import shim from '@joplin/lib/shim'; const os = require('os'); import { readFile } from 'fs/promises'; const { filename } = require('@joplin/lib/path-utils'); import HtmlToMd from '@joplin/lib/HtmlToMd'; describe('HtmlToMd', () => { it('should convert from Html to Markdown', (async () => { const basePath = `${__dirname}/html_to_md`; const files = await shim.fsDriver().readDirStats(basePath); const htmlToMd = new HtmlToMd(); for (let i = 0; i < files.length; i++) { const htmlFilename = files[i].path; if (htmlFilename.indexOf('.html') < 0) continue; const htmlPath = `${basePath}/${htmlFilename}`; const mdPath = `${basePath}/${filename(htmlFilename)}.md`; // if (htmlFilename !== 'anchor_same_title_and_url.html') continue; // if (htmlFilename.indexOf('image_preserve_size') !== 0) continue; const htmlToMdOptions: any = {}; if (htmlFilename === 'anchor_local.html') { // Normally the list of anchor names in the document are retrieved from the HTML code // This is straightforward when the document is still in DOM format, as with the clipper, // but otherwise it would need to be somehow parsed out from the HTML. Here we just // hard code the anchors that we know are in the file. htmlToMdOptions.anchorNames = ['first', 'second', 'fourth']; } if (htmlFilename.indexOf('image_preserve_size') === 0) { htmlToMdOptions.preserveImageTagsWithSize = true; } if (htmlFilename.indexOf('preserve_nested_tables') === 0) { htmlToMdOptions.preserveNestedTables = true; } if (htmlFilename.indexOf('text_color') === 0) { htmlToMdOptions.preserveColorStyles = true; } const html = await readFile(htmlPath, 'utf8'); let expectedMd = await readFile(mdPath, 'utf8'); let actualMd = await htmlToMd.parse(`