static-cms/packages/netlify-cms-widget-markdown/src/serializers/__tests__/commonmark.spec.js

import { flow } from 'lodash';
import { tests as commonmarkSpec } from 'commonmark-spec';
import commonmark from 'commonmark';
import { markdownToSlate, slateToMarkdown } from '../index.js';

const skips = [
  {
    number: [456],
    reason: 'Remark ¯\\_(ツ)_/¯',
  },
  {
    number: [416, 417, 424, 425, 426, 431, 457, 460, 462, 464, 467],
    reason: 'Remark does not support infinite (redundant) nested marks',
  },
  {
    number: [455, 469, 470, 471],
    reason: 'Remark parses the initial set of identical nested delimiters first',
  },
  {
    number: [473, 476, 478, 480],
    reason: 'we convert underscores to asterisks for strong/emphasis',
  },
  { number: 490, reason: 'Remark strips pointy enclosing pointy brackets from link url' },
  { number: 503, reason: 'Remark allows non-breaking space between link url and title' },
  { number: 507, reason: 'Remark allows a space between link alt and url' },
  {
    number: [
      511,
      516,
      525,
      528,
      529,
      530,
      532,
      533,
      534,
      540,
      541,
      542,
      543,
      546,
      548,
      560,
      565,
      567,
    ],
    reason: 'we convert link references to standard links, but Remark also fails these',
  },
  {
    number: [569, 570, 571, 572, 573, 581, 585],
    reason: 'Remark does not recognize or remove marks in image alt text',
  },
  { number: 589, reason: 'Remark does not honor backslash escape of image exclamation point' },
  { number: 593, reason: 'Remark removes "mailto:" from autolink text' },
  { number: 599, reason: 'Remark does not escape all expected entities' },
  { number: 602, reason: 'Remark allows autolink emails to contain backslashes' },
];

const onlys = [
  // just add the spec number, eg:
  // 431,
];

/**
 * Each test receives input markdown and output html as expected for Commonmark
 * compliance. To test all of our handling in one go, we serialize the markdown
 * into our Slate AST, then back to raw markdown, and finally to HTML.
 */
const reader = new commonmark.Parser();
const writer = new commonmark.HtmlRenderer();
const parseWithCommonmark = markdown => {
  const parsed = reader.parse(markdown);
  return writer.render(parsed);
};
const parse = flow([markdownToSlate, slateToMarkdown]);

/**
 * Passing this test suite requires 100% Commonmark compliance. There are 624
 * tests, of which we're passing about 300 as of introduction of this suite. To
 * work on improving Commonmark support, update __fixtures__/commonmarkExpected.json
 */
describe.skip('Commonmark support', function() {
  const specs =
    onlys.length > 0
      ? commonmarkSpec.filter(({ number }) => onlys.includes(number))
      : commonmarkSpec;
  specs.forEach(spec => {
    const skip = skips.find(({ number }) => {
      return Array.isArray(number) ? number.includes(spec.number) : number === spec.number;
    });
    const specUrl = `https://spec.commonmark.org/0.29/#example-${spec.number}`;
    const parsed = parse(spec.markdown);
    const commonmarkParsedHtml = parseWithCommonmark(parsed);
    const description = `
${spec.section}
${specUrl}

Spec:
${JSON.stringify(spec, null, 2)}

Markdown input:
${spec.markdown}

Markdown parsed through Slate/Remark and back to Markdown:
${parsed}

HTML output:
${commonmarkParsedHtml}

Expected HTML output:
${spec.html}
    `;
    if (skip) {
      const showMessage = Array.isArray(skip.number) ? skip.number[0] === spec.number : true;
      if (showMessage) {
        //console.log(`skipping spec ${skip.number}\n${skip.reason}\n${specUrl}`);
      }
    }
    const testFn = skip ? test.skip : test;
    testFn(description, () => {
      expect(commonmarkParsedHtml).toEqual(spec.html);
    });
  });
});