OpenAPI Validator

This command line tool lets you validate OpenAPI documents according to their specification, either 2.0 or 3.0, as well as IBM-defined best practices.

Prerequisites

• Node 10.x
• NPM 5.x

Table of contents

• Getting Started
  • Customization
• Installation
  • Install with NPM (recommended)
  • Build from source
  • Platform specific binaries
  • Docker container
• Usage
  • Command line
  • Node module
    • API
    • Validation results
• Configuration
  • Setup
  • Definitions
    • Specs
    • Categories
    • Rules
    • Statuses
    • Configuration Options
  • Configuration file
  • Default mode
    • Default values
  • Spectral configuration
    • Changing Spectral rule severity
    • Custom Spectral rules
• Warnings Limit
• Turning off update-notifier
• License

Getting Started

The validator analyzes your API definition and reports any problems within. The validator is highly customizable, and supports both OpenAPI 3.0 and OpenAPI 2.0 (Swagger 2.0) formats. The tool also supports a number of rules from Spectral. You can easily extend the tool with custom rules to meet your specific needs and ensure compliance to your standards.

The default configuration uses both OpenAPI 3.0 rules as well as Spectral rules. The default mode section decscribes these rules. Get started by installing the tool, then run the tool on your API definition.

Customization

You can modify the behavior of the validator for your project to meet your preferred standards. Specific validation "rules" can be turned off, or configured to trigger an error, warning, info, or hint message in the validator output.

Some validations can be configured even further, such as switching the case convention for parameter names. There are also currently some validations that cannot be disabled or configured to a different severity. You can see the rule associated with each message produced by the validator with the -v command line option. Rules that are not configurable will show the name builtin.

The validator also employs the Spectral validation/linting engine to detect certain issues in the API document. Spectral rules can also be configured to trigger an error, warning, info, or hint message in the validator output with the .spectral.yaml configuration file. When the validator issues a message as the result of a Spectral rule, the rule name displayed will correspond to the Spectral rule. You must add a Spectral configuration if you want to create a new rule, as opposed to modifying an existing one.

To get started configuring the validator, set up a configuration file. See the Spectral configuration section for more details on customization with Spectral.

Installation

There are two main ways to install the validator, either using NPM or building from source. Installing with NPM is reccomended.

Install with NPM (recommended)

npm install -g ibm-openapi-validator

The -g flag installs the tool globally so that the validator can be run from anywhere in the file system. Alternatively, you can pass no flag or the --save-dev flag to add the validator as a dependency to your project and run it from your NPM scripts.

Build from source

1. Clone or download this repository
2. Navigate to the root directory of this project.
3. Install the dependencies using npm install
4. Build the command line tool by running npm run link.

If you installed the validator using npm install -g ibm-openapi-validator, you will need to run npm uninstall -g ibm-openapi-validator before running npm run link.

Platform specific binaries

It is possible to build platform specific binaries for Linux, MacOS, and Windows that do not depend on having node.js installed.

To build these, run npm run pkg in the root. The binaries (lint-openapi-linux, lint-openapi-macos, lint-openapi-windows.exe respectively) are built in the 'bin' directory in the root.

Docker container

A community Docker image is publicly available on Docker hub.

docker pull jamescooke/openapi-validator

Once pulled, the container can be run directly, but mount a volume containing the OpenAPI specification file so that it can be accessed.

docker run --volume "$PWD":/data jamescooke/openapi-validator [options] [command] [<files>]

Usage

Command line

lint-openapi [options] [command] [<files>]

[options]

• -c (--config) <path/to/your/config> : Path to a validator configuration file. If provided, this is used instead of .validaterc.
• -d (--default_mode) : This option turns off configuration and runs the validator in default mode.
• -e (--errors_only) : Only print the errors, ignore the warnings.
• -j (--json) : Output results as a JSON object
• -n (--no_colors) : The output is colored by default. If this bothers you, this flag will turn off the coloring.
• -p (--print_validator_modules) : Print the name of the validator source file the error/warning was caught it. This can be helpful for developing validations.
• -v (--verbose) : Increase verbosity of reported results. Use this option to display the rule for each reported result.
• -r (--ruleset) <path/to/your/ruleset> : Path to Spectral ruleset file, used instead of .spectral.yaml if provided.
• -s (--report_statistics) : Print a simple report at the end of the output showing the frequency, in percentage, of each error/warning.
• --debug : Enable debugging output.
• --version : Print the current semantic version of the validator
• -h (--help) : This option prints the usage menu.

These options only apply to running the validator on a file, not to any commands.

[command]

$ lint-openapi init

• init : The init command initializes a .validaterc file, used to configure the validator. It can also be used to reset the configurable rules to their default values.

[command]

$ lint-openapi migrate

• migrate : The migrate command migrates a .validaterc file from the legacy format to the current format, retaining all custom rules. The new format is required - this command provides an option to keep custom rules without manually updating the file or initializing a new configuration file with all rules set to the defaults using lint-openapi init.

None of the above options pertain to these commands.

<files>

• The OpenAPI document(s) to be validated. All files must be a valid JSON or YAML (only .json, .yml, and .yaml file extensions are supported).
• Multiple, space-separated files can be passed in and each will be validated. This includes support for globs (e.g. lint-openapi files/* will run the validator on all files in files/)

Node module

const validator = require('ibm-openapi-validator');

validator(openApiDoc)
  .then(validationResults => {
    console.log(JSON.stringify(validationResults, null, 2));
  });

// or, if inside `async` function
const validationResults = await validator(openApiDoc);
console.log(JSON.stringify(validationResults, null, 2));

API

validator(openApiDoc, [defaultMode = false])

Returns a Promise with the validation results.

openApiDoc

Type: Object An object that represents an OpenAPI document.

defaultMode

Type: boolean Default: false If set to true, the validator will ignore the .validaterc file and will use the configuration defaults.

Validation results

The Promise returned from the validator resolves into a JSON object. The structure of the object is:

{
  errors:
  [
    {
      path: 'path.to.error.in.object'
      message: 'Major problem in the OpenAPI document.'
    }
  ],
  warnings:
  [
    {
      path: 'path.to.warning.in.object'
      message: 'Minor problem in the OpenAPI document.'
    }
  ]
}

The object will always have errors and warnings keys that map to arrays. If an array is empty, that means there were no errors/warnings in the OpenAPI document.

Configuration

Use these sections to customize your traditional and Spectral rules to reflect your API standards.

Setup

To set up the configuration capability, simply run the command lint-openapi init. This will create (or overwrite) a .validaterc file with all rules set to their default value. This command does not create a .validateignore. That file must be created manually. These rules can then be changed to configure the validator. Continue reading for more details.

WARNING: If a .validaterc file already exists and has been customized, this command will reset all rules to their default values.

It is recommended to place these files in the root directory of your project. The code will recursively search up the filesystem for these files from wherever the validator is being run. Wherever in the file system the validator is being run, the nearest versions of these files will be used.

Definitions

Specs

The validator supports two API definition specifications - OpenAPI 2.0, aka Swagger 2.0, and OpenAPI 3.0. The validator will automatically determine which spec a document is written in. There are some rules in the validator that only apply to one of the specs and some rules that apply to both. The configuration structure is organized by these "specs". The supported specs are described below:

Categories

Rules are further organized by categories. Not every category is supported in every spec - these are a superset of the available categories. For the actual structure, see the default values. The supported categories are described below:

Rules

Each category contains a group of rules. The spec that each rule applies to is marked in the third column. For the actual configuration structure, see the default values. You can use the -v option of the CLI validator to display the rule for each reported result. The supported rules are described below:

operations

pagination

parameters

paths

responses

schemas

security_definitions

security

walker

Statuses

Each rule can be assigned a status. The supported statuses are error, warning, info, hint and off. Some rules can be configured further with configuration options. The format of this configuration is to provide an array, rather than just a string. e.g. "param_name_case_convention": ["error", "lower_camel_case"] If just a string is provided for these rule, the default configuration option will be used. If only one value is provided in the array, it MUST be a status. The default configuration option will be used in this case as well. The rules that support configuration options will have two values in the defaults table.

Configuration Options

For rules that accept additional configuration, there will be a limited set of available options.

Case Convention Options

• Some rules check strings for adherence to a specific case convention. In some cases, the case convention checked is configurable.
• Rules with configurable case conventions will end in _case_convention, such as param_name_case_convention.

Configuration file

Non-Spectral configurations are defined in a file, titled .validaterc. Spectral rules must be configured in .spectral.yaml rather than in .validaterc. Additionally, certain files can be ignored by the validator. Any glob placed in a file called .validateignore will always be ignored by the validator at runtime. This is set up like a .gitignore or a .eslintignore file.

The configuration file must be structured as a JSON object with specs as first-level keys, categories as second-level keys, rules as third-level keys, and statuses as values for the 'rules' objects.

If a rule is not included in the file, that rule will be set to the default status automatically. See the Default Values for more info.

For an example of the structure, see the defaults file.

The easiest way to create a .validaterc file is using the initialization command.

Default mode

The validator has a set of predefined default statuses for each rule that are used in 'default mode'.

Default mode can be turned on with the command line option -d. If this option is given, the .validaterc file will be ignored.

If a .validaterc file does not exist at the root directory of your project, the validator will automatically run in default mode.

The default values for each rule are described below.

Default values

swagger2

operations

oas3

operations

parameters

responses

schemas

shared

operations

pagination

parameters

paths

responses

security_definitions

security

schemas

walker

Spectral configuration

Currently the validator configures Spectral to check the following rules from its “oas" ruleset:

no-eval-in-markdown: true
no-script-tags-in-markdown: true
openapi-tags: true
operation-description: true
operation-operationId-unique: true
operation-parameters: true
operation-tags: true
operation-tag-defined: true
path-keys-no-trailing-slash: true
path-not-include-query: true
typed-enum: true
oas2-api-host: true
oas2-api-schemes: true
oas2-host-trailing-slash: true
oas2-valid-example: true
oas2-valid-definition-example: true
oas2-anyOf: true
oas2-oneOf: true
oas2-operation-formData-consume-check: true
oas3-api-servers: true
oas3-examples-value-or-externalValue: true
oas3-server-trailing-slash: true
oas3-valid-example: true
oas3-valid-schema-example: true

This ruleset has the alias ibm:oas, and you can "extend" this ruleset or specify your own custom ruleset with a Spectral ruleset file. Note that all of the rules in the spectral:oas ruleset are defined in ibm:oas but only the rules listed above are enabled by default.

The ibm:oas ruleset also includes custom Spectral rules, documented here. These are configurable rules in the ibm:oas ruleset that are not part of the spectral:oas ruleset.

You can provide a Spectral ruleset file to the IBM OpenAPI validator in a file named .spectral.yaml in the current directory or with the --ruleset command line option of the validator.

Changing Spectral rule severity

Any rule in the ibm:oas ruleset can be configured to trigger an error, warning, info, or hint message in the validator output. For example, to configure the openapi-tags rule to trigger an info message instead of a warning, specify the following in your Spectral ruleset file:

extends: ibm:oas
rules:
  openapi-tags: warn

To completely disable a rule, use the severity of off. For example, to disable the operation-tags rule, specify the following in your Spectral ruleset file:

extends: ibm:oas
rules:
  operation-tags: off

Since the ibm:oas ruleset includes all the rules in spectral:oas, you can also enable rules from that ruleset that are disabled by default in ibm:oas. For example, to enable the info-contact rule with it's default severity (warning), specify the following in your Spectral ruleset file:

extends: ibm:oas
rules:
  info-contact: true

You could also set the severity of info-contact explicitly to error, warn, info, or hint.

Custom Spectral rules

New custom rules can be specified using a simple but powerful yaml syntax or with custom Javascript functions. Use the documentation on Spectral custom rules in order to add these to your .spectral.yaml file.

Warnings Limit

You may impose a warning limit on your API definitions. If the number of warnings issued exceeds the warning limit, the exit code will be set to 1. If the Validator is part of your CI build, this will cause the build to fail.

To impose a warnings limit on a project, add a .thresholdrc to your project. It is recommended to add this file to the root of the project. The validator recursively searches up the filesystem from whichever directory the validator is invoked, and the nearest .thresholdrc will be used.

The format for the .thresholdrc file is a top-level JSON object with a "warnings" field (shown below).

{
  "warnings": 0
}

limits

Turning off update-notifier

This package uses update-notifier to alert users when new versions of the tool are available. To turn this feature off, follow these instructions from the package authors. It is recommended to keep this feature on to help stay up to date with the latest changes.

License

Copyright 2017 SmartBear Software

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.