Versions

Node.js API Reference

While ESLint is designed to be run on the command line, it’s possible to use ESLint programmatically through the Node.js API. The purpose of the Node.js API is to allow plugin and tool authors to use the ESLint functionality directly, without going through the command line interface.

Note: Use undocumented parts of the API at your own risk. Only those parts that are specifically mentioned in this document are approved for use and will remain stable and reliable. Anything left undocumented is unstable and may change or be removed at any point.

ESLint class

The ESLint class is the primary class to use in Node.js applications.

This class depends on the Node.js fs module and the file system, so you cannot use it in browsers. If you want to lint code on browsers, use the Linter class instead.

Here’s a simple example of using the ESLint class:

const { ESLint } = require("eslint");

(async function main() {
    // 1. Create an instance.
    const eslint = new ESLint();

    // 2. Lint files.
    const results = await eslint.lintFiles(["lib/**/*.js"]);

    // 3. Format the results.
    const formatter = await eslint.loadFormatter("stylish");
    const resultText = formatter.format(results);

    // 4. Output it.
    console.log(resultText);
})().catch((error) => {
    process.exitCode = 1;
    console.error(error);
});

Here’s an example that autofixes lint problems:

const { ESLint } = require("eslint");

(async function main() {
    // 1. Create an instance with the `fix` option.
    const eslint = new ESLint({ fix: true });

    // 2. Lint files. This doesn't modify target files.
    const results = await eslint.lintFiles(["lib/**/*.js"]);

    // 3. Modify the files with the fixed code.
    await ESLint.outputFixes(results);

    // 4. Format the results.
    const formatter = await eslint.loadFormatter("stylish");
    const resultText = formatter.format(results);

    // 5. Output it.
    console.log(resultText);
})().catch((error) => {
    process.exitCode = 1;
    console.error(error);
});

And here is an example of using the ESLint class with lintText API:

const { ESLint } = require("eslint");

const testCode = `
  const name = "eslint";
  if(true) {
    console.log("constant condition warning")
  };
`;

(async function main() {
    // 1. Create an instance
    const eslint = new ESLint({
        overrideConfigFile: true,
        overrideConfig: {
            languageOptions: {
                ecmaVersion: 2018,
                sourceType: "commonjs"
            }
        },
    });

    // 2. Lint text.
    const results = await eslint.lintText(testCode);

    // 3. Format the results.
    const formatter = await eslint.loadFormatter("stylish");
    const resultText = formatter.format(results);

    // 4. Output it.
    console.log(resultText);
})().catch((error) => {
    process.exitCode = 1;
    console.error(error);
});

◆ new ESLint(options)

const eslint = new ESLint(options);

Create a new ESLint instance.

Parameters

The ESLint constructor takes an options object. If you omit the options object then it uses default values for all options. The options object has the following properties.

File Enumeration
  • options.cwd (string)
    Default is process.cwd(). The working directory. This must be an absolute path.
  • options.errorOnUnmatchedPattern (boolean)
    Default is true. Unless set to false, the eslint.lintFiles() method will throw an error when no target files are found.
  • options.globInputPaths (boolean)
    Default is true. If false is present, the eslint.lintFiles() method doesn’t interpret glob patterns.
  • options.ignore (boolean)
    Default is true. If false is present, the eslint.lintFiles() method doesn’t respect ignorePatterns in your configuration.
  • options.ignorePatterns (string[] | null)
    Default is null. Ignore file patterns to use in addition to config ignores. These patterns are relative to cwd.
  • options.passOnNoPatterns (boolean)
    Default is false. When set to true, missing patterns cause the linting operation to short circuit and not report any failures.
  • options.warnIgnored (boolean)
    Default is true. Show warnings when the file list includes ignored files.
Linting
  • options.allowInlineConfig (boolean)
    Default is true. If false is present, ESLint suppresses directive comments in source code. If this option is false, it overrides the noInlineConfig setting in your configurations.
  • options.baseConfig (ConfigData | ConfigData[] | null)
    Default is null. Configuration object, extended by all configurations used with this instance. You can use this option to define the default settings that will be used if your configuration files don’t configure it.
  • options.overrideConfig (ConfigData | ConfigData[] | null)
    Default is null. Configuration object, overrides all configurations used with this instance. You can use this option to define the settings that will be used even if your configuration files configure it.
  • options.overrideConfigFile (string | boolean)
    Default is false. The path to a configuration file, overrides all configurations used with this instance. The options.overrideConfig option is applied after this option is applied.
  • options.plugins (Record<string, Plugin> | null)
    Default is null. The plugin implementations that ESLint uses for the plugins setting of your configuration. This is a map-like object. Those keys are plugin IDs and each value is implementation.
  • options.ruleFilter (({ruleId: string, severity: number}) => boolean)
    Default is () => true. A predicate function that filters rules to be run. This function is called with an object containing ruleId and severity, and returns true if the rule should be run.
  • options.stats (boolean)
    Default is false. When set to true, additional statistics are added to the lint results (see Stats type).
Autofix
  • options.fix (boolean | (message: LintMessage) => boolean)
    Default is false. If true is present, the eslint.lintFiles() and eslint.lintText() methods work in autofix mode. If a predicate function is present, the methods pass each lint message to the function, then use only the lint messages for which the function returned true.
  • options.fixTypes (("directive" | "problem" | "suggestion" | "layout")[] | null)
    Default is null. The types of the rules that the eslint.lintFiles() and eslint.lintText() methods use for autofix.
  • options.cache (boolean)
    Default is false. If true is present, the eslint.lintFiles() method caches lint results and uses it if each target file is not changed. Please mind that ESLint doesn’t clear the cache when you upgrade ESLint plugins. In that case, you have to remove the cache file manually. The eslint.lintText() method doesn’t use caches even if you pass the options.filePath to the method.
  • options.cacheLocation (string)
    Default is .eslintcache. The eslint.lintFiles() method writes caches into this file.
  • options.cacheStrategy (string)
    Default is "metadata". Strategy for the cache to use for detecting changed files. Can be either "metadata" or "content".
Other Options
  • options.flags (string[])
    Default is []. The feature flags to enable for this instance.

◆ eslint.lintFiles(patterns)

const results = await eslint.lintFiles(patterns);

This method lints the files that match the glob patterns and then returns the results.

Parameters

  • patterns (string | string[])
    The lint target files. This can contain any of file paths, directory paths, and glob patterns.

Return Value

  • (Promise<LintResult[]>)
    The promise that will be fulfilled with an array of LintResult objects.

◆ eslint.lintText(code, options)

const results = await eslint.lintText(code, options);

This method lints the given source code text and then returns the results.

By default, this method uses the configuration that applies to files in the current working directory (the cwd constructor option). If you want to use a different configuration, pass options.filePath, and ESLint will load the same configuration that eslint.lintFiles() would use for a file at options.filePath.

If the options.filePath value is configured to be ignored, this method returns an empty array. If the options.warnIgnored option is set along with the options.filePath option, this method returns a LintResult object. In that case, the result may contain a warning that indicates the file was ignored.

Parameters

The second parameter options is omittable.

  • code (string)
    The source code text to check.
  • options.filePath (string)
    Optional. The path to the file of the source code text. If omitted, the result.filePath becomes the string "<text>".
  • options.warnIgnored (boolean)
    Optional, defaults to options.warnIgnored passed to the constructor. If true is present and the options.filePath is a file ESLint should ignore, this method returns a lint result contains a warning message.

Return Value

  • (Promise<LintResult[]>)
    The promise that will be fulfilled with an array of LintResult objects. This is an array (despite there being only one lint result) in order to keep the interfaces between this and the eslint.lintFiles() method similar.

◆ eslint.getRulesMetaForResults(results)

const results = await eslint.lintFiles(patterns);
const rulesMeta = eslint.getRulesMetaForResults(results);

This method returns an object containing meta information for each rule that triggered a lint error in the given results.

Parameters

  • results (LintResult[])
    An array of LintResult objects returned from a call to ESLint#lintFiles() or ESLint#lintText().

Return Value

  • (Object)
    An object whose property names are the rule IDs from the results and whose property values are the rule’s meta information (if available).

◆ eslint.calculateConfigForFile(filePath)

const config = await eslint.calculateConfigForFile(filePath);

This method calculates the configuration for a given file, which can be useful for debugging purposes.

Parameters

  • filePath (string)
    The path to the file whose configuration you would like to calculate. Directory paths are forbidden because ESLint cannot handle the overrides setting.

Return Value

  • (Promise<Object>)
    The promise that will be fulfilled with a configuration object.

◆ eslint.isPathIgnored(filePath)

const isPathIgnored = await eslint.isPathIgnored(filePath);

This method checks if a given file is ignored by your configuration.

Parameters

  • filePath (string)
    The path to the file you want to check.

Return Value

  • (Promise<boolean>)
    The promise that will be fulfilled with whether the file is ignored or not. If the file is ignored, then it will return true.

◆ eslint.loadFormatter(nameOrPath)

const formatter = await eslint.loadFormatter(nameOrPath);

This method loads a formatter. Formatters convert lint results to a human- or machine-readable string.

Parameters

  • nameOrPath (string | undefined)
    The path to the file you want to check. The following values are allowed:
    • undefined. In this case, loads the "stylish" built-in formatter.
    • A name of built-in formatters.
    • A name of third-party formatters. For examples:
      • "foo" will load eslint-formatter-foo.
      • "@foo" will load @foo/eslint-formatter.
      • "@foo/bar" will load @foo/eslint-formatter-bar.
    • A path to the file that defines a formatter. The path must contain one or more path separators (/) in order to distinguish if it’s a path or not. For example, start with ./.

Return Value

  • (Promise<LoadedFormatter>)
    The promise that will be fulfilled with a LoadedFormatter object.

◆ eslint.hasFlag(flagName)

This method is used to determine if a given feature flag is set, as in this example:

if (eslint.hasFlag("x_feature")) {
    // handle flag
}

Parameters

  • flagName (string)
    The flag to check.

Return Value

  • (boolean)
    True if the flag is enabled.

◆ ESLint.version

const version = ESLint.version;

The version string of ESLint. E.g. "7.0.0".

This is a static property.

◆ ESLint.defaultConfig

const defaultConfig = ESLint.defaultConfig;

The default configuration that ESLint uses internally. This is provided for tooling that wants to calculate configurations using the same defaults as ESLint. Keep in mind that the default configuration may change from version to version, so you shouldn’t rely on any particular keys or values to be present.

This is a static property.

◆ ESLint.outputFixes(results)

await ESLint.outputFixes(results);

This method writes code modified by ESLint’s autofix feature into its respective file. If any of the modified files don’t exist, this method does nothing.

This is a static method.

Parameters

  • results (LintResult[])
    The LintResult objects to write.

Return Value

  • (Promise<void>)
    The promise that will be fulfilled after all files are written.

◆ ESLint.getErrorResults(results)

const filteredResults = ESLint.getErrorResults(results);

This method copies the given results and removes warnings. The returned value contains only errors.

This is a static method.

Parameters

  • results (LintResult[])
    The LintResult objects to filter.

Return Value

◆ LintResult type

The LintResult value is the information of the linting result of each file. The eslint.lintFiles() and eslint.lintText() methods return it. It has the following properties:

  • filePath (string)
    The absolute path to the file of this result. This is the string "<text>" if the file path is unknown (when you didn’t pass the options.filePath option to the eslint.lintText() method).
  • messages (LintMessage[])
    The array of LintMessage objects.
  • suppressedMessages (SuppressedLintMessage[])
    The array of SuppressedLintMessage objects.
  • fixableErrorCount (number)
    The number of errors that can be fixed automatically by the fix constructor option.
  • fixableWarningCount (number)
    The number of warnings that can be fixed automatically by the fix constructor option.
  • errorCount (number)
    The number of errors. This includes fixable errors and fatal errors.
  • fatalErrorCount (number)
    The number of fatal errors.
  • warningCount (number)
    The number of warnings. This includes fixable warnings.
  • output (string | undefined)
    The modified source code text. This property is undefined if any fixable messages didn’t exist.
  • source (string | undefined)
    The original source code text. This property is undefined if any messages didn’t exist or the output property exists.
  • stats (Stats | undefined)
    The Stats object. This contains the lint performance statistics collected with the stats option.
  • usedDeprecatedRules ({ ruleId: string; replacedBy: string[] }[])
    The information about the deprecated rules that were used to check this file.

◆ LintMessage type

The LintMessage value is the information of each linting error. The messages property of the LintResult type contains it. It has the following properties:

  • ruleId (string | null)
    The rule name that generates this lint message. If this message is generated by the ESLint core rather than rules, this is null.
  • severity (1 | 2)
    The severity of this message. 1 means warning and 2 means error.
  • fatal (boolean | undefined)
    true if this is a fatal error unrelated to a rule, like a parsing error.
  • message (string)
    The error message.
  • messageId (string | undefined)
    The message ID of the lint error. This property is undefined if the rule does not use message IDs.
  • line (number | undefined)
    The 1-based line number of the begin point of this message.
  • column (number | undefined)
    The 1-based column number of the begin point of this message.
  • endLine (number | undefined)
    The 1-based line number of the end point of this message. This property is undefined if this message is not a range.
  • endColumn (number | undefined)
    The 1-based column number of the end point of this message. This property is undefined if this message is not a range.
  • fix (EditInfo | undefined)
    The EditInfo object of autofix. This property is undefined if this message is not fixable.
  • suggestions ({ desc: string; fix: EditInfo; messageId?: string; data?: object }[] | undefined)
    The list of suggestions. Each suggestion is the pair of a description and an EditInfo object to fix code. API users such as editor integrations can choose one of them to fix the problem of this message. This property is undefined if this message doesn’t have any suggestions.

◆ SuppressedLintMessage type

The SuppressedLintMessage value is the information of each suppressed linting error. The suppressedMessages property of the LintResult type contains it. It has the following properties:

  • ruleId (string | null)
    Same as ruleId in LintMessage type.
  • severity (1 | 2)
    Same as severity in LintMessage type.
  • fatal (boolean | undefined)
    Same as fatal in LintMessage type.
  • message (string)
    Same as message in LintMessage type.
  • messageId (string | undefined)
    Same as messageId in LintMessage type.
  • line (number | undefined)
    Same as line in LintMessage type.
  • column (number | undefined)
    Same as column in LintMessage type.
  • endLine (number | undefined)
    Same as endLine in LintMessage type.
  • endColumn (number | undefined)
    Same as endColumn in LintMessage type.
  • fix (EditInfo | undefined)
    Same as fix in LintMessage type.
  • suggestions ({ desc: string; fix: EditInfo; messageId?: string; data?: object }[] | undefined)
    Same as suggestions in LintMessage type.
  • suppressions ({ kind: string; justification: string}[])
    The list of suppressions. Each suppression is the pair of a kind and a justification.

◆ EditInfo type

The EditInfo value is information to edit text. The fix and suggestions properties of LintMessage type contain it. It has following properties:

  • range ([number, number])
    The pair of 0-based indices in source code text to remove.
  • text (string)
    The text to add.

This edit information means replacing the range of the range property by the text property value. It’s like sourceCodeText.slice(0, edit.range[0]) + edit.text + sourceCodeText.slice(edit.range[1]). Therefore, it’s an add if the range[0] and range[1] property values are the same value, and it’s removal if the text property value is empty string.

◆ LoadedFormatter type

The LoadedFormatter value is the object to convert the LintResult objects to text. The eslint.loadFormatter() method returns it. It has the following method:

  • format ((results: LintResult[], resultsMeta?: ResultsMeta) => string | Promise<string>)
    The method to convert the LintResult objects to text. resultsMeta is an optional parameter that is primarily intended for use by the ESLint CLI and can contain only a maxWarningsExceeded property that would be passed through the context object when this method calls the underlying formatter function. Note that ESLint automatically generates cwd and rulesMeta properties of the context object, so you typically don’t need to pass in the second argument when calling this method.

loadESLint()

The loadESLint() function is used for integrations that wish to support both the current configuration system (flat config) and the old configuration system (eslintrc). This function returns the correct ESLint class implementation based on the arguments provided:

const { loadESLint } = require("eslint");

// loads the default ESLint that the CLI would use based on process.cwd()
const DefaultESLint = await loadESLint();

// loads the flat config version specifically
const FlatESLint = await loadESLint({ useFlatConfig: true });

// loads the legacy version specifically
const LegacyESLint = await loadESLint({ useFlatConfig: false });

You can then use the returned constructor to instantiate a new ESLint instance, like this:

// loads the default ESLint that the CLI would use based on process.cwd()
const DefaultESLint = await loadESLint();
const eslint = new DefaultESLint();

If you’re ever unsure which config system the returned constructor uses, check the configType property, which is either "flat" or "eslintrc":

// loads the default ESLint that the CLI would use based on process.cwd()
const DefaultESLint = await loadESLint();

if (DefaultESLint.configType === "flat") {
    // do something specific to flat config
}

If you don’t need to support both the old and new configuration systems, then it’s recommended to just use the ESLint constructor directly.


SourceCode

The SourceCode type represents the parsed source code that ESLint executes on. It’s used internally in ESLint and is also available so that already-parsed code can be used. You can create a new instance of SourceCode by passing in the text string representing the code and an abstract syntax tree (AST) in ESTree format (including location information, range information, comments, and tokens):

const SourceCode = require("eslint").SourceCode;

const code = new SourceCode("var foo = bar;", ast);

The SourceCode constructor throws an error if the AST is missing any of the required information.

The SourceCode constructor strips Unicode BOM. Please note the AST also should be parsed from stripped text.

const SourceCode = require("eslint").SourceCode;

const code = new SourceCode("\uFEFFvar foo = bar;", ast);

assert(code.hasBOM === true);
assert(code.text === "var foo = bar;");

SourceCode#splitLines()

This is a static function on SourceCode that is used to split the source code text into an array of lines.

const SourceCode = require("eslint").SourceCode;

const code = "var a = 1;\nvar b = 2;"

// split code into an array
const codeLines = SourceCode.splitLines(code);

/*
    Value of codeLines will be
    [
        "var a = 1;",
        "var b = 2;"
    ]
 */

Linter

The Linter object does the actual evaluation of the JavaScript code. It doesn’t do any filesystem operations, it simply parses and reports on the code. In particular, the Linter object does not process configuration files. Unless you are working in the browser, you probably want to use the ESLint class instead.

The Linter is a constructor, and you can create a new instance by passing in the options you want to use. The available options are:

  • cwd - Path to a directory that should be considered as the current working directory. It is accessible to rules from context.cwd or by calling context.getCwd() (see The Context Object). If cwd is undefined, it will be normalized to process.cwd() if the global process object is defined (for example, in the Node.js runtime) , or undefined otherwise.

For example:

const Linter = require("eslint").Linter;
const linter1 = new Linter({ cwd: 'path/to/project' });
const linter2 = new Linter();

In this example, rules run on linter1 will get path/to/project from context.cwd or when calling context.getCwd(). Those run on linter2 will get process.cwd() if the global process object is defined or undefined otherwise (e.g. on the browser https://eslint.org/demo).

Linter#verify

The most important method on Linter is verify(), which initiates linting of the given text. This method accepts three arguments:

  • code - the source code to lint (a string or instance of SourceCode).
  • config - a Configuration object or an array of configuration objects.
  • options - (optional) Additional options for this run.
    • filename - (optional) the filename to associate with the source code.
    • preprocess - (optional) A function that Processors in Plugins documentation describes as the preprocess method.
    • postprocess - (optional) A function that Processors in Plugins documentation describes as the postprocess method.
    • filterCodeBlock - (optional) A function that decides which code blocks the linter should adopt. The function receives two arguments. The first argument is the virtual filename of a code block. The second argument is the text of the code block. If the function returned true then the linter adopts the code block. If the function was omitted, the linter adopts only *.js code blocks. If you provided a filterCodeBlock function, it overrides this default behavior, so the linter doesn’t adopt *.js code blocks automatically.
    • disableFixes - (optional) when set to true, the linter doesn’t make either the fix or suggestions property of the lint result.
    • allowInlineConfig - (optional) set to false to disable inline comments from changing ESLint rules.
    • reportUnusedDisableDirectives - (optional) when set to true, adds reported errors for unused eslint-disable and eslint-enable directives when no problems would be reported in the disabled area anyway.
    • ruleFilter - (optional) A function predicate that decides which rules should run. It receives an object containing ruleId and severity, and returns true if the rule should be run.

If the third argument is a string, it is interpreted as the filename.

You can call verify() like this:

const Linter = require("eslint").Linter;
const linter = new Linter();

const messages = linter.verify("var foo;", {
    rules: {
        semi: 2
    }
}, { filename: "foo.js" });

// or using SourceCode

const Linter = require("eslint").Linter,
    linter = new Linter(),
    SourceCode = require("eslint").SourceCode;

const code = new SourceCode("var foo = bar;", ast);

const messages = linter.verify(code, {
    rules: {
        semi: 2
    }
}, { filename: "foo.js" });

The verify() method returns an array of objects containing information about the linting warnings and errors. Here’s an example:

[
    {
        fatal: false,
        ruleId: "semi",
        severity: 2,
        line: 1,
        column: 23,
        message: "Expected a semicolon.",
        fix: {
            range: [1, 15],
            text: ";"
        }
    }
]

The information available for each linting message is:

  • column - the column on which the error occurred.
  • fatal - usually omitted, but will be set to true if there’s a parsing error (not related to a rule).
  • line - the line on which the error occurred.
  • message - the message that should be output.
  • messageId - the ID of the message used to generate the message (this property is omitted if the rule does not use message IDs).
  • nodeType - (Deprecated: This property will be removed in a future version of ESLint.) the node or token type that was reported with the problem.
  • ruleId - the ID of the rule that triggered the messages (or null if fatal is true).
  • severity - either 1 or 2, depending on your configuration.
  • endColumn - the end column of the range on which the error occurred (this property is omitted if it’s not range).
  • endLine - the end line of the range on which the error occurred (this property is omitted if it’s not range).
  • fix - an object describing the fix for the problem (this property is omitted if no fix is available).
  • suggestions - an array of objects describing possible lint fixes for editors to programmatically enable (see details in the Working with Rules docs).

You can get the suppressed messages from the previous run by getSuppressedMessages() method. If there is not a previous run, getSuppressedMessage() will return an empty list.

const Linter = require("eslint").Linter;
const linter = new Linter();

const messages = linter.verify("var foo = bar; // eslint-disable-line -- Need to suppress", {
    rules: {
        semi: ["error", "never"]
    }
}, { filename: "foo.js" });
const suppressedMessages = linter.getSuppressedMessages();

console.log(suppressedMessages[0].suppressions); // [{ "kind": "directive", "justification": "Need to suppress" }]

You can also get an instance of the SourceCode object used inside of linter by using the getSourceCode() method:

const Linter = require("eslint").Linter;
const linter = new Linter();

const messages = linter.verify("var foo = bar;", {
    rules: {
        semi: 2
    }
}, { filename: "foo.js" });

const code = linter.getSourceCode();

console.log(code.text);     // "var foo = bar;"

In this way, you can retrieve the text and AST used for the last run of linter.verify().

Linter#verifyAndFix()

This method is similar to verify except that it also runs autofixing logic, similar to the --fix flag on the command line. The result object will contain the autofixed code, along with any remaining linting messages for the code that were not autofixed.

const Linter = require("eslint").Linter;
const linter = new Linter();

const messages = linter.verifyAndFix("var foo", {
    rules: {
        semi: 2
    }
});

Output object from this method:

{
    fixed: true,
    output: "var foo;",
    messages: []
}

The information available is:

  • fixed - True, if the code was fixed.
  • output - Fixed code text (might be the same as input if no fixes were applied).
  • messages - Collection of all messages for the given code (It has the same information as explained above under verify block).

Linter#version/Linter.version

Each instance of Linter has a version property containing the semantic version number of ESLint that the Linter instance is from.

const Linter = require("eslint").Linter;
const linter = new Linter();

linter.version; // => '9.0.0'

There is also a Linter.version property that you can read without instantiating Linter:

const Linter = require("eslint").Linter;

Linter.version; // => '9.0.0'

Linter#getTimes()

This method is used to get the times spent on (parsing, fixing, linting) a file. See times property of the Stats object.

Linter#getFixPassCount()

This method is used to get the number of autofix passes made. See fixPasses property of the Stats object.

Linter#hasFlag()

This method is used to determine if a given feature flag is set, as in this example:

const Linter = require("eslint").Linter;
const linter = new Linter({ flags: ["x_feature"] });

console.log(linter.hasFlag("x_feature")); // true

RuleTester

eslint.RuleTester is a utility to write tests for ESLint rules. It is used internally for the bundled rules that come with ESLint, and it can also be used by plugins.

Example usage:

"use strict";

const rule = require("../../../lib/rules/my-rule"),
    RuleTester = require("eslint").RuleTester;

const ruleTester = new RuleTester();

ruleTester.run("my-rule", rule, {
    valid: [
        {
            code: "var foo = true",
            options: [{ allowFoo: true }]
        }
    ],

    invalid: [
        {
            code: "var invalidVariable = true",
            errors: [{ message: "Unexpected invalid variable." }]
        },
        {
            code: "var invalidVariable = true",
            errors: [{ message: /^Unexpected.+variable/ }]
        }
    ]
});

The RuleTester constructor accepts an optional object argument, which can be used to specify defaults for your test cases. For example, if all of your test cases use ES2015, you can set it as a default:

const ruleTester = new RuleTester({ languageOptions: { ecmaVersion: 2015 } });

The RuleTester#run() method is used to run the tests. It should be passed the following arguments:

  • The name of the rule (string)
  • The rule object itself (see “working with rules”)
  • An object containing valid and invalid properties, each of which is an array containing test cases.

A test case is an object with the following properties:

  • name (string, optional): The name to use for the test case, to make it easier to find
  • code (string, required): The source code that the rule should be run on
  • options (array, optional): The options passed to the rule. The rule severity should not be included in this list.
  • before (function, optional): Function to execute before testing the case.
  • after (function, optional): Function to execute after testing the case regardless of its result.
  • filename (string, optional): The filename for the given case (useful for rules that make assertions about filenames).
  • only (boolean, optional): Run this case exclusively for debugging in supported test frameworks.

In addition to the properties above, invalid test cases can also have the following properties:

  • errors (number or array, required): Asserts some properties of the errors that the rule is expected to produce when run on this code. If this is a number, asserts the number of errors produced. Otherwise, this should be a list of objects, each containing information about a single reported error. The following properties can be used for an error (all are optional unless otherwise noted):

    • message (string/regexp): The message for the error. Must provide this or messageId
    • messageId (string): The ID for the error. Must provide this or message. See testing errors with messageId for details
    • data (object): Placeholder data which can be used in combination with messageId
    • type (string): (Deprecated: This property will be removed in a future version of ESLint.) The type of the reported AST node
    • line (number): The 1-based line number of the reported location
    • column (number): The 1-based column number of the reported location
    • endLine (number): The 1-based line number of the end of the reported location
    • endColumn (number): The 1-based column number of the end of the reported location
    • suggestions (array): An array of objects with suggestion details to check. Required if the rule produces suggestions. See Testing Suggestions for details

    If a string is provided as an error instead of an object, the string is used to assert the message of the error.

  • output (string, required if the rule fixes code): Asserts the output that will be produced when using this rule for a single pass of autofixing (e.g. with the --fix command line flag). If this is null or omitted, asserts that none of the reported problems suggest autofixes.

Any additional properties of a test case will be passed directly to the linter as config options. For example, a test case can have a languageOptions property to configure parser behavior:

{
    code: "let foo;",
    languageOptions: { ecmaVersion: 2015 }
}

If a valid test case only uses the code property, it can optionally be provided as a string containing the code, rather than an object with a code key.

Testing Errors with messageId

If the rule under test uses messageIds, you can use messageId property in a test case to assert reported error’s messageId instead of its message.

{
    code: "let foo;",
    errors: [{ messageId: "unexpected" }]
}

For messages with placeholders, a test case can also use data property to additionally assert reported error’s message.

{
    code: "let foo;",
    errors: [{ messageId: "unexpected", data: { name: "foo" } }]
}

Please note that data in a test case does not assert data passed to context.report. Instead, it is used to form the expected message text which is then compared with the received message.

Testing Fixes

The result of applying fixes can be tested by using the output property of an invalid test case. The output property should be used only when you expect a fix to be applied to the specified code; you can safely omit output if no changes are expected to the code. Here’s an example:

ruleTester.run("my-rule-for-no-foo", rule, {
    valid: [],
    invalid: [{
        code: "var foo;",
        output: "var bar;",
        errors: [{
            messageId: "shouldBeBar",
            line: 1,
            column: 5
        }]
    }]
})

A the end of this invalid test case, RuleTester expects a fix to be applied that results in the code changing from var foo; to var bar;. If the output after applying the fix doesn’t match, then the test fails.

Testing Suggestions

Suggestions can be tested by defining a suggestions key on an errors object. If this is a number, it asserts the number of suggestions provided for the error. Otherwise, this should be an array of objects, each containing information about a single provided suggestion. The following properties can be used:

  • desc (string): The suggestion desc value. Must provide this or messageId
  • messageId (string): The suggestion messageId value for suggestions that use messageIds. Must provide this or desc
  • data (object): Placeholder data which can be used in combination with messageId.
  • output (string, required): A code string representing the result of applying the suggestion fix to the input code

Example:

ruleTester.run("my-rule-for-no-foo", rule, {
    valid: [],
    invalid: [{
        code: "var foo;",
        errors: [{
            suggestions: [{
                desc: "Rename identifier 'foo' to 'bar'",
                output: "var bar;"
            }]
        }]
    }]
})

messageId and data properties in suggestion test objects work the same way as in error test objects. See testing errors with messageId for details.

ruleTester.run("my-rule-for-no-foo", rule, {
    valid: [],
    invalid: [{
        code: "var foo;",
        errors: [{
            suggestions: [{
                messageId: "renameFoo",
                data: { newName: "bar" },
                output: "var bar;"
            }]
        }]
    }]
})

Customizing RuleTester

RuleTester depends on two functions to run tests: describe and it. These functions can come from various places:

  1. If RuleTester.describe and RuleTester.it have been set to function values, RuleTester will use RuleTester.describe and RuleTester.it to run tests. You can use this to customize the behavior of RuleTester to match a test framework that you’re using.

    If RuleTester.itOnly has been set to a function value, RuleTester will call RuleTester.itOnly instead of RuleTester.it to run cases with only: true. If RuleTester.itOnly is not set but RuleTester.it has an only function property, RuleTester will fall back to RuleTester.it.only.

  2. Otherwise, if describe and it are present as globals, RuleTester will use globalThis.describe and globalThis.it to run tests and globalThis.it.only to run cases with only: true. This allows RuleTester to work when using frameworks like Mocha without any additional configuration.

  3. Otherwise, RuleTester#run will simply execute all of the tests in sequence, and will throw an error if one of them fails. This means you can simply execute a test file that calls RuleTester.run using Node.js, without needing a testing framework.

RuleTester#run calls the describe function with two arguments: a string describing the rule, and a callback function. The callback calls the it function with a string describing the test case, and a test function. The test function will return successfully if the test passes, and throw an error if the test fails. The signature for only is the same as it. RuleTester calls either it or only for every case even when some cases have only: true, and the test framework is responsible for implementing test case exclusivity. (Note that this is the standard behavior for test suites when using frameworks like Mocha; this information is only relevant if you plan to customize RuleTester.describe, RuleTester.it, or RuleTester.itOnly.)

Example of customizing RuleTester:

"use strict";

const RuleTester = require("eslint").RuleTester,
    test = require("my-test-runner"),
    myRule = require("../../../lib/rules/my-rule");

RuleTester.describe = function(text, method) {
    RuleTester.it.title = text;
    return method.call(this);
};

RuleTester.it = function(text, method) {
    test(RuleTester.it.title + ": " + text, method);
};

// then use RuleTester as documented

const ruleTester = new RuleTester();

ruleTester.run("my-rule", myRule, {
    valid: [
        // valid test cases
    ],
    invalid: [
        // invalid test cases
    ]
})

Change Language