/**
* @fileoverview Main Linter Class
* @author Gyandeep Singh
*/
"use strict";
//------------------------------------------------------------------------------
// Requirements
//------------------------------------------------------------------------------
const eslintScope = require("eslint-scope"),
levn = require("levn"),
lodash = require("lodash"),
blankScriptAST = require("../conf/blank-script.json"),
defaultConfig = require("../conf/default-config-options.js"),
CodePathAnalyzer = require("./code-path-analysis/code-path-analyzer"),
ConfigOps = require("./config/config-ops"),
validator = require("./config/config-validator"),
Environments = require("./config/environments"),
applyDisableDirectives = require("./util/apply-disable-directives"),
createEmitter = require("./util/safe-emitter"),
NodeEventGenerator = require("./util/node-event-generator"),
SourceCode = require("./util/source-code"),
Traverser = require("./util/traverser"),
createReportTranslator = require("./report-translator"),
Rules = require("./rules"),
timing = require("./timing"),
astUtils = require("./ast-utils"),
pkg = require("../package.json"),
SourceCodeFixer = require("./util/source-code-fixer");
const debug = require("debug")("eslint:linter");
const MAX_AUTOFIX_PASSES = 10;
//------------------------------------------------------------------------------
// Typedefs
//------------------------------------------------------------------------------
/**
* The result of a parsing operation from parseForESLint()
* @typedef {Object} CustomParseResult
* @property {ASTNode} ast The ESTree AST Program node.
* @property {Object} services An object containing additional services related
* to the parser.
*/
//------------------------------------------------------------------------------
// Helpers
//------------------------------------------------------------------------------
/**
* Parses a list of "name:boolean_value" or/and "name" options divided by comma or
* whitespace.
* @param {string} string The string to parse.
* @param {Comment} comment The comment node which has the string.
* @returns {Object} Result map object of names and boolean values
*/
function parseBooleanConfig(string, comment) {
const items = {};
// Collapse whitespace around `:` and `,` to make parsing easier
string = string.replace(/\s*([:,])\s*/g, "$1");
string.split(/\s|,+/).forEach(name => {
if (!name) {
return;
}
const pos = name.indexOf(":");
let value;
if (pos !== -1) {
value = name.slice(pos + 1);
name = name.slice(0, pos);
}
items[name] = {
value: (value === "true"),
comment
};
});
return items;
}
/**
* Parses a JSON-like config.
* @param {string} string The string to parse.
* @param {Object} location Start line and column of comments for potential error message.
* @returns {({success: true, config: Object}|{success: false, error: Problem})} Result map object
*/
function parseJsonConfig(string, location) {
let items = {};
// Parses a JSON-like comment by the same way as parsing CLI option.
try {
items = levn.parse("Object", string) || {};
// Some tests say that it should ignore invalid comments such as `/*eslint no-alert:abc*/`.
// Also, commaless notations have invalid severity:
// "no-alert: 2 no-console: 2" --> {"no-alert": "2 no-console: 2"}
// Should ignore that case as well.
if (ConfigOps.isEverySeverityValid(items)) {
return {
success: true,
config: items
};
}
} catch (ex) {
// ignore to parse the string by a fallback.
}
/*
* Optionator cannot parse commaless notations.
* But we are supporting that. So this is a fallback for that.
*/
items = {};
string = string.replace(/([a-zA-Z0-9\-/]+):/g, "\"$1\":").replace(/(]|[0-9])\s+(?=")/, "$1,");
try {
items = JSON.parse(`{${string}}`);
} catch (ex) {
return {
success: false,
error: {
ruleId: null,
fatal: true,
severity: 2,
source: null,
message: `Failed to parse JSON from '${string}': ${ex.message}`,
line: location.start.line,
column: location.start.column + 1
}
};
}
return {
success: true,
config: items
};
}
/**
* Parses a config of values separated by comma.
* @param {string} string The string to parse.
* @returns {Object} Result map of values and true values
*/
function parseListConfig(string) {
const items = {};
// Collapse whitespace around ,
string = string.replace(/\s*,\s*/g, ",");
string.split(/,+/).forEach(name => {
name = name.trim();
if (!name) {
return;
}
items[name] = true;
});
return items;
}
/**
* Ensures that variables representing built-in properties of the Global Object,
* and any globals declared by special block comments, are present in the global
* scope.
* @param {Scope} globalScope The global scope.
* @param {Object} config The existing configuration data.
* @param {Environments} envContext Env context
* @returns {void}
*/
function addDeclaredGlobals(globalScope, config, envContext) {
const declaredGlobals = {},
exportedGlobals = {},
explicitGlobals = {},
builtin = envContext.get("builtin");
Object.assign(declaredGlobals, builtin);
Object.keys(config.env).filter(name => config.env[name]).forEach(name => {
const env = envContext.get(name),
environmentGlobals = env && env.globals;
if (environmentGlobals) {
Object.assign(declaredGlobals, environmentGlobals);
}
});
Object.assign(exportedGlobals, config.exported);
Object.assign(declaredGlobals, config.globals);
Object.assign(explicitGlobals, config.astGlobals);
Object.keys(declaredGlobals).forEach(name => {
let variable = globalScope.set.get(name);
if (!variable) {
variable = new eslintScope.Variable(name, globalScope);
variable.eslintExplicitGlobal = false;
globalScope.variables.push(variable);
globalScope.set.set(name, variable);
}
variable.writeable = declaredGlobals[name];
});
Object.keys(explicitGlobals).forEach(name => {
let variable = globalScope.set.get(name);
if (!variable) {
variable = new eslintScope.Variable(name, globalScope);
variable.eslintExplicitGlobal = true;
variable.eslintExplicitGlobalComment = explicitGlobals[name].comment;
globalScope.variables.push(variable);
globalScope.set.set(name, variable);
}
variable.writeable = explicitGlobals[name].value;
});
// mark all exported variables as such
Object.keys(exportedGlobals).forEach(name => {
const variable = globalScope.set.get(name);
if (variable) {
variable.eslintUsed = true;
}
});
/*
* "through" contains all references which definitions cannot be found.
* Since we augment the global scope using configuration, we need to update
* references and remove the ones that were added by configuration.
*/
globalScope.through = globalScope.through.filter(reference => {
const name = reference.identifier.name;
const variable = globalScope.set.get(name);
if (variable) {
/*
* Links the variable and the reference.
* And this reference is removed from `Scope#through`.
*/
reference.resolved = variable;
variable.references.push(reference);
return false;
}
return true;
});
}
/**
* Creates a collection of disable directives from a comment
* @param {("disable"|"enable"|"disable-line"|"disable-next-line")} type The type of directive comment
* @param {{line: number, column: number}} loc The 0-based location of the comment token
* @param {string} value The value after the directive in the comment
* comment specified no specific rules, so it applies to all rules (e.g. `eslint-disable`)
* @returns {{
* type: ("disable"|"enable"|"disable-line"|"disable-next-line"),
* line: number,
* column: number,
* ruleId: (string|null)
* }[]} Directives from the comment
*/
function createDisableDirectives(type, loc, value) {
const ruleIds = Object.keys(parseListConfig(value));
const directiveRules = ruleIds.length ? ruleIds : [null];
return directiveRules.map(ruleId => ({ type, line: loc.line, column: loc.column + 1, ruleId }));
}
/**
* Parses comments in file to extract file-specific config of rules, globals
* and environments and merges them with global config; also code blocks
* where reporting is disabled or enabled and merges them with reporting config.
* @param {string} filename The file being checked.
* @param {ASTNode} ast The top node of the AST.
* @param {Object} config The existing configuration data.
* @param {function(string): {create: Function}} ruleMapper A map from rule IDs to defined rules
* @returns {{
* config: Object,
* problems: Problem[],
* disableDirectives: {
* type: ("disable"|"enable"|"disable-line"|"disable-next-line"),
* line: number,
* column: number,
* ruleId: (string|null)
* }[]
* }} Modified config object, along with any problems encountered
* while parsing config comments
*/
function modifyConfigsFromComments(filename, ast, config, ruleMapper) {
const commentConfig = {
exported: {},
astGlobals: {},
rules: {},
env: {}
};
const commentRules = {};
const problems = [];
const disableDirectives = [];
ast.comments.filter(token => token.type !== "Shebang").forEach(comment => {
let value = comment.value.trim();
const match = /^(eslint(-\w+){0,3}|exported|globals?)(\s|$)/.exec(value);
if (match) {
value = value.slice(match.index + match[1].length);
if (comment.type === "Block") {
switch (match[1]) {
case "exported":
Object.assign(commentConfig.exported, parseBooleanConfig(value, comment));
break;
case "globals":
case "global":
Object.assign(commentConfig.astGlobals, parseBooleanConfig(value, comment));
break;
case "eslint-disable":
[].push.apply(disableDirectives, createDisableDirectives("disable", comment.loc.start, value));
break;
case "eslint-enable":
[].push.apply(disableDirectives, createDisableDirectives("enable", comment.loc.start, value));
break;
case "eslint": {
const parseResult = parseJsonConfig(value, comment.loc);
if (parseResult.success) {
Object.keys(parseResult.config).forEach(name => {
const ruleValue = parseResult.config[name];
validator.validateRuleOptions(ruleMapper(name), name, ruleValue, `${filename} line ${comment.loc.start.line}`);
commentRules[name] = ruleValue;
});
} else {
problems.push(parseResult.error);
}
break;
}
// no default
}
} else { // comment.type === "Line"
if (match[1] === "eslint-disable-line") {
[].push.apply(disableDirectives, createDisableDirectives("disable-line", comment.loc.start, value));
} else if (match[1] === "eslint-disable-next-line") {
[].push.apply(disableDirectives, createDisableDirectives("disable-next-line", comment.loc.start, value));
}
}
}
});
Object.assign(commentConfig.rules, commentRules);
return {
config: ConfigOps.merge(config, commentConfig),
problems,
disableDirectives
};
}
/**
* Normalize ECMAScript version from the initial config
* @param {number} ecmaVersion ECMAScript version from the initial config
* @param {boolean} isModule Whether the source type is module or not
* @returns {number} normalized ECMAScript version
*/
function normalizeEcmaVersion(ecmaVersion, isModule) {
// Need at least ES6 for modules
if (isModule && (!ecmaVersion || ecmaVersion < 6)) {
ecmaVersion = 6;
}
/*
* Calculate ECMAScript edition number from official year version starting with
* ES2015, which corresponds with ES6 (or a difference of 2009).
*/
if (ecmaVersion >= 2015) {
ecmaVersion -= 2009;
}
return ecmaVersion;
}
/**
* Process initial config to make it safe to extend by file comment config
* @param {Object} config Initial config
* @param {Environments} envContext Env context
* @returns {Object} Processed config
*/
function prepareConfig(config, envContext) {
config.globals = config.globals || {};
const copiedRules = {};
let parserOptions = {};
if (typeof config.rules === "object") {
Object.keys(config.rules).forEach(k => {
const rule = config.rules[k];
if (rule === null) {
throw new Error(`Invalid config for rule '${k}'.`);
}
if (Array.isArray(rule)) {
copiedRules[k] = rule.slice();
} else {
copiedRules[k] = rule;
}
});
}
// merge in environment parserOptions
if (typeof config.env === "object") {
Object.keys(config.env).forEach(envName => {
const env = envContext.get(envName);
if (config.env[envName] && env && env.parserOptions) {
parserOptions = ConfigOps.merge(parserOptions, env.parserOptions);
}
});
}
const preparedConfig = {
rules: copiedRules,
parser: config.parser || defaultConfig.parser,
globals: ConfigOps.merge(defaultConfig.globals, config.globals),
env: ConfigOps.merge(defaultConfig.env, config.env || {}),
settings: ConfigOps.merge(defaultConfig.settings, config.settings || {}),
parserOptions: ConfigOps.merge(parserOptions, config.parserOptions || {})
};
const isModule = preparedConfig.parserOptions.sourceType === "module";
if (isModule) {
// can't have global return inside of modules
preparedConfig.parserOptions.ecmaFeatures = Object.assign({}, preparedConfig.parserOptions.ecmaFeatures, { globalReturn: false });
}
preparedConfig.parserOptions.ecmaVersion = normalizeEcmaVersion(preparedConfig.parserOptions.ecmaVersion, isModule);
return preparedConfig;
}
const eslintEnvPattern = /\/\*\s*eslint-env\s(.+?)\*\//g;
/**
* Checks whether or not there is a comment which has "eslint-env *" in a given text.
* @param {string} text - A source code text to check.
* @returns {Object|null} A result of parseListConfig() with "eslint-env *" comment.
*/
function findEslintEnv(text) {
let match, retv;
eslintEnvPattern.lastIndex = 0;
while ((match = eslintEnvPattern.exec(text))) {
retv = Object.assign(retv || {}, parseListConfig(match[1]));
}
return retv;
}
/**
* Strips Unicode BOM from a given text.
*
* @param {string} text - A text to strip.
* @returns {string} The stripped text.
*/
function stripUnicodeBOM(text) {
/*
* Check Unicode BOM.
* In JavaScript, string data is stored as UTF-16, so BOM is 0xFEFF.
* http://www.ecma-international.org/ecma-262/6.0/#sec-unicode-format-control-characters
*/
if (text.charCodeAt(0) === 0xFEFF) {
return text.slice(1);
}
return text;
}
/**
* Get the options for a rule (not including severity), if any
* @param {Array|number} ruleConfig rule configuration
* @returns {Array} of rule options, empty Array if none
*/
function getRuleOptions(ruleConfig) {
if (Array.isArray(ruleConfig)) {
return ruleConfig.slice(1);
}
return [];
}
/**
* Parses text into an AST. Moved out here because the try-catch prevents
* optimization of functions, so it's best to keep the try-catch as isolated
* as possible
* @param {string} text The text to parse.
* @param {Object} providedParserOptions Options to pass to the parser
* @param {Object} parser The parser module
* @param {string} filePath The path to the file being parsed.
* @returns {{success: false, error: Problem}|{success: true,ast: ASTNode, services: Object}}
* An object containing the AST and parser services if parsing was successful, or the error if parsing failed
* @private
*/
function parse(text, providedParserOptions, parser, filePath) {
const parserOptions = Object.assign({}, providedParserOptions, {
loc: true,
range: true,
raw: true,
tokens: true,
comment: true,
filePath
});
/*
* Check for parsing errors first. If there's a parsing error, nothing
* else can happen. However, a parsing error does not throw an error
* from this method - it's just considered a fatal error message, a
* problem that ESLint identified just like any other.
*/
try {
if (typeof parser.parseForESLint === "function") {
const parseResult = parser.parseForESLint(text, parserOptions);
return {
success: true,
ast: parseResult.ast,
services: parseResult.services || {}
};
}
return {
success: true,
ast: parser.parse(text, parserOptions),
services: {}
};
} catch (ex) {
// If the message includes a leading line number, strip it:
const message = `Parsing error: ${ex.message.replace(/^line \d+:/i, "").trim()}`;
const source = ex.lineNumber ? SourceCode.splitLines(text)[ex.lineNumber - 1] : null;
return {
success: false,
error: {
ruleId: null,
fatal: true,
severity: 2,
source,
message,
line: ex.lineNumber,
column: ex.column
}
};
}
}
/**
* Gets the scope for the current node
* @param {ScopeManager} scopeManager The scope manager for this AST
* @param {ASTNode} currentNode The node to get the scope of
* @param {number} ecmaVersion The `ecmaVersion` setting that this code was parsed with
* @returns {eslint-scope.Scope} The scope information for this node
*/
function getScope(scopeManager, currentNode, ecmaVersion) {
let initialNode;
// if current node introduces a scope, add it to the list
if (
["FunctionDeclaration", "FunctionExpression", "ArrowFunctionExpression"].indexOf(currentNode.type) >= 0 ||
ecmaVersion >= 6 && ["BlockStatement", "SwitchStatement", "CatchClause"].indexOf(currentNode.type) >= 0
) {
initialNode = currentNode;
} else {
initialNode = currentNode.parent;
}
// Ascend the current node's parents
for (let node = initialNode; node; node = node.parent) {
// Get the innermost scope
const scope = scopeManager.acquire(node, true);
if (scope) {
if (scope.type === "function-expression-name") {
return scope.childScopes[0];
}
return scope;
}
}
return scopeManager.scopes[0];
}
/**
* Marks a variable as used in the current scope
* @param {ScopeManager} scopeManager The scope manager for this AST. The scope may be mutated by this function.
* @param {ASTNode} currentNode The node currently being traversed
* @param {Object} parserOptions The options used to parse this text
* @param {string} name The name of the variable that should be marked as used.
* @returns {boolean} True if the variable was found and marked as used, false if not.
*/
function markVariableAsUsed(scopeManager, currentNode, parserOptions, name) {
const hasGlobalReturn = parserOptions.ecmaFeatures && parserOptions.ecmaFeatures.globalReturn;
const specialScope = hasGlobalReturn || parserOptions.sourceType === "module";
const currentScope = getScope(scopeManager, currentNode, parserOptions.ecmaVersion);
// Special Node.js scope means we need to start one level deeper
const initialScope = currentScope.type === "global" && specialScope ? currentScope.childScopes[0] : currentScope;
for (let scope = initialScope; scope; scope = scope.upper) {
const variable = scope.variables.find(scopeVar => scopeVar.name === name);
if (variable) {
variable.eslintUsed = true;
return true;
}
}
return false;
}
// methods that exist on SourceCode object
const DEPRECATED_SOURCECODE_PASSTHROUGHS = {
getSource: "getText",
getSourceLines: "getLines",
getAllComments: "getAllComments",
getNodeByRangeIndex: "getNodeByRangeIndex",
getComments: "getComments",
getCommentsBefore: "getCommentsBefore",
getCommentsAfter: "getCommentsAfter",
getCommentsInside: "getCommentsInside",
getJSDocComment: "getJSDocComment",
getFirstToken: "getFirstToken",
getFirstTokens: "getFirstTokens",
getLastToken: "getLastToken",
getLastTokens: "getLastTokens",
getTokenAfter: "getTokenAfter",
getTokenBefore: "getTokenBefore",
getTokenByRangeStart: "getTokenByRangeStart",
getTokens: "getTokens",
getTokensAfter: "getTokensAfter",
getTokensBefore: "getTokensBefore",
getTokensBetween: "getTokensBetween"
};
const BASE_TRAVERSAL_CONTEXT = Object.freeze(
Object.keys(DEPRECATED_SOURCECODE_PASSTHROUGHS).reduce(
(contextInfo, methodName) =>
Object.assign(contextInfo, {
[methodName]() {
const sourceCode = this.getSourceCode();
return sourceCode[DEPRECATED_SOURCECODE_PASSTHROUGHS[methodName]].apply(sourceCode, arguments);
}
}),
{}
)
);
const lastSourceCodes = new WeakMap();
//------------------------------------------------------------------------------
// Public Interface
//------------------------------------------------------------------------------
/**
* Object that is responsible for verifying JavaScript text
* @name eslint
*/
module.exports = class Linter {
constructor() {
lastSourceCodes.set(this, null);
this.version = pkg.version;
this.rules = new Rules();
this._parsers = new Map();
this.environments = new Environments();
}
/**
* Configuration object for the `verify` API. A JS representation of the eslintrc files.
* @typedef {Object} ESLintConfig
* @property {Object} rules The rule configuration to verify against.
* @property {string} [parser] Parser to use when generatig the AST.
* @property {Object} [parserOptions] Options for the parsed used.
* @property {Object} [settings] Global settings passed to each rule.
* @property {Object} [env] The environment to verify in.
* @property {Object} [globals] Available globals to the code.
*/
/**
* Same as linter.verify, except without support for processors.
* @param {string|SourceCode} textOrSourceCode The text to parse or a SourceCode object.
* @param {ESLintConfig} config An ESLintConfig instance to configure everything.
* @param {(string|Object)} [filenameOrOptions] The optional filename of the file being checked.
* If this is not set, the filename will default to '' in the rule context. If
* an object, then it has "filename", "saveState", and "allowInlineConfig" properties.
* @param {boolean} [filenameOrOptions.allowInlineConfig=true] Allow/disallow inline comments' ability to change config once it is set. Defaults to true if not supplied.
* Useful if you want to validate JS without comments overriding rules.
* @param {boolean} [filenameOrOptions.reportUnusedDisableDirectives=false] Adds reported errors for unused
* eslint-disable directives
* @returns {Object[]} The results as an array of messages or null if no messages.
*/
_verifyWithoutProcessors(textOrSourceCode, config, filenameOrOptions) {
let text,
parserServices,
allowInlineConfig,
providedFilename,
reportUnusedDisableDirectives;
// evaluate arguments
if (typeof filenameOrOptions === "object") {
providedFilename = filenameOrOptions.filename;
allowInlineConfig = filenameOrOptions.allowInlineConfig;
reportUnusedDisableDirectives = filenameOrOptions.reportUnusedDisableDirectives;
} else {
providedFilename = filenameOrOptions;
}
const filename = typeof providedFilename === "string" ? providedFilename : "";
if (typeof textOrSourceCode === "string") {
lastSourceCodes.set(this, null);
text = textOrSourceCode;
} else {
lastSourceCodes.set(this, textOrSourceCode);
text = textOrSourceCode.text;
}
// search and apply "eslint-env *".
const envInFile = findEslintEnv(text);
config = Object.assign({}, config);
if (envInFile) {
if (config.env) {
config.env = Object.assign({}, config.env, envInFile);
} else {
config.env = envInFile;
}
}
// process initial config to make it safe to extend
config = prepareConfig(config, this.environments);
if (lastSourceCodes.get(this)) {
parserServices = {};
} else {
// there's no input, just exit here
if (text.trim().length === 0) {
lastSourceCodes.set(this, new SourceCode(text, blankScriptAST));
return [];
}
let parser;
try {
parser = this._parsers.get(config.parser) || require(config.parser);
} catch (ex) {
return [{
ruleId: null,
fatal: true,
severity: 2,
source: null,
message: ex.message,
line: 0,
column: 0
}];
}
const parseResult = parse(
stripUnicodeBOM(text).replace(astUtils.SHEBANG_MATCHER, (match, captured) => `//${captured}`),
config.parserOptions,
parser,
filename
);
if (!parseResult.success) {
return [parseResult.error];
}
parserServices = parseResult.services;
lastSourceCodes.set(this, new SourceCode(text, parseResult.ast));
}
const problems = [];
const sourceCode = lastSourceCodes.get(this);
let disableDirectives;
// parse global comments and modify config
if (allowInlineConfig !== false) {
const modifyConfigResult = modifyConfigsFromComments(filename, sourceCode.ast, config, ruleId => this.rules.get(ruleId));
config = modifyConfigResult.config;
modifyConfigResult.problems.forEach(problem => problems.push(problem));
disableDirectives = modifyConfigResult.disableDirectives;
} else {
disableDirectives = [];
}
const emitter = createEmitter();
const traverser = new Traverser();
const ecmaFeatures = config.parserOptions.ecmaFeatures || {};
const ecmaVersion = config.parserOptions.ecmaVersion || 5;
const scopeManager = eslintScope.analyze(sourceCode.ast, {
ignoreEval: true,
nodejsScope: ecmaFeatures.globalReturn,
impliedStrict: ecmaFeatures.impliedStrict,
ecmaVersion,
sourceType: config.parserOptions.sourceType || "script",
fallback: Traverser.getKeys
});
/*
* Create a frozen object with the ruleContext properties and methods that are shared by all rules.
* All rule contexts will inherit from this object. This avoids the performance penalty of copying all the
* properties once for each rule.
*/
const sharedTraversalContext = Object.freeze(
Object.assign(
Object.create(BASE_TRAVERSAL_CONTEXT),
{
getAncestors: () => traverser.parents(),
getDeclaredVariables: scopeManager.getDeclaredVariables.bind(scopeManager),
getFilename: () => filename,
getScope: () => getScope(scopeManager, traverser.current(), config.parserOptions.ecmaVersion),
getSourceCode: () => sourceCode,
markVariableAsUsed: name => markVariableAsUsed(scopeManager, traverser.current(), config.parserOptions, name),
parserOptions: config.parserOptions,
parserPath: config.parser,
parserServices,
settings: config.settings,
/**
* This is used to avoid breaking rules that used to monkeypatch the `Linter#report` method
* by using the `_linter` property on rule contexts.
*
* This should be removed in a major release after we create a better way to
* lint for unused disable comments.
* https://github.com/eslint/eslint/issues/9193
*/
_linter: {
report() {},
on: emitter.on
}
}
)
);
// enable appropriate rules
Object.keys(config.rules).forEach(ruleId => {
const severity = ConfigOps.getRuleSeverity(config.rules[ruleId]);
if (severity === 0) {
return;
}
const rule = this.rules.get(ruleId);
let reportTranslator = null;
const ruleContext = Object.freeze(
Object.assign(
Object.create(sharedTraversalContext),
{
id: ruleId,
options: getRuleOptions(config.rules[ruleId]),
report() {
/*
* Create a report translator lazily.
* In a vast majority of cases, any given rule reports zero errors on a given
* piece of code. Creating a translator lazily avoids the performance cost of
* creating a new translator function for each rule that usually doesn't get
* called.
*
* Using lazy report translators improves end-to-end performance by about 3%
* with Node 8.4.0.
*/
if (reportTranslator === null) {
reportTranslator = createReportTranslator({ ruleId, severity, sourceCode });
}
const problem = reportTranslator.apply(null, arguments);
if (problem.fix && rule.meta && !rule.meta.fixable) {
throw new Error("Fixable rules should export a `meta.fixable` property.");
}
problems.push(problem);
/*
* This is used to avoid breaking rules that used monkeypatch Linter, and relied on
* `linter.report` getting called with report info every time a rule reports a problem.
* To continue to support this, make sure that `context._linter.report` is called every
* time a problem is reported by a rule, even though `context._linter` is no longer a
* `Linter` instance.
*
* This should be removed in a major release after we create a better way to
* lint for unused disable comments.
* https://github.com/eslint/eslint/issues/9193
*/
sharedTraversalContext._linter.report( // eslint-disable-line no-underscore-dangle
problem.ruleId,
problem.severity,
{ loc: { start: { line: problem.line, column: problem.column - 1 } } },
problem.message
);
}
}
)
);
try {
const ruleListeners = rule.create(ruleContext);
// add all the selectors from the rule as listeners
Object.keys(ruleListeners).forEach(selector => {
emitter.on(
selector,
timing.enabled
? timing.time(ruleId, ruleListeners[selector])
: ruleListeners[selector]
);
});
} catch (ex) {
ex.message = `Error while loading rule '${ruleId}': ${ex.message}`;
throw ex;
}
});
// augment global scope with declared global variables
addDeclaredGlobals(scopeManager.scopes[0], config, this.environments);
const eventGenerator = new CodePathAnalyzer(new NodeEventGenerator(emitter));
/*
* Each node has a type property. Whenever a particular type of
* node is found, an event is fired. This allows any listeners to
* automatically be informed that this type of node has been found
* and react accordingly.
*/
traverser.traverse(sourceCode.ast, {
enter(node, parent) {
node.parent = parent;
eventGenerator.enterNode(node);
},
leave(node) {
eventGenerator.leaveNode(node);
}
});
return applyDisableDirectives({
directives: disableDirectives,
problems: problems.sort((problemA, problemB) => problemA.line - problemB.line || problemA.column - problemB.column),
reportUnusedDisableDirectives
});
}
/**
* Verifies the text against the rules specified by the second argument.
* @param {string|SourceCode} textOrSourceCode The text to parse or a SourceCode object.
* @param {ESLintConfig} config An ESLintConfig instance to configure everything.
* @param {(string|Object)} [filenameOrOptions] The optional filename of the file being checked.
* If this is not set, the filename will default to '' in the rule context. If
* an object, then it has "filename", "saveState", and "allowInlineConfig" properties.
* @param {boolean} [saveState] Indicates if the state from the last run should be saved.
* Mostly useful for testing purposes.
* @param {boolean} [filenameOrOptions.allowInlineConfig] Allow/disallow inline comments' ability to change config once it is set. Defaults to true if not supplied.
* Useful if you want to validate JS without comments overriding rules.
* @param {function(string): string[]} [filenameOrOptions.preprocess] preprocessor for source text. If provided,
* this should accept a string of source text, and return an array of code blocks to lint.
* @param {function(Array): Object[]} [filenameOrOptions.postprocess] postprocessor for report messages. If provided,
* this should accept an array of the message lists for each code block returned from the preprocessor,
* apply a mapping to the messages as appropriate, and return a one-dimensional array of messages
* @returns {Object[]} The results as an array of messages or null if no messages.
*/
verify(textOrSourceCode, config, filenameOrOptions) {
const preprocess = filenameOrOptions && filenameOrOptions.preprocess || (rawText => [rawText]);
const postprocess = filenameOrOptions && filenameOrOptions.postprocess || lodash.flatten;
return postprocess(
preprocess(textOrSourceCode).map(
textBlock => this._verifyWithoutProcessors(textBlock, config, filenameOrOptions)
)
);
}
/**
* Gets the SourceCode object representing the parsed source.
* @returns {SourceCode} The SourceCode object.
*/
getSourceCode() {
return lastSourceCodes.get(this);
}
/**
* Defines a new linting rule.
* @param {string} ruleId A unique rule identifier
* @param {Function} ruleModule Function from context to object mapping AST node types to event handlers
* @returns {void}
*/
defineRule(ruleId, ruleModule) {
this.rules.define(ruleId, ruleModule);
}
/**
* Defines many new linting rules.
* @param {Object} rulesToDefine map from unique rule identifier to rule
* @returns {void}
*/
defineRules(rulesToDefine) {
Object.getOwnPropertyNames(rulesToDefine).forEach(ruleId => {
this.defineRule(ruleId, rulesToDefine[ruleId]);
});
}
/**
* Gets an object with all loaded rules.
* @returns {Map} All loaded rules
*/
getRules() {
return this.rules.getAllLoadedRules();
}
/**
* Define a new parser module
* @param {any} parserId Name of the parser
* @param {any} parserModule The parser object
* @returns {void}
*/
defineParser(parserId, parserModule) {
this._parsers.set(parserId, parserModule);
}
/**
* Performs multiple autofix passes over the text until as many fixes as possible
* have been applied.
* @param {string} text The source text to apply fixes to.
* @param {Object} config The ESLint config object to use.
* @param {Object} options The ESLint options object to use.
* @param {string} options.filename The filename from which the text was read.
* @param {boolean} options.allowInlineConfig Flag indicating if inline comments
* should be allowed.
* @param {boolean|Function} options.fix Determines whether fixes should be applied
* @param {Function} options.preprocess preprocessor for source text. If provided, this should
* accept a string of source text, and return an array of code blocks to lint.
* @param {Function} options.postprocess postprocessor for report messages. If provided,
* this should accept an array of the message lists for each code block returned from the preprocessor,
* apply a mapping to the messages as appropriate, and return a one-dimensional array of messages
* @returns {Object} The result of the fix operation as returned from the
* SourceCodeFixer.
*/
verifyAndFix(text, config, options) {
let messages = [],
fixedResult,
fixed = false,
passNumber = 0;
const debugTextDescription = options && options.filename || `${text.slice(0, 10)}...`;
const shouldFix = options && typeof options.fix !== "undefined" ? options.fix : true;
/**
* This loop continues until one of the following is true:
*
* 1. No more fixes have been applied.
* 2. Ten passes have been made.
*
* That means anytime a fix is successfully applied, there will be another pass.
* Essentially, guaranteeing a minimum of two passes.
*/
do {
passNumber++;
debug(`Linting code for ${debugTextDescription} (pass ${passNumber})`);
messages = this.verify(text, config, options);
debug(`Generating fixed text for ${debugTextDescription} (pass ${passNumber})`);
fixedResult = SourceCodeFixer.applyFixes(text, messages, shouldFix);
/*
* stop if there are any syntax errors.
* 'fixedResult.output' is a empty string.
*/
if (messages.length === 1 && messages[0].fatal) {
break;
}
// keep track if any fixes were ever applied - important for return value
fixed = fixed || fixedResult.fixed;
// update to use the fixed output instead of the original text
text = fixedResult.output;
} while (
fixedResult.fixed &&
passNumber < MAX_AUTOFIX_PASSES
);
/*
* If the last result had fixes, we need to lint again to be sure we have
* the most up-to-date information.
*/
if (fixedResult.fixed) {
fixedResult.messages = this.verify(text, config, options);
}
// ensure the last result properly reflects if fixes were done
fixedResult.fixed = fixed;
fixedResult.output = text;
return fixedResult;
}
};