A flexible JSON object linter with out of the box support for OpenAPI v2 and v3
- Create custom rules to lint any JSON object
- Use JSON paths to apply rules / functions to specific parts of your JSON objects
- Built-in set of functions to help build custom rules. Functions include pattern checks, parameter checks, alphabetical ordering, a specified number of characters, provided keys are present in an object, etc
- Create custom functions for advanced use cases
- Optional ready to use rules and functions to validate and lint OpenAPI v2 and v3 documents
- Validate JSON with Ajv
npm install @stoplight/spectral
npm install -g @stoplight/spectral
Supports Node v8.3+.
For users without Node and/or NPM/Yarn, we provide standalone packages for all major platforms:
- x64 Windows
- x64 MacOS
- x64 Linux
You can find them here. Once downloaded, you can proceed with the standard procedure for running any CLI tool.
./spectral-macos lint petstore.yaml
Note, the binaries are not auto-updatable, therefore you will need to download a new version on your own.
sudo mv ./spectral-linux /usr/local/bin/spectral
You may need to restart your terminal.
Now, spectral
command will be accessible in your terminal.
Head over to releases for the latest binaries.
Spectral can be run via the command-line:
spectral lint petstore.yaml
Other options include:
-e, --encoding=encoding [default: utf8] text encoding to use
-f, --format=json|stylish [default: stylish] formatter to use for outputting results
-h, --help show CLI help
-m, --maxResults=maxResults [default: all] maximum results to show
-o, --output=output output to a file instead of stdout
-v, --verbose increase verbosity
Note: The Spectral CLI supports both YAML and JSON.
Currently, the CLI supports validation of OpenAPI documents and lints them based on our default ruleset. It does not support custom rulesets at this time. Although if you want to build and run custom rulesets outside of the CLI, see Customization.
Spectral includes a number of ready made rules and functions for OpenAPI v2 and v3 documents.
This example uses the OpenAPI v3 rules to lint a document.
const { Spectral } = require('@stoplight/spectral');
const { oas3Functions, oas3Rules } = require('@stoplight/spectral/rulesets/oas3');
// an OASv3 document
const myOAS = {
// ... properties in your document
responses: {
'200': {
description: '',
schema: {
$ref: '#/definitions/error-response',
},
},
},
// ... properties in your document
};
// create a new instance of spectral with all of the baked in rulesets
const spectral = new Spectral();
spectral.addFunctions(oas3Functions());
spectral.addRules(oas3Rules());
spectral.addRules({
// .. extend with your own custom rules
});
// run!
spectral.run(myOAS).then(results => {
console.log(JSON.stringify(results, null, 4));
});
You can also add to these rules to create a customized linting style guide for your OpenAPI documents.
The existing OAS rules are opinionated. There might be some rules that you prefer to change. We encourage you to create your rules to fit your use case. We welcome additions to the existing rulesets as well!
You can find all about rulesets here.
There are three key concepts in Spectral: Rulesets, Rules and Functions.
- Ruleset is a container for a collection of rules and functions.
- Rule filters your object down to a set of target values, and specify the function that should evaluate those values.
- Function accept a value and return issue(s) if the value is incorrect.
Think of a set of rules and functions as a flexible and customizable style guide for your JSON objects.
Spectral has a built-in set of functions which you can reference in your rules. This example uses the RuleFunction.PATTERN
to create a rule that checks that all property values are in snake case.
const { RuleFunction, Spectral } = require('@stoplight/spectral');
const spectral = new Spectral();
spectral.addRules({
snake_case: {
summary: 'Checks for snake case pattern',
// evaluate every property
given: '$..*',
then: {
function: RuleFunction.PATTERN,
functionOptions: {
match: '^[a-z]+[a-z0-9_]*[a-z0-9]+$',
},
},
},
});
// run!
spectral.run({name: 'helloWorld',}).then(results => {
console.log(JSON.stringify(results, null, 4));
});
// => outputs a single result since `helloWorld` is not snake_case
// [
// {
// "code": "snake_case",
// "message": "must match the pattern '^[a-z]+[a-z0-9_]*[a-z0-9]+$'",
// "severity": 1,
// "path": [
// "name"
// ]
// }
// ]
Sometimes the built-in functions don't cover your use case. This example creates a custom function, customNotThatFunction
, and then uses it within a rule, openapi_not_swagger
. The custom function checks that you are not using a specific string (e.g., "Swagger") and suggests what to use instead (e.g., "OpenAPI").
const { Spectral } = require('@stoplight/spectral');
// custom function
const customNotThatFunction = (targetValue, options) => {
const { match, suggestion } = options;
if (targetValue && targetValue.match(new RegExp(match))) {
// return the single error
return [
{
message: `Use ${suggestion} instead of ${match}!`,
},
];
}
};
const spectral = new Spectral();
spectral.addFunctions({
notThat: customNotThatFunction,
});
spectral.addRules({
openapi_not_swagger: {
summary: 'Checks for use of Swagger, and suggests OpenAPI.',
// check every property
given: '$..*',
then: {
// reference the function we added!
function: 'notThat',
// pass it the options it needs
functionOptions: {
match: 'Swagger',
suggestion: 'OpenAPI',
},
},
},
});
// run!
spectral.run({description: 'Swagger is pretty cool!',}).then(results => {
console.log(JSON.stringify(results, null, 4));
});
// => outputs a single result since we are using the term `Swagger` in our object
// [
// {
// "code": "openapi_not_swagger",
// "message": "Use OpenAPI instead of Swagger!",
// "severity": 1,
// "path": [
// "description"
// ]
// }
// ]
How is this different than Ajv?
Ajv is a JSON Schema validator, not a linter. Spectral does expose a schema
function that you can use in your rules to validate all or part of the target object with JSON Schema (Ajv is used under the hood). However, Spectral also provides a number of other functions and utilities that you can use to build up a linting ruleset to validates things that JSON Schema is not well suited for.
I want to lint my OpenAPI documents but don't want to implement Spectral right now.
No problem! A hosted version of Spectral comes free with the Stoplight platform. Sign up for a free account here.
What is the difference between Spectral and Speccy?
With Spectral, lint rules can be applied to any JSON object. Speccy is designed to work with OpenAPI v3 only. The rule structure is different between the two. Spectral uses JSONPath path
parameters instead of the object
parameters (which are OpenAPI specific). Rules are also more clearly defined (thanks to TypeScript typings) and now require specifying a type
parameter. Some rule types have been enhanced to be a little more flexible along with being able to create your own rules based on the built-in and custom functions.
If you are interested in contributing to Spectral itself, check out our contributing docs to get started.
Also, most of the interesting projects are built with Spectral. Please consider using Spectral in a project or contribute to an existing one.
If you are using Spectral in your project and want to be listed in the examples section, we encourage you to open an issue.
- Stoplight's Custom Style and Validation Rules uses Spectral to validate and lint OpenAPI documents on the Stoplight platform
- Spectral GitHub Bot, a GitHub pull request bot that lints your repo's OpenAPI document that uses the Probot framework, built by Taylor Barnett
- Spectral GitHub Action, a GitHub Action that lints your repo's OpenAPI document, built by Vincenzo Chianese
- JSONPath Online Evaluator, a helpful tool to determine what
path
you want - stoplightio/json, a library of useful functions for when working with JSON
- stoplightio/yaml, a library of useful functions for when working with YAML, including parsing YAML into JSON, and a few helper functions such as
getJsonPathForPosition
orgetLocationForJsonPath
- Phil Sturgeon for collaboration and creating Speccy
- Mike Ralphson for kicking off the Spectral CLI
If you have a bug or feature request, please open an issue here.
If you need help using Spectral or have a support question, please use the Stoplight Community forum. We've created an open source category for these questions. It's also a great place to share your implementations.
If you want to discuss something in private, you can reach out to Stoplight support at support@stoplight.io.