编辑器模块(笔记)初版
This commit is contained in:
1 parent
7e65825d3e
commit
ad987111a3
66 files changed
+17470
-283
No files matched your search
@@ -0,0 +1,296 @@
|
||||
/**
|
||||
* markdownExport 单测(Node 内置 test runner)。
|
||||
*
|
||||
* # 为什么用 `renderMarkdownFragmentRaw`
|
||||
*
|
||||
* DOMPurify 需要 `window`/`document`:Node 里 `isSupported === false` 且对象上**根本
|
||||
* 没有** `sanitize`,所以「消毒后的片段」在纯 Node 下无法断言(本文件里如实
|
||||
* `t.skip()`,不装 jsdom、更不为跑通测试降低消毒强度)。
|
||||
* 但「fence 语言识别、任务列表、表格、链接、图片」这些**渲染**行为是纯字符串逻辑,
|
||||
* 必须能脱离浏览器回归,所以走模块里明确标注为「仅供单测」的未消毒入口。
|
||||
*
|
||||
* # 为什么两个方向都断言
|
||||
*
|
||||
* 一边断言「未消毒的片段里确实带着 `onerror=` / `<script>`」(证明消毒不是可选项),
|
||||
* 一边断言「无 DOM 时消毒入口失败关闭,返回空串而不是原文」(证明降级方向是安全的那
|
||||
* 一侧)。有 DOM 的环境(浏览器/未来的 DOM shim)会自动改跑真正的消毒断言。
|
||||
*/
|
||||
import { test } from 'node:test'
|
||||
import assert from 'node:assert/strict'
|
||||
import DOMPurify from 'dompurify'
|
||||
import {
|
||||
markdownToPlainText,
|
||||
renderMarkdownFragment,
|
||||
renderMarkdownFragmentRaw,
|
||||
renderNoteHtml,
|
||||
stripFrontMatter
|
||||
} from './markdownExport.ts'
|
||||
|
||||
/** 当前环境能不能真消毒(浏览器/happy-dom 之类为 true,纯 Node 为 false) */
|
||||
const domAvailable = DOMPurify.isSupported && typeof DOMPurify.sanitize === 'function'
|
||||
|
||||
// ===== front-matter =====
|
||||
|
||||
test('stripFrontMatter:剥掉闭合的 front-matter', () => {
|
||||
const src = ['---', 'title: 周报', 'tags:', ' - 工作', '---', '# 正文', '', '内容'].join('\n')
|
||||
assert.equal(stripFrontMatter(src), '# 正文\n\n内容')
|
||||
})
|
||||
|
||||
test('stripFrontMatter:闭合行是末行(末尾没有换行)也能处理', () => {
|
||||
assert.equal(stripFrontMatter('---\ntitle: x\n---'), '')
|
||||
assert.equal(stripFrontMatter('---\ntitle: x\n---\n正文'), '正文')
|
||||
})
|
||||
|
||||
test('stripFrontMatter:未闭合时整篇当正文,一个字符都不吃', () => {
|
||||
const unclosed = '---\ntitle: x\n# 正文\n--- 少了点什么'
|
||||
assert.equal(stripFrontMatter(unclosed), unclosed)
|
||||
// 只有一行 `---` 时同样不认
|
||||
assert.equal(stripFrontMatter('---'), '---')
|
||||
})
|
||||
|
||||
test('stripFrontMatter:CRLF 与 UTF-8 BOM', () => {
|
||||
assert.equal(stripFrontMatter('---\r\ntitle: x\r\n---\r\n正文\r\n'), '正文\r\n')
|
||||
assert.equal(stripFrontMatter('\uFEFF---\r\ntitle: x\r\n---\r\n正文'), '正文')
|
||||
// 没有 front-matter 时也要剥 BOM:留着它 markdown-it 会渲染成不可见怪字符
|
||||
assert.equal(stripFrontMatter('\uFEFF# 正文'), '# 正文')
|
||||
})
|
||||
|
||||
test('stripFrontMatter:第一行不是整行 --- 时不认', () => {
|
||||
assert.equal(stripFrontMatter('--- x\na: 1\n---\n正文'), '--- x\na: 1\n---\n正文')
|
||||
assert.equal(stripFrontMatter(' ---\na: 1\n---\n正文'), ' ---\na: 1\n---\n正文')
|
||||
})
|
||||
|
||||
test('stripFrontMatter:只剥第一段,正文里的 --- 原样保留', () => {
|
||||
assert.equal(stripFrontMatter('---\na: 1\n---\n正文\n---\n更多'), '正文\n---\n更多')
|
||||
})
|
||||
|
||||
// ===== 纯文本摘要 =====
|
||||
|
||||
test('markdownToPlainText:去掉标记、保留文字、折叠空白', () => {
|
||||
const src = [
|
||||
'---',
|
||||
'title: 元数据不该出现',
|
||||
'---',
|
||||
'# 标题一',
|
||||
'',
|
||||
'这是 **粗体**、*斜体*、`行内代码`,还有 [链接文字](https://example.com/a) 和 。',
|
||||
'',
|
||||
'> 引用行',
|
||||
'',
|
||||
'- [ ] 待办',
|
||||
'- 普通项',
|
||||
'',
|
||||
'```js',
|
||||
'const a = 1',
|
||||
'```',
|
||||
'',
|
||||
'<p>html 段落</p>'
|
||||
].join('\n')
|
||||
const text = markdownToPlainText(src)
|
||||
assert.ok(!text.includes('元数据不该出现'), 'front-matter 必须被剥掉')
|
||||
for (const kept of ['标题一', '粗体', '斜体', '行内代码', '链接文字', '图片说明', '引用行', '待办', '普通项', 'const a = 1', 'html 段落']) {
|
||||
assert.ok(text.includes(kept), `摘要应保留「${kept}」:${text}`)
|
||||
}
|
||||
for (const dropped of ['**', '`', '](', 'https://example.com', 'images/x.png', '<p>', '#', '>']) {
|
||||
assert.ok(!text.includes(dropped), `摘要不该残留「${dropped}」:${text}`)
|
||||
}
|
||||
assert.ok(!/ {2}/.test(text), `空白应折叠成单空格:${text}`)
|
||||
})
|
||||
|
||||
test('markdownToPlainText:默认 120 字符截断并加省略号', () => {
|
||||
const text = markdownToPlainText(`# ${'x'.repeat(200)}`)
|
||||
assert.equal(text.length, 121)
|
||||
assert.ok(text.endsWith('…'))
|
||||
assert.equal(text, `${'x'.repeat(120)}…`)
|
||||
})
|
||||
|
||||
test('markdownToPlainText:按码点截断,emoji 不会被劈成半个', () => {
|
||||
const text = markdownToPlainText('😀'.repeat(10), 3)
|
||||
assert.equal(text, '😀😀😀…')
|
||||
assert.equal(Array.from(text).length, 4)
|
||||
})
|
||||
|
||||
test('markdownToPlainText:maxChars 为 0 时返回空串', () => {
|
||||
assert.equal(markdownToPlainText('随便什么内容', 0), '')
|
||||
})
|
||||
|
||||
test('markdownToPlainText:#tag 不算标题,分隔线要清干净', () => {
|
||||
// 与 outline.ts 同一套判断:`#` 后必须有空白才算 ATX 标题
|
||||
assert.equal(markdownToPlainText('#hashtag 不是标题'), '#hashtag 不是标题')
|
||||
assert.equal(markdownToPlainText('* * *\n\n正文'), '正文')
|
||||
assert.equal(markdownToPlainText('***'), '')
|
||||
assert.equal(markdownToPlainText('标题\n==='), '标题')
|
||||
// 表格:对齐分隔行是纯标记,要丢;单元格文字留着
|
||||
const table = markdownToPlainText('| 模块 | 状态 |\n| --- | --- |\n| 导出 | 完成 |')
|
||||
assert.ok(!table.includes('---'), table)
|
||||
assert.ok(table.includes('模块'), table)
|
||||
assert.ok(table.includes('导出'), table)
|
||||
})
|
||||
|
||||
// ===== 渲染(未消毒片段)=====
|
||||
|
||||
test('围栏代码块:info string 命中注册表时按该语言高亮', () => {
|
||||
const html = renderMarkdownFragmentRaw('```ts\nconst n: number = 1\n```')
|
||||
assert.match(html, /<pre class="hljs">/)
|
||||
assert.match(html, /<code class="hljs language-ts">/)
|
||||
assert.match(html, /hljs-keyword/)
|
||||
})
|
||||
|
||||
test('围栏代码块:`.js` 这类写法也能归一成语言名', () => {
|
||||
const html = renderMarkdownFragmentRaw('```.js\nlet a = 1\n```')
|
||||
assert.match(html, /<code class="hljs language-js">/)
|
||||
})
|
||||
|
||||
test('围栏代码块:未注册的语言不抛错,猜不准的散文保持纯文本', () => {
|
||||
const html = renderMarkdownFragmentRaw('```notalang\nplain text here\n```')
|
||||
assert.match(html, /<pre class="hljs">/)
|
||||
assert.match(html, /plain text here/)
|
||||
// 相关度低于下限时不许上色:给散文瞎上色比没颜色更糟
|
||||
assert.match(html, /<code class="hljs">/)
|
||||
assert.doesNotMatch(html, /language-notalang/)
|
||||
})
|
||||
|
||||
test('围栏代码块:语言名拼错时在候选集里自动识别', () => {
|
||||
const html = renderMarkdownFragmentRaw('```typescrpt\ndef f(x):\n return x + 1\n```')
|
||||
assert.match(html, /<code class="hljs language-python">/)
|
||||
assert.match(html, /hljs-keyword/)
|
||||
})
|
||||
|
||||
test('围栏代码块:内容被转义,原始 HTML 不会漏出去', () => {
|
||||
const html = renderMarkdownFragmentRaw('```\n<b>&"x\n```')
|
||||
assert.match(html, /<b>&"x/)
|
||||
assert.doesNotMatch(html, /<b>/)
|
||||
})
|
||||
|
||||
test('表格渲染', () => {
|
||||
const html = renderMarkdownFragmentRaw('| 列 A | 列 B |\n| --- | --- |\n| 1 | 2 |')
|
||||
assert.match(html, /<table>/)
|
||||
assert.match(html, /<th>列 A<\/th>/)
|
||||
assert.match(html, /<td>2<\/td>/)
|
||||
})
|
||||
|
||||
test('任务列表渲染成 disabled 复选框', () => {
|
||||
const html = renderMarkdownFragmentRaw('- [ ] 未完成\n- [x] 已完成\n- 普通项')
|
||||
assert.match(html, /<input class="task-list-item-checkbox" type="checkbox" disabled>/)
|
||||
assert.match(html, /type="checkbox" disabled checked>/)
|
||||
assert.match(html, /class="task-list-item"/)
|
||||
assert.match(html, /普通项/)
|
||||
})
|
||||
|
||||
test('链接渲染与裸 URL 自动识别', () => {
|
||||
const html = renderMarkdownFragmentRaw('[Thing](https://example.com/a) 与 https://example.org/b')
|
||||
assert.match(html, /<a href="https:\/\/example\.com\/a">Thing<\/a>/)
|
||||
assert.match(html, /href="https:\/\/example\.org\/b"/)
|
||||
})
|
||||
|
||||
test('图片渲染(保留相对路径与 alt)', () => {
|
||||
const html = renderMarkdownFragmentRaw('')
|
||||
assert.match(html, /<img src="images\/demo\.png" alt="示意图">/)
|
||||
})
|
||||
|
||||
test('未消毒片段确实会带出 XSS 载荷(说明消毒不是可选项)', () => {
|
||||
const raw = renderMarkdownFragmentRaw('<img src=x onerror=alert(1)>\n\n<script>alert(1)</script>')
|
||||
assert.match(raw, /onerror=/)
|
||||
assert.match(raw, /<script>/)
|
||||
})
|
||||
|
||||
// ===== 消毒 =====
|
||||
|
||||
test('无 DOM 环境:消毒入口失败关闭,返回空串而不是原文', t => {
|
||||
if (domAvailable) {
|
||||
t.skip('本环境有 DOM,消毒可用,该降级分支不适用')
|
||||
return
|
||||
}
|
||||
assert.equal(renderMarkdownFragment('<script>alert(1)</script>'), '')
|
||||
assert.equal(renderMarkdownFragment('# 正常标题'), '')
|
||||
})
|
||||
|
||||
test('XSS:<script>、onerror、javascript: 被剥掉,该留的留着', t => {
|
||||
if (!domAvailable) {
|
||||
t.skip('Node 无 window/document:DOMPurify.isSupported=false,无 sanitize 可调')
|
||||
return
|
||||
}
|
||||
const dirty = [
|
||||
'<script>alert(1)</script>',
|
||||
'',
|
||||
'<img src=x onerror=alert(1)>',
|
||||
'',
|
||||
'[点我](javascript:alert(1))',
|
||||
'',
|
||||
'<span style="position:fixed;top:0">覆盖界面</span>',
|
||||
'',
|
||||
'<img src="a.png" srcset="https://evil.example/x.png 1x">',
|
||||
'',
|
||||
'<style>body{display:none}</style>'
|
||||
].join('\n')
|
||||
const html = renderMarkdownFragment(dirty)
|
||||
assert.ok(!html.includes('onerror='), html)
|
||||
assert.ok(!html.includes('<script'), html)
|
||||
assert.ok(!html.includes('javascript:'), html)
|
||||
assert.ok(!html.includes('style='), html)
|
||||
assert.ok(!html.includes('srcset='), html)
|
||||
assert.ok(!html.includes('<style'), html)
|
||||
|
||||
const kept = renderMarkdownFragment(
|
||||
'```ts\nconst a = 1\n```\n\n| a |\n| --- |\n| 1 |\n\n- [x] 完成\n\n\n\n<a href="https://example.com" target="_blank">外链</a>'
|
||||
)
|
||||
assert.match(kept, /class="hljs"/)
|
||||
assert.match(kept, /<table>/)
|
||||
assert.match(kept, /type="checkbox" disabled checked/)
|
||||
assert.match(kept, /<img src="a\.png"/)
|
||||
assert.match(kept, /target="_blank"/)
|
||||
})
|
||||
|
||||
// ===== 导出文档 =====
|
||||
|
||||
test('renderNoteHtml:结构完整、<title> 与页首标题都转义', () => {
|
||||
const html = renderNoteHtml({
|
||||
title: 'A & B <C>',
|
||||
fileName: 'a.md',
|
||||
markdown: '# 正文',
|
||||
exportedAt: new Date(2026, 0, 2, 3, 4)
|
||||
})
|
||||
assert.ok(html.startsWith('<!doctype html>'))
|
||||
assert.ok(html.includes('<html lang="zh-CN">'))
|
||||
assert.ok(html.includes('<meta charset="utf-8">'))
|
||||
assert.ok(html.includes('name="viewport"'))
|
||||
assert.ok(html.includes('<title>A & B <C></title>'))
|
||||
assert.ok(html.includes('<h1 class="note-title">A & B <C></h1>'))
|
||||
assert.ok(html.includes('导出自 Thing · 2026-01-02 03:04'))
|
||||
assert.ok(html.includes('@media print'))
|
||||
})
|
||||
|
||||
test('renderNoteHtml:内联样式自带排版与高亮配色,不依赖外部资源', () => {
|
||||
const html = renderNoteHtml({ title: '排版', fileName: 'style.md', markdown: '正文' })
|
||||
assert.ok(html.includes('780px'))
|
||||
assert.ok(html.includes('line-height: 1.7'))
|
||||
assert.ok(html.includes('overflow-x: auto'))
|
||||
assert.ok(html.includes('background: #f6f8fa'))
|
||||
assert.ok(html.includes('border-collapse: collapse'))
|
||||
assert.ok(html.includes('border-left: 4px solid'))
|
||||
assert.ok(html.includes('img { max-width: 100%'))
|
||||
assert.ok(html.includes('.note-body mark'))
|
||||
for (const token of ['hljs-keyword', 'hljs-string', 'hljs-comment', 'hljs-number', 'hljs-title', 'hljs-attr']) {
|
||||
assert.ok(html.includes(`.${token}`), `缺少高亮配色 ${token}`)
|
||||
}
|
||||
})
|
||||
|
||||
test('renderNoteHtml:样式与外壳里没有任何外部地址', () => {
|
||||
const html = renderNoteHtml({
|
||||
title: '离线笔记',
|
||||
fileName: 'offline.md',
|
||||
markdown: '# 标题\n\n纯文本,没有任何链接。'
|
||||
})
|
||||
const style = /<style>([\s\S]*?)<\/style>/.exec(html)?.[1] ?? ''
|
||||
assert.ok(style.length > 0)
|
||||
assert.ok(!/https?:\/\//.test(style), '内联样式里不能出现外链')
|
||||
assert.ok(!/https?:\/\//.test(html), '正文没有链接时整份文档都不该有外链')
|
||||
assert.ok(!/<link|<script|@import/i.test(html))
|
||||
})
|
||||
|
||||
test('renderNoteHtml:标题为空串时用 fileName 兜底,时间缺省用当前时间', () => {
|
||||
const html = renderNoteHtml({ title: ' ', fileName: '周报.md', markdown: '正文' })
|
||||
assert.ok(html.includes('<title>周报.md</title>'))
|
||||
assert.ok(html.includes('<h1 class="note-title">周报.md</h1>'))
|
||||
assert.match(html, /导出自 Thing · \d{4}-\d{2}-\d{2} \d{2}:\d{2}/)
|
||||
})
|
||||
Reference in new issue
Block a user