node/tools/eslint/lib/rules/no-multi-spaces.js

/**
 * @fileoverview Disallow use of multiple spaces.
 * @author Nicholas C. Zakas
 */

"use strict";

//------------------------------------------------------------------------------
// Rule Definition
//------------------------------------------------------------------------------

module.exports = {
    meta: {
        docs: {
            description: "disallow multiple spaces",
            category: "Best Practices",
            recommended: false
        },

        fixable: "whitespace",

        schema: [
            {
                type: "object",
                properties: {
                    exceptions: {
                        type: "object",
                        patternProperties: {
                            "^([A-Z][a-z]*)+$": {
                                type: "boolean"
                            }
                        },
                        additionalProperties: false
                    }
                },
                additionalProperties: false
            }
        ]
    },

    create: function(context) {

        // the index of the last comment that was checked
        var exceptions = { Property: true },
            hasExceptions = true,
            options = context.options[0],
            lastCommentIndex = 0;

        if (options && options.exceptions) {
            Object.keys(options.exceptions).forEach(function(key) {
                if (options.exceptions[key]) {
                    exceptions[key] = true;
                } else {
                    delete exceptions[key];
                }
            });
            hasExceptions = Object.keys(exceptions).length > 0;
        }

        /**
         * Determines if a given source index is in a comment or not by checking
         * the index against the comment range. Since the check goes straight
         * through the file, once an index is passed a certain comment, we can
         * go to the next comment to check that.
         * @param {int} index The source index to check.
         * @param {ASTNode[]} comments An array of comment nodes.
         * @returns {boolean} True if the index is within a comment, false if not.
         * @private
         */
        function isIndexInComment(index, comments) {

            var comment;

            while (lastCommentIndex < comments.length) {

                comment = comments[lastCommentIndex];

                if (comment.range[0] <= index && index < comment.range[1]) {
                    return true;
                } else if (index > comment.range[1]) {
                    lastCommentIndex++;
                } else {
                    break;
                }

            }

            return false;
        }

        //--------------------------------------------------------------------------
        // Public
        //--------------------------------------------------------------------------

        return {
            Program: function() {

                var source = context.getSource(),
                    allComments = context.getAllComments(),
                    pattern = /[^\n\r\u2028\u2029\t ].? {2,}/g,  // note: repeating space
                    token,
                    previousToken,
                    parent;


                /**
                 * Creates a fix function that removes the multiple spaces between the two tokens
                 * @param {RuleFixer} leftToken left token
                 * @param {RuleFixer} rightToken right token
                 * @returns {function} fix function
                 * @private
                 */
                function createFix(leftToken, rightToken) {
                    return function(fixer) {
                        return fixer.replaceTextRange([leftToken.range[1], rightToken.range[0]], " ");
                    };
                }

                while (pattern.test(source)) {

                    // do not flag anything inside of comments
                    if (!isIndexInComment(pattern.lastIndex, allComments)) {

                        token = context.getTokenByRangeStart(pattern.lastIndex);
                        if (token) {
                            previousToken = context.getTokenBefore(token);

                            if (hasExceptions) {
                                parent = context.getNodeByRangeIndex(pattern.lastIndex - 1);
                            }

                            if (!parent || !exceptions[parent.type]) {
                                context.report({
                                    node: token,
                                    loc: token.loc.start,
                                    message: "Multiple spaces found before '{{value}}'.",
                                    data: { value: token.value },
                                    fix: createFix(previousToken, token)
                                });
                            }
                        }

                    }
                }
            }
        };

    }
};