for large schemas and data
- More comprehensive support for JSON Schema draft versions
- Extensive customization options and plugins
Cons of ajv
- Steeper learning curve due to more advanced features
- Larger package size, which may impact bundle size in browser environments
Code comparison
ajv:
const Ajv = require('ajv');
const ajv = new Ajv();
const validate = ajv.compile(schema);
const valid = validate(data);
if (!valid) console.log(validate.errors);
jsonschema:
const Validator = require('jsonschema').Validator;
const v = new Validator();
const result = v.validate(data, schema);
if (!result.valid) console.log(result.errors);
Key differences
- ajv offers better performance and more features, making it suitable for complex validation scenarios
- jsonschema has a simpler API and smaller package size, which may be preferable for basic use cases
- ajv provides more detailed error messages and customization options
- jsonschema has a more straightforward setup process and may be easier for beginners to use
Both libraries are widely used and maintained, with ajv being more popular due to its performance and feature set. The choice between them depends on project requirements, performance needs, and developer preferences.
PHP implementation of JSON schema. Fork of the http://jsonschemaphpv.sourceforge.net/ project
Pros of json-schema
- More actively maintained with recent updates
- Supports newer JSON Schema versions (up to draft-07)
- Better documentation and examples provided in the repository
Cons of json-schema
- Smaller community and fewer contributors
- Less comprehensive test suite compared to jsonschema
- May have fewer features or edge case handling
Code Comparison
jsonschema:
var JSV = require('JSV').JSV;
var env = JSV.createEnvironment();
var schema = {...};
var data = {...};
var report = env.validate(data, schema);
json-schema:
const Validator = require('jsonschema').Validator;
const v = new Validator();
const schema = {...};
const data = {...};
const validationResult = v.validate(data, schema);
Both libraries provide similar functionality for JSON Schema validation, but json-schema offers a more modern and actively maintained codebase. However, jsonschema has a larger community and potentially more robust testing. The choice between the two may depend on specific project requirements and the need for newer JSON Schema version support.
Tiny Validator for JSON Schema v4
Pros of tv4
- More comprehensive support for JSON Schema draft 4
- Better performance for large-scale validation tasks
- Includes a command-line interface for easy integration into build processes
Cons of tv4
- Less frequent updates and maintenance
- May have compatibility issues with newer JSON Schema drafts
- Limited support for custom formats and keywords
Code Comparison
tv4:
var valid = tv4.validate(data, schema);
if (valid) {
console.log('Valid');
} else {
console.log(tv4.error);
}
jsonschema:
var validator = new Validator();
var result = validator.validate(data, schema);
if (result.valid) {
console.log('Valid');
} else {
console.log(result.errors);
}
Key Differences
- tv4 uses a global object for validation, while jsonschema creates instances
- tv4 returns a boolean for validation, jsonschema returns a result object
- Error handling differs: tv4 uses a single error object, jsonschema provides an array of errors
Use Cases
- tv4: Ideal for projects requiring high-performance validation and CLI integration
- jsonschema: Better suited for applications needing flexibility and custom extensions
Community and Ecosystem
- tv4: Larger user base, more third-party extensions available
- jsonschema: Smaller community, but actively maintained and updated
The JSON Schema specification
Pros of json-schema-spec
- Official specification repository, providing the most up-to-date and authoritative information
- Comprehensive documentation and examples for implementing JSON Schema
- Active community involvement and regular updates
Cons of json-schema-spec
- Not a ready-to-use implementation, requires additional effort to integrate
- May be overwhelming for developers seeking a simple validation solution
- Lacks language-specific implementations or bindings
Code comparison
json-schema-spec (example schema):
{
"$schema": "https://json-schema.org/draft/2020-12/schema",
"type": "object",
"properties": {
"name": { "type": "string" },
"age": { "type": "integer", "minimum": 0 }
}
}
jsonschema (usage example):
var Validator = require('jsonschema').Validator;
var v = new Validator();
var instance = { name: "John", age: 30 };
var schema = { "type": "object", "properties": { "name": { "type": "string" }, "age": { "type": "integer", "minimum": 0 } } };
console.log(v.validate(instance, schema));
Summary
json-schema-spec is the official specification repository, offering comprehensive documentation and community support. However, it's not a ready-to-use implementation like jsonschema. The latter provides a simpler, language-specific solution for JSON validation in JavaScript, making it more accessible for quick integration in projects. Developers should choose based on their specific needs: comprehensive specification or easy-to-use implementation.
Convert designs to code with AI
Introducing Visual Copilot: A new AI model to turn Figma designs to high quality code using your components.
Try Visual CopilotREADME
jsonschema
JSON schema validator, which is designed to be fast and simple to use. JSON Schema versions through draft-07 are fully supported.
Contributing & bugs
Please fork the repository, make the changes in your fork and include tests. Once you're done making changes, send in a pull request.
Bug reports
Please include a test which shows why the code fails.
Usage
Simple
Simple object validation using JSON schemas.
var Validator = require('jsonschema').Validator;
var v = new Validator();
var instance = 4;
var schema = {"type": "number"};
console.log(v.validate(instance, schema));
Even simpler
var validate = require('jsonschema').validate;
console.log(validate(4, {"type": "number"}));
Complex example, with split schemas and references
var Validator = require('jsonschema').Validator;
var v = new Validator();
// Address, to be embedded on Person
var addressSchema = {
"id": "/SimpleAddress",
"type": "object",
"properties": {
"lines": {
"type": "array",
"items": {"type": "string"}
},
"zip": {"type": "string"},
"city": {"type": "string"},
"country": {"type": "string"}
},
"required": ["country"]
};
// Person
var schema = {
"id": "/SimplePerson",
"type": "object",
"properties": {
"name": {"type": "string"},
"address": {"$ref": "/SimpleAddress"},
"votes": {"type": "integer", "minimum": 1}
}
};
var p = {
"name": "Barack Obama",
"address": {
"lines": [ "1600 Pennsylvania Avenue Northwest" ],
"zip": "DC 20500",
"city": "Washington",
"country": "USA"
},
"votes": "lots"
};
v.addSchema(addressSchema, '/SimpleAddress');
console.log(v.validate(p, schema));
Returned ValidatorResult object, will show this example is NOT valid since: "votes": "lots"
is not an integer.
Example for Array schema
var arraySchema = {
"type": "array",
"items": {
"properties": {
"name": { "type": "string" },
"lastname": { "type": "string" }
},
"required": ["name", "lastname"]
}
}
For a comprehensive, annotated example illustrating all possible validation options, see examples/all.js
Features
Definitions
All schema definitions are supported, $schema is ignored.
Types
All types are supported
Handling undefined
undefined
is not a value known to JSON, and by default, the validator treats it as if it is not invalid. i.e., it will return valid.
var res = validate(undefined, {type: 'string'});
res.valid // true
This behavior may be changed with the "required" option:
var res = validate(undefined, {type: 'string'}, {required: true});
res.valid // false
Formats
Disabling the format keyword.
You may disable format validation by providing disableFormat: true
to the validator
options.
String Formats
All formats are supported, phone numbers are expected to follow the E.123 standard.
Custom Formats
You may add your own custom format functions. Format functions accept the input
being validated and return a boolean value. If the returned value is true
, then
validation succeeds. If the returned value is false
, then validation fails.
- Formats added to
Validator.prototype.customFormats
do not affect previously instantiated Validators. This is to prevent validator instances from being altered once created. It is conceivable that multiple validators may be created to handle multiple schemas with different formats in a program. - Formats added to
validator.customFormats
affect only that Validator instance.
Here is an example that uses custom formats:
Validator.prototype.customFormats.myFormat = function(input) {
return input === 'myFormat';
};
var validator = new Validator();
validator.validate('myFormat', {type: 'string', format: 'myFormat'}).valid; // true
validator.validate('foo', {type: 'string', format: 'myFormat'}).valid; // false
Results
By default, results will be returned in a ValidatorResult
object with the following properties:
instance
: any.schema
: Schema.errors
: ValidationError[].valid
: boolean.
Each item in errors
is a ValidationError
with the following properties:
- path: array. An array of property keys or array offsets, indicating where inside objects or arrays the instance was found.
- property: string. Describes the property path. Starts with
instance
, and is delimited with a dot (.
). - message: string. A human-readable message for debugging use. Provided in English and subject to change.
- schema: object. The schema containing the keyword that failed
- instance: any. The instance that failed
- name: string. The keyword within the schema that failed.
- argument: any. Provides information about the keyword that failed.
The validator can be configured to throw in the event of a validation error:
-
If the
throwFirst
option is set, the validator will terminate validation at the first encountered error and throw aValidatorResultError
object. -
If the
throwAll
option is set, the validator will throw aValidatorResultError
object after the entire instance has been validated. -
If the
throwError
option is set, it will throw at the first encountered validation error (likethrowFirst
), but theValidationError
object itself will be thrown. Note that, despite the name, this does not inherit from Error likeValidatorResultError
does.
The ValidatorResultError
object has the same properties as ValidatorResult
and additionally inherits from Error.
"nestedErrors" option
When oneOf
or anyOf
validations fail, errors that caused any of the sub-schemas referenced therein to fail are normally suppressed, because it is not necessary to fix all of them. And in the case of oneOf
, it would itself be an error to fix all of the listed errors.
This behavior may be configured with options.nestedErrors
. If truthy, it will emit all the errors from the subschemas. This option may be useful when troubleshooting validation errors in complex schemas:
var schema = {
oneOf: [
{ type: 'string', minLength: 32, maxLength: 32 },
{ type: 'string', maxLength: 16 },
{ type: 'number' },
]
};
var validator = new Validator();
var result = validator.validate('This string is 28 chars long', schema, {nestedErrors: true});
// result.toString() reads out:
// 0: instance does not meet minimum length of 32
// 1: instance does not meet maximum length of 16
// 2: instance is not of a type(s) number
// 3: instance is not exactly one from [subschema 0],[subschema 1],[subschema 2]
Localizing Error Messages
To provide localized, human-readable errors, use the name
string as a translation key. Feel free to open an issue for support relating to localizing error messages. For example:
var localized = result.errors.map(function(err){
return localeService.translate(err.name);
});
Custom keywords
Specify your own JSON Schema keywords with the validator.attributes property:
validator.attributes.contains = function validateContains(instance, schema, options, ctx) {
if(typeof instance !== 'string') return;
if(typeof schema.contains !== 'string') throw new jsonschema.SchemaError('"contains" expects a string', schema);
if(instance.indexOf(schema.contains)<0){
return 'does not contain the string ' + JSON.stringify(schema.contains);
}
}
var result = validator.validate("I am an instance", { type:"string", contains: "I am" });
// result.valid === true;
The instance passes validation if the function returns nothing. A single validation error is produced
if the function returns a string. Any number of errors (maybe none at all) may be returned by passing a
ValidatorResult
object, which may be used like so:
var result = new ValidatorResult(instance, schema, options, ctx);
while(someErrorCondition()){
result.addError('fails some validation test');
}
return result;
Dereferencing schemas
Sometimes you may want to download schemas from remote sources, like a database, or over HTTP. When importing a schema,
unknown references are inserted into the validator.unresolvedRefs
Array. Asynchronously shift elements off this array and import
them:
var Validator = require('jsonschema').Validator;
var v = new Validator();
v.addSchema(initialSchema);
function importNextSchema(){
var nextSchema = v.unresolvedRefs.shift();
if(!nextSchema){ done(); return; }
databaseGet(nextSchema, function(schema){
v.addSchema(schema);
importNextSchema();
});
}
importNextSchema();
Disallowing unknown attributes
Sometimes you may want to disallow unknown attributes passed in the body of the request, in order to disallow those unknown attributes before the validation of the body, you need to set additionalProperties to false.
{
"$schema": "http://json-schema.org/draft-04/schema#",
"title": "Accounting Resource - Add Item",
"type": "object",
"additionalProperties": false,
"properties": {
"itemNumber": {
"type":"string"
},
"title": {
"type":"string"
},
"description": {
"type":"string"
}
},
"required": [
"itemNumber",
"title",
"description"
]
}
Default base URI
Schemas should typically have an id
with an absolute, full URI. However if the schema you are using contains only relative URI references, the base
option will be used to resolve these.
This following example would throw a SchemaError
if the base
option were unset:
var result = validate(["Name"], {
id: "/schema.json",
type: "array",
items: { $ref: "http://example.com/schema.json#/definitions/item" },
definitions: {
item: { type: "string" },
},
}, { base: 'http://example.com/' });
Rewrite Hook
The rewrite
option lets you change the value of an instance after it has successfully been validated. This will mutate the instance
passed to the validate function. This can be useful for unmarshalling data and parsing it into native instances, such as changing a string to a Date
instance.
The rewrite
option accepts a function with the following arguments:
- instance: any
- schema: object
- options: object
- ctx: object
- return value: any new value for the instance
The value may be removed by returning undefined
.
If you don't want to change the value, call return instance
.
Here is an example that can convert a property expecting a date into a Date instance:
const schema = {
properties: {
date: {id: 'http://example.com/date', type: 'string'},
},
};
const value = {
date: '2020-09-30T23:39:27.060Z',
};
function unmarshall(instance, schema){
if(schema.id === 'http://example.com/date'){
return new Date(instance);
}
return instance;
}
const v = new Validator();
const res = v.validate(value, schema, {rewrite: unmarshall});
assert(res.instance.date instanceof Date);
Pre-Property Validation Hook
If some processing of properties is required prior to validation a function may be passed via the options parameter of the validate function. For example, say you needed to perform type coercion for some properties:
// See examples/coercion.js
function preValidateProperty(object, key, schema, options, ctx) {
var value = object[key];
if (typeof value === 'undefined') return;
// Test if the schema declares a type, but the type keyword fails validation
if (schema.type && validator.attributes.type.call(validator, value, schema, options, ctx.makeChild(schema, key))) {
// If the type is "number" but the instance is not a number, cast it
if(schema.type==='number' && typeof value!=='number'){
object[key] = parseFloat(value);
return;
}
// If the type is "string" but the instance is not a string, cast it
if(schema.type==='string' && typeof value!=='string'){
object[key] = String(value).toString();
return;
}
}
};
// And now, to actually perform validation with the coercion hook!
v.validate(instance, schema, { preValidateProperty });
Skip validation of certain keywords
Use the "skipAttributes" option to skip validation of certain keywords. Provide an array of keywords to ignore.
For skipping the "format" keyword, see the disableFormat option.
Fail on unknown keywords
By default, JSON Schema is supposed to ignore unknown schema keywords.
You can change this behavior to require that all keywords used in a schema have a defined behavior, by using setting the "allowUnknownAttributes" option to false.
This example will throw a SchemaError
:
var schema = {
type: "string",
format: "email",
example: "foo",
};
var result = validate("Name", schema, { allowUnknownAttributes: false });
Tests
Uses JSON Schema Test Suite as well as our own tests. You'll need to update and init the git submodules:
git submodule update --init
npm test
Contributions
This library would not be possible without the valuable contributions by:
- Austin Wright
... and many others!
License
jsonschema is licensed under MIT license.
Copyright (C) 2012-2019 Tom de Grunt <tom@degrunt.nl>
Permission is hereby granted, free of charge, to any person obtaining a copy of
this software and associated documentation files (the "Software"), to deal in
the Software without restriction, including without limitation the rights to
use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies
of the Software, and to permit persons to whom the Software is furnished to do
so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
Top Related Projects
The fastest JSON schema Validator. Supports JSON Schema draft-04/06/07/2019-09/2020-12 and JSON Type Definition (RFC8927)
PHP implementation of JSON schema. Fork of the http://jsonschemaphpv.sourceforge.net/ project
Tiny Validator for JSON Schema v4
The JSON Schema specification
Convert designs to code with AI
Introducing Visual Copilot: A new AI model to turn Figma designs to high quality code using your components.
Try Visual Copilot