Automatically generate table of contents in text (#2213)

* automatic toc generation in markdown

* don't open hash links in new tab

* only process toc for top level items

---------

Co-authored-by: ekzyis <ek@stacker.news>

components/table-of-contents.js

@@ -2,11 +2,8 @@ import React, { useMemo, useState } from 'react'
 import Dropdown from 'react-bootstrap/Dropdown'
 import FormControl from 'react-bootstrap/FormControl'
 import TocIcon from '@/svgs/list-unordered.svg'
-import { fromMarkdown } from 'mdast-util-from-markdown'
-import { visit } from 'unist-util-visit'
-import { toString } from 'mdast-util-to-string'
-import { slug } from 'github-slugger'
 import { useRouter } from 'next/router'
+import { extractHeadings } from '@/lib/toc'
 
 export default function Toc ({ text }) {
   const router = useRouter()
@@ -14,16 +11,7 @@ export default function Toc ({ text }) {
     return null
   }
 
-  const toc = useMemo(() => {
-    const tree = fromMarkdown(text)
-    const toc = []
-    visit(tree, 'heading', (node, position, parent) => {
-      const str = toString(node)
-      toc.push({ heading: str, slug: slug(str.replace(/[^\w\-\s]+/gi, '')), depth: node.depth })
-    })
-
-    return toc
-  }, [text])
+  const toc = useMemo(() => extractHeadings(text), [text])
 
   if (toc.length === 0) {
     return null

components/text.js

@@ -20,6 +20,7 @@ import rehypeSN from '@/lib/rehype-sn'
 import remarkUnicode from '@/lib/remark-unicode'
 import Embed from './embed'
 import remarkMath from 'remark-math'
+import remarkToc from '@/lib/remark-toc'
 
 const rehypeSNStyled = () => rehypeSN({
   stylers: [{
@@ -33,7 +34,11 @@ const rehypeSNStyled = () => rehypeSN({
   }]
 })
 
-const remarkPlugins = [gfm, remarkUnicode, [remarkMath, { singleDollarTextMath: false }]]
+const baseRemarkPlugins = [
+  gfm,
+  remarkUnicode,
+  [remarkMath, { singleDollarTextMath: false }]
+]
 
 export function SearchText ({ text }) {
   return (
@@ -49,6 +54,9 @@ export function SearchText ({ text }) {
 
 // this is one of the slowest components to render
 export default memo(function Text ({ rel = UNKNOWN_LINK_REL, imgproxyUrls, children, tab, itemId, outlawed, topLevel }) {
+  // include remarkToc if topLevel
+  const remarkPlugins = topLevel ? [...baseRemarkPlugins, remarkToc] : baseRemarkPlugins
+
   // would the text overflow on the current screen size?
   const [overflowing, setOverflowing] = useState(false)
   // should we show the full text?

lib/remark-toc.js

@@ -0,0 +1,61 @@
+import { SKIP, visit } from 'unist-util-visit'
+import { extractHeadings } from './toc'
+
+export default function remarkToc () {
+  return function transformer (tree) {
+    const headings = extractHeadings(tree)
+
+    visit(tree, 'paragraph', (node, index, parent) => {
+      if (
+        node.children?.length === 1 &&
+        node.children[0].type === 'text' &&
+        node.children[0].value.trim() === '{:toc}'
+      ) {
+        parent.children.splice(index, 1, buildToc(headings))
+        return [SKIP, index]
+      }
+    })
+  }
+}
+
+function buildToc (headings) {
+  const root = { type: 'list', ordered: false, spread: false, children: [] }
+  const stack = [{ depth: 0, node: root }] // holds the current chain of parents
+
+  for (const { heading, slug, depth } of headings) {
+    // walk up the stack to find the parent of the current heading
+    while (stack.length && depth <= stack[stack.length - 1].depth) {
+      stack.pop()
+    }
+    let parent = stack[stack.length - 1].node
+
+    // if the parent is a li, gets its child ul
+    if (parent.type === 'listItem') {
+      let ul = parent.children.find(c => c.type === 'list')
+      if (!ul) {
+        ul = { type: 'list', ordered: false, spread: false, children: [] }
+        parent.children.push(ul)
+      }
+      parent = ul
+    }
+
+    // build the li from the current heading
+    const listItem = {
+      type: 'listItem',
+      spread: false,
+      children: [{
+        type: 'paragraph',
+        children: [{
+          type: 'link',
+          url: `#${slug}`,
+          children: [{ type: 'text', value: heading }]
+        }]
+      }]
+    }
+
+    parent.children.push(listItem)
+    stack.push({ depth, node: listItem })
+  }
+
+  return root
+}

lib/toc.js

@@ -0,0 +1,23 @@
+import { fromMarkdown } from 'mdast-util-from-markdown'
+import { visit } from 'unist-util-visit'
+import { toString } from 'mdast-util-to-string'
+import { slug } from 'github-slugger'
+
+export function extractHeadings (markdownOrTree) {
+  const tree = typeof markdownOrTree === 'string'
+    ? fromMarkdown(markdownOrTree)
+    : markdownOrTree
+
+  const headings = []
+
+  visit(tree, 'heading', node => {
+    const str = toString(node)
+    headings.push({
+      heading: str,
+      slug: slug(str.replace(/[^\w\-\s]+/gi, '')),
+      depth: node.depth
+    })
+  })
+
+  return headings
+}