lib/rules/multiline-comment-style.js

/**
 * @fileoverview enforce a particular style for multiline comments
 * @author Teddy Katz
 */
"use strict";

const astUtils = require("./utils/ast-utils");

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

module.exports = {
    meta: {
        type: "suggestion",

        docs: {
            description: "enforce a particular style for multiline comments",
            category: "Stylistic Issues",
            recommended: false,
            url: "https://eslint.org/docs/rules/multiline-comment-style"
        },

        fixable: "whitespace",
        schema: [{ enum: ["starred-block", "separate-lines", "bare-block"] }],
        messages: {
            expectedBlock: "Expected a block comment instead of consecutive line comments.",
            expectedBareBlock: "Expected a block comment without padding stars.",
            startNewline: "Expected a linebreak after '/*'.",
            endNewline: "Expected a linebreak before '*/'.",
            missingStar: "Expected a '*' at the start of this line.",
            alignment: "Expected this line to be aligned with the start of the comment.",
            expectedLines: "Expected multiple line comments instead of a block comment."
        }
    },

    create(context) {
        const sourceCode = context.getSourceCode();
        const option = context.options[0] || "starred-block";

        //----------------------------------------------------------------------
        // Helpers
        //----------------------------------------------------------------------

        /**
         * Checks if a comment line is starred.
         * @param {string} line A string representing a comment line.
         * @returns {boolean} Whether or not the comment line is starred.
         */
        function isStarredCommentLine(line) {
            return /^\s*\*/u.test(line);
        }

        /**
         * Checks if a comment group is in starred-block form.
         * @param {Token[]} commentGroup A group of comments, containing either multiple line comments or a single block comment.
         * @returns {boolean} Whether or not the comment group is in starred block form.
         */
        function isStarredBlockComment([firstComment]) {
            if (firstComment.type !== "Block") {
                return false;
            }

            const lines = firstComment.value.split(astUtils.LINEBREAK_MATCHER);

            // The first and last lines can only contain whitespace.
            return lines.length > 0 && lines.every((line, i) => (i === 0 || i === lines.length - 1 ? /^\s*$/u : /^\s*\*/u).test(line));
        }

        /**
         * Checks if a comment group is in JSDoc form.
         * @param {Token[]} commentGroup A group of comments, containing either multiple line comments or a single block comment.
         * @returns {boolean} Whether or not the comment group is in JSDoc form.
         */
        function isJSDocComment([firstComment]) {
            if (firstComment.type !== "Block") {
                return false;
            }

            const lines = firstComment.value.split(astUtils.LINEBREAK_MATCHER);

            return /^\*\s*$/u.test(lines[0]) &&
                lines.slice(1, -1).every(line => /^\s* /u.test(line)) &&
                /^\s*$/u.test(lines[lines.length - 1]);
        }

        /**
         * Processes a comment group that is currently in separate-line form, calculating the offset for each line.
         * @param {Token[]} commentGroup A group of comments containing multiple line comments.
         * @returns {string[]} An array of the processed lines.
         */
        function processSeparateLineComments(commentGroup) {
            const allLinesHaveLeadingSpace = commentGroup
                .map(({ value }) => value)
                .filter(line => line.trim().length)
                .every(line => line.startsWith(" "));

            return commentGroup.map(({ value }) => (allLinesHaveLeadingSpace ? value.replace(/^ /u, "") : value));
        }

        /**
         * Processes a comment group that is currently in starred-block form, calculating the offset for each line.
         * @param {Token} comment A single block comment token in starred-block form.
         * @returns {string[]} An array of the processed lines.
         */
        function processStarredBlockComment(comment) {
            const lines = comment.value.split(astUtils.LINEBREAK_MATCHER)
                .filter((line, i, linesArr) => !(i === 0 || i === linesArr.length - 1))
                .map(line => line.replace(/^\s*$/u, ""));
            const allLinesHaveLeadingSpace = lines
                .map(line => line.replace(/\s*\*/u, ""))
                .filter(line => line.trim().length)
                .every(line => line.startsWith(" "));

            return lines.map(line => line.replace(allLinesHaveLeadingSpace ? /\s*\* ?/u : /\s*\*/u, ""));
        }

        /**
         * Processes a comment group that is currently in bare-block form, calculating the offset for each line.
         * @param {Token} comment A single block comment token in bare-block form.
         * @returns {string[]} An array of the processed lines.
         */
        function processBareBlockComment(comment) {
            const lines = comment.value.split(astUtils.LINEBREAK_MATCHER).map(line => line.replace(/^\s*$/u, ""));
            const leadingWhitespace = `${sourceCode.text.slice(comment.range[0] - comment.loc.start.column, comment.range[0])}   `;
            let offset = "";

            /*
             * Calculate the offset of the least indented line and use that as the basis for offsetting all the lines.
             * The first line should not be checked because it is inline with the opening block comment delimiter.
             */
            for (const [i, line] of lines.entries()) {
                if (!line.trim().length || i === 0) {
                    continue;
                }

                const [, lineOffset] = line.match(/^(\s*\*?\s*)/u);

                if (lineOffset.length < leadingWhitespace.length) {
                    const newOffset = leadingWhitespace.slice(lineOffset.length - leadingWhitespace.length);

                    if (newOffset.length > offset.length) {
                        offset = newOffset;
                    }
                }
            }

            return lines.map(line => {
                const match = line.match(/^(\s*\*?\s*)(.*)/u);
                const [, lineOffset, lineContents] = match;

                if (lineOffset.length > leadingWhitespace.length) {
                    return `${lineOffset.slice(leadingWhitespace.length - (offset.length + lineOffset.length))}${lineContents}`;
                }

                if (lineOffset.length < leadingWhitespace.length) {
                    return `${lineOffset.slice(leadingWhitespace.length)}${lineContents}`;
                }

                return lineContents;
            });
        }

        /**
         * Gets a list of comment lines in a group, formatting leading whitespace as necessary.
         * @param {Token[]} commentGroup A group of comments containing either multiple line comments or a single block comment.
         * @returns {string[]} A list of comment lines.
         */
        function getCommentLines(commentGroup) {
            const [firstComment] = commentGroup;

            if (firstComment.type === "Line") {
                return processSeparateLineComments(commentGroup);
            }

            if (isStarredBlockComment(commentGroup)) {
                return processStarredBlockComment(firstComment);
            }

            return processBareBlockComment(firstComment);
        }

        /**
         * Gets the initial offset (whitespace) from the beginning of a line to a given comment token.
         * @param {Token} comment The token to check.
         * @returns {string} The offset from the beginning of a line to the token.
         */
        function getInitialOffset(comment) {
            return sourceCode.text.slice(comment.range[0] - comment.loc.start.column, comment.range[0]);
        }

        /**
         * Converts a comment into starred-block form
         * @param {Token} firstComment The first comment of the group being converted
         * @param {string[]} commentLinesList A list of lines to appear in the new starred-block comment
         * @returns {string} A representation of the comment value in starred-block form, excluding start and end markers
         */
        function convertToStarredBlock(firstComment, commentLinesList) {
            const initialOffset = getInitialOffset(firstComment);

            return `/*\n${commentLinesList.map(line => `${initialOffset} * ${line}`).join("\n")}\n${initialOffset} */`;
        }

        /**
         * Converts a comment into separate-line form
         * @param {Token} firstComment The first comment of the group being converted
         * @param {string[]} commentLinesList A list of lines to appear in the new starred-block comment
         * @returns {string} A representation of the comment value in separate-line form
         */
        function convertToSeparateLines(firstComment, commentLinesList) {
            return commentLinesList.map(line => `// ${line}`).join(`\n${getInitialOffset(firstComment)}`);
        }

        /**
         * Converts a comment into bare-block form
         * @param {Token} firstComment The first comment of the group being converted
         * @param {string[]} commentLinesList A list of lines to appear in the new starred-block comment
         * @returns {string} A representation of the comment value in bare-block form
         */
        function convertToBlock(firstComment, commentLinesList) {
            return `/* ${commentLinesList.join(`\n${getInitialOffset(firstComment)}   `)} */`;
        }

        /**
         * Each method checks a group of comments to see if it's valid according to the given option.
         * @param {Token[]} commentGroup A list of comments that appear together. This will either contain a single
         * block comment or multiple line comments.
         * @returns {void}
         */
        const commentGroupCheckers = {
            "starred-block"(commentGroup) {
                const [firstComment] = commentGroup;
                const commentLines = getCommentLines(commentGroup);

                if (commentLines.some(value => value.includes("*/"))) {
                    return;
                }

                if (commentGroup.length > 1) {
                    context.report({
                        loc: {
                            start: firstComment.loc.start,
                            end: commentGroup[commentGroup.length - 1].loc.end
                        },
                        messageId: "expectedBlock",
                        fix(fixer) {
                            const range = [firstComment.range[0], commentGroup[commentGroup.length - 1].range[1]];

                            return commentLines.some(value => value.startsWith("/"))
                                ? null
                                : fixer.replaceTextRange(range, convertToStarredBlock(firstComment, commentLines));
                        }
                    });
                } else {
                    const lines = firstComment.value.split(astUtils.LINEBREAK_MATCHER);
                    const expectedLeadingWhitespace = getInitialOffset(firstComment);
                    const expectedLinePrefix = `${expectedLeadingWhitespace} *`;

                    if (!/^\*?\s*$/u.test(lines[0])) {
                        const start = firstComment.value.startsWith("*") ? firstComment.range[0] + 1 : firstComment.range[0];

                        context.report({
                            loc: {
                                start: firstComment.loc.start,
                                end: { line: firstComment.loc.start.line, column: firstComment.loc.start.column + 2 }
                            },
                            messageId: "startNewline",
                            fix: fixer => fixer.insertTextAfterRange([start, start + 2], `\n${expectedLinePrefix}`)
                        });
                    }

                    if (!/^\s*$/u.test(lines[lines.length - 1])) {
                        context.report({
                            loc: {
                                start: { line: firstComment.loc.end.line, column: firstComment.loc.end.column - 2 },
                                end: firstComment.loc.end
                            },
                            messageId: "endNewline",
                            fix: fixer => fixer.replaceTextRange([firstComment.range[1] - 2, firstComment.range[1]], `\n${expectedLinePrefix}/`)
                        });
                    }

                    for (let lineNumber = firstComment.loc.start.line + 1; lineNumber <= firstComment.loc.end.line; lineNumber++) {
                        const lineText = sourceCode.lines[lineNumber - 1];
                        const errorType = isStarredCommentLine(lineText)
                            ? "alignment"
                            : "missingStar";

                        if (!lineText.startsWith(expectedLinePrefix)) {
                            context.report({
                                loc: {
                                    start: { line: lineNumber, column: 0 },
                                    end: { line: lineNumber, column: lineText.length }
                                },
                                messageId: errorType,
                                fix(fixer) {
                                    const lineStartIndex = sourceCode.getIndexFromLoc({ line: lineNumber, column: 0 });

                                    if (errorType === "alignment") {
                                        const [, commentTextPrefix = ""] = lineText.match(/^(\s*\*)/u) || [];
                                        const commentTextStartIndex = lineStartIndex + commentTextPrefix.length;

                                        return fixer.replaceTextRange([lineStartIndex, commentTextStartIndex], expectedLinePrefix);
                                    }

                                    const [, commentTextPrefix = ""] = lineText.match(/^(\s*)/u) || [];
                                    const commentTextStartIndex = lineStartIndex + commentTextPrefix.length;
                                    let offset;

                                    for (const [idx, line] of lines.entries()) {
                                        if (!/\S+/u.test(line)) {
                                            continue;
                                        }

                                        const lineTextToAlignWith = sourceCode.lines[firstComment.loc.start.line - 1 + idx];
                                        const [, prefix = "", initialOffset = ""] = lineTextToAlignWith.match(/^(\s*(?:\/?\*)?(\s*))/u) || [];

                                        offset = `${commentTextPrefix.slice(prefix.length)}${initialOffset}`;

                                        if (/^\s*\//u.test(lineText) && offset.length === 0) {
                                            offset += " ";
                                        }
                                        break;
                                    }

                                    return fixer.replaceTextRange([lineStartIndex, commentTextStartIndex], `${expectedLinePrefix}${offset}`);
                                }
                            });
                        }
                    }
                }
            },
            "separate-lines"(commentGroup) {
                const [firstComment] = commentGroup;

                if (firstComment.type !== "Block" || isJSDocComment(commentGroup)) {
                    return;
                }

                const commentLines = getCommentLines(commentGroup);
                const tokenAfter = sourceCode.getTokenAfter(firstComment, { includeComments: true });

                if (tokenAfter && firstComment.loc.end.line === tokenAfter.loc.start.line) {
                    return;
                }

                context.report({
                    loc: {
                        start: firstComment.loc.start,
                        end: { line: firstComment.loc.start.line, column: firstComment.loc.start.column + 2 }
                    },
                    messageId: "expectedLines",
                    fix(fixer) {
                        return fixer.replaceText(firstComment, convertToSeparateLines(firstComment, commentLines));
                    }
                });
            },
            "bare-block"(commentGroup) {
                if (isJSDocComment(commentGroup)) {
                    return;
                }

                const [firstComment] = commentGroup;
                const commentLines = getCommentLines(commentGroup);

                // Disallows consecutive line comments in favor of using a block comment.
                if (firstComment.type === "Line" && commentLines.length > 1 &&
                    !commentLines.some(value => value.includes("*/"))) {
                    context.report({
                        loc: {
                            start: firstComment.loc.start,
                            end: commentGroup[commentGroup.length - 1].loc.end
                        },
                        messageId: "expectedBlock",
                        fix(fixer) {
                            return fixer.replaceTextRange(
                                [firstComment.range[0], commentGroup[commentGroup.length - 1].range[1]],
                                convertToBlock(firstComment, commentLines)
                            );
                        }
                    });
                }

                // Prohibits block comments from having a * at the beginning of each line.
                if (isStarredBlockComment(commentGroup)) {
                    context.report({
                        loc: {
                            start: firstComment.loc.start,
                            end: { line: firstComment.loc.start.line, column: firstComment.loc.start.column + 2 }
                        },
                        messageId: "expectedBareBlock",
                        fix(fixer) {
                            return fixer.replaceText(firstComment, convertToBlock(firstComment, commentLines));
                        }
                    });
                }
            }
        };

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

        return {
            Program() {
                return sourceCode.getAllComments()
                    .filter(comment => comment.type !== "Shebang")
                    .filter(comment => !astUtils.COMMENTS_IGNORE_PATTERN.test(comment.value))
                    .filter(comment => {
                        const tokenBefore = sourceCode.getTokenBefore(comment, { includeComments: true });

                        return !tokenBefore || tokenBefore.loc.end.line < comment.loc.start.line;
                    })
                    .reduce((commentGroups, comment, index, commentList) => {
                        const tokenBefore = sourceCode.getTokenBefore(comment, { includeComments: true });

                        if (
                            comment.type === "Line" &&
                            index && commentList[index - 1].type === "Line" &&
                            tokenBefore && tokenBefore.loc.end.line === comment.loc.start.line - 1 &&
                            tokenBefore === commentList[index - 1]
                        ) {
                            commentGroups[commentGroups.length - 1].push(comment);
                        } else {
                            commentGroups.push([comment]);
                        }

                        return commentGroups;
                    }, [])
                    .filter(commentGroup => !(commentGroup.length === 1 && commentGroup[0].loc.start.line === commentGroup[0].loc.end.line))
                    .forEach(commentGroupCheckers[option]);
            }
        };
    }
};