139 lines
4.7 KiB
JavaScript
139 lines
4.7 KiB
JavaScript
/**
|
|
* @fileoverview Rule to flag use of comma operator
|
|
* @author Brandon Mills
|
|
*/
|
|
|
|
"use strict";
|
|
|
|
//------------------------------------------------------------------------------
|
|
// Requirements
|
|
//------------------------------------------------------------------------------
|
|
|
|
const astUtils = require("./utils/ast-utils");
|
|
|
|
//------------------------------------------------------------------------------
|
|
// Helpers
|
|
//------------------------------------------------------------------------------
|
|
|
|
const DEFAULT_OPTIONS = {
|
|
allowInParentheses: true
|
|
};
|
|
|
|
//------------------------------------------------------------------------------
|
|
// Rule Definition
|
|
//------------------------------------------------------------------------------
|
|
|
|
/** @type {import('../shared/types').Rule} */
|
|
module.exports = {
|
|
meta: {
|
|
type: "suggestion",
|
|
|
|
docs: {
|
|
description: "Disallow comma operators",
|
|
recommended: false,
|
|
url: "https://eslint.org/docs/latest/rules/no-sequences"
|
|
},
|
|
|
|
schema: [{
|
|
properties: {
|
|
allowInParentheses: {
|
|
type: "boolean",
|
|
default: true
|
|
}
|
|
},
|
|
additionalProperties: false
|
|
}],
|
|
|
|
messages: {
|
|
unexpectedCommaExpression: "Unexpected use of comma operator."
|
|
}
|
|
},
|
|
|
|
create(context) {
|
|
const options = Object.assign({}, DEFAULT_OPTIONS, context.options[0]);
|
|
const sourceCode = context.sourceCode;
|
|
|
|
/**
|
|
* Parts of the grammar that are required to have parens.
|
|
*/
|
|
const parenthesized = {
|
|
DoWhileStatement: "test",
|
|
IfStatement: "test",
|
|
SwitchStatement: "discriminant",
|
|
WhileStatement: "test",
|
|
WithStatement: "object",
|
|
ArrowFunctionExpression: "body"
|
|
|
|
/*
|
|
* Omitting CallExpression - commas are parsed as argument separators
|
|
* Omitting NewExpression - commas are parsed as argument separators
|
|
* Omitting ForInStatement - parts aren't individually parenthesised
|
|
* Omitting ForStatement - parts aren't individually parenthesised
|
|
*/
|
|
};
|
|
|
|
/**
|
|
* Determines whether a node is required by the grammar to be wrapped in
|
|
* parens, e.g. the test of an if statement.
|
|
* @param {ASTNode} node The AST node
|
|
* @returns {boolean} True if parens around node belong to parent node.
|
|
*/
|
|
function requiresExtraParens(node) {
|
|
return node.parent && parenthesized[node.parent.type] &&
|
|
node === node.parent[parenthesized[node.parent.type]];
|
|
}
|
|
|
|
/**
|
|
* Check if a node is wrapped in parens.
|
|
* @param {ASTNode} node The AST node
|
|
* @returns {boolean} True if the node has a paren on each side.
|
|
*/
|
|
function isParenthesised(node) {
|
|
return astUtils.isParenthesised(sourceCode, node);
|
|
}
|
|
|
|
/**
|
|
* Check if a node is wrapped in two levels of parens.
|
|
* @param {ASTNode} node The AST node
|
|
* @returns {boolean} True if two parens surround the node on each side.
|
|
*/
|
|
function isParenthesisedTwice(node) {
|
|
const previousToken = sourceCode.getTokenBefore(node, 1),
|
|
nextToken = sourceCode.getTokenAfter(node, 1);
|
|
|
|
return isParenthesised(node) && previousToken && nextToken &&
|
|
astUtils.isOpeningParenToken(previousToken) && previousToken.range[1] <= node.range[0] &&
|
|
astUtils.isClosingParenToken(nextToken) && nextToken.range[0] >= node.range[1];
|
|
}
|
|
|
|
return {
|
|
SequenceExpression(node) {
|
|
|
|
// Always allow sequences in for statement update
|
|
if (node.parent.type === "ForStatement" &&
|
|
(node === node.parent.init || node === node.parent.update)) {
|
|
return;
|
|
}
|
|
|
|
// Wrapping a sequence in extra parens indicates intent
|
|
if (options.allowInParentheses) {
|
|
if (requiresExtraParens(node)) {
|
|
if (isParenthesisedTwice(node)) {
|
|
return;
|
|
}
|
|
} else {
|
|
if (isParenthesised(node)) {
|
|
return;
|
|
}
|
|
}
|
|
}
|
|
|
|
const firstCommaToken = sourceCode.getTokenAfter(node.expressions[0], astUtils.isCommaToken);
|
|
|
|
context.report({ node, loc: firstCommaToken.loc, messageId: "unexpectedCommaExpression" });
|
|
}
|
|
};
|
|
|
|
}
|
|
};
|