Update schema to reflect changes to GH

*   `ariaHidden` is not supported by GH
*   ARIA attributes `ariaDescribedBy`, `ariaLabel`, and `ariaLabelledBy` are
    allowed on `a`, `dl`, `ol`, and `ul` by GH
*   Add support for footnotes and their properties
*   Add `language-*` classes on `code`
*   Add `contains-task-list` class on `ol`, `ul`
*   Remove `vSpace`
*   Remove `type`, other than on `input`
*   Add `ariaDescribedBy`, `ariaLabelledBy` to clobbered attributes
*   Remove `abbr`, `bdo`, `caption`, `cite`, `dfn`, `figcaption`, `figure`,
    `mark`, `small`, `time`, `wbr` elements
*   Add `picture`, `source` elements

Author: wooorm

.prettierignore

@@ -1,2 +1,3 @@
 coverage/
+*.html
 *.md

input.html

@@ -1,3 +1,6 @@
+// Couple of ARIA attributes allowed in several, but not all, places.
+const aria = ['ariaDescribedBy', 'ariaLabel', 'ariaLabelledBy']
+
 /**
  * Default schema.
  *
@@ -15,18 +18,53 @@ export const defaultSchema = {
     tr: ['table']
   },
   attributes: {
-    a: ['href'],
+    a: [
+      ...aria,
+      // Note: these 3 are used by GFM footnotes, they do work on all links.
+      'dataFootnoteBackref',
+      'dataFootnoteRef',
+      ['className', 'data-footnote-backref'],
+      'href'
+    ],
     blockquote: ['cite'],
+    // Note: this class is not normally allowed by GH, when manually writing
+    // `code` as HTML in markdown, they adds it some other way.
+    // We can’t do that, so we have to allow it.
+    code: [['className', /^language-./]],
     del: ['cite'],
     div: ['itemScope', 'itemType'],
-    img: ['longDesc', 'src'],
+    dl: [...aria],
+    // Note: these 2 are used by GFM footnotes, they *sometimes* work.
+    h2: [
+      ['id', 'footnote-label'],
+      ['className', 'sr-only']
+    ],
+    img: [...aria, 'longDesc', 'src'],
+    // Note: `input` is not normally allowed by GH, when manually writing
+    // it in markdown, they add it from tasklists some other way.
+    // We can’t do that, so we have to allow it.
     input: [
       ['disabled', true],
       ['type', 'checkbox']
     ],
     ins: ['cite'],
+    // Note: this class is not normally allowed by GH, when manually writing
+    // `li` as HTML in markdown, they adds it some other way.
+    // We can’t do that, so we have to allow it.
     li: [['className', 'task-list-item']],
+    // Note: this class is not normally allowed by GH, when manually writing
+    // `ol` as HTML in markdown, they adds it some other way.
+    // We can’t do that, so we have to allow it.
+    ol: [...aria, ['className', 'contains-task-list']],
     q: ['cite'],
+    section: ['dataFootnotes', ['className', 'footnotes']],
+    source: ['srcSet'],
+    summary: [...aria],
+    table: [...aria],
+    // Note: this class is not normally allowed by GH, when manually writing
+    // `ol` as HTML in markdown, they adds it some other way.
+    // We can’t do that, so we have to allow it.
+    ul: [...aria, ['className', 'contains-task-list']],
     '*': [
       'abbr',
       'accept',
@@ -35,10 +73,6 @@ export const defaultSchema = {
       'action',
       'align',
       'alt',
-      'ariaDescribedBy',
-      'ariaHidden',
-      'ariaLabel',
-      'ariaLabelledBy',
       'axis',
       'border',
       'cellPadding',
@@ -94,16 +128,14 @@ export const defaultSchema = {
       'tabIndex',
       'target',
       'title',
-      'type',
       'useMap',
       'vAlign',
-      'vSpace',
       'value',
       'width'
     ]
   },
   clobberPrefix: 'user-content-',
-  clobber: ['id', 'name'],
+  clobber: ['ariaDescribedBy', 'ariaLabelledBy', 'id', 'name'],
   protocols: {
     href: ['http', 'https', 'irc', 'ircs', 'mailto', 'xmpp'],
     cite: ['http', 'https'],
@@ -116,24 +148,17 @@ export const defaultSchema = {
   strip: ['script'],
   tagNames: [
     'a',
-    'abbr',
     'b',
-    'bdo',
     'blockquote',
     'br',
-    'caption',
-    'cite',
     'code',
     'dd',
     'del',
     'details',
-    'dfn',
     'div',
     'dl',
     'dt',
     'em',
-    'figcaption',
-    'figure',
     'h1',
     'h2',
     'h3',
@@ -143,21 +168,25 @@ export const defaultSchema = {
     'hr',
     'i',
     'img',
+    // Note: `input` is not normally allowed by GH, when manually writing
+    // it in markdown, they add it from tasklists some other way.
+    // We can’t do that, so we have to allow it.
     'input',
     'ins',
     'kbd',
     'li',
-    'mark',
     'ol',
     'p',
+    'picture',
     'pre',
     'q',
     'rp',
     'rt',
     'ruby',
     's',
     'samp',
-    'small',
+    'section',
+    'source',
     'span',
     'strike',
     'strong',
@@ -170,11 +199,9 @@ export const defaultSchema = {
     'tfoot',
     'th',
     'thead',
-    'time',
     'tr',
     'tt',
     'ul',
-    'var',
-    'wbr'
+    'var'
   ]
 }

lib/schema.js

@@ -43,23 +43,28 @@
   "devDependencies": {
     "@types/node": "^20.0.0",
     "@types/ungap__structured-clone": "^0.3.0",
+    "aria-attributes": "^2.0.0",
     "c8": "^8.0.0",
     "deepmerge": "^4.0.0",
+    "hast-util-from-html": "^1.0.0",
     "hast-util-to-html": "^8.0.0",
     "hastscript": "^8.0.0",
+    "html-element-attributes": "^3.0.0",
+    "html-tag-names": "^2.0.0",
     "prettier": "^3.0.0",
     "remark-cli": "^11.0.0",
     "remark-preset-wooorm": "^9.0.0",
     "type-coverage": "^2.0.0",
     "typescript": "^5.0.0",
     "unist-builder": "^4.0.0",
+    "unist-util-visit": "^5.0.0",
     "xo": "^0.55.0"
   },
   "scripts": {
     "prepack": "npm run build && npm run format",
     "build": "tsc --build --clean && tsc --build && type-coverage",
     "format": "remark . -qfo && prettier . -w --log-level warn && xo --fix",
-    "test-api": "node --conditions development test.js",
+    "test-api": "node --conditions development test/index.js",
     "test-coverage": "c8 --100 --reporter lcov npm run test-api",
     "test": "npm run build && npm run format && npm run test-coverage"
   },
@@ -90,7 +95,7 @@
     "overrides": [
       {
         "files": [
-          "test.js"
+          "test/**/*.js"
         ],
         "rules": {
           "max-nested-callbacks": "off",