Skip to content

An open source design system that's fully customizable and agent ready

TypeScript

fix(Markdown): end a list at a thematic break (#7087)

Cindy Zhang34d32e6528b5
diff --git a/.changeset/markdown-thematic-break-ends-list.md b/.changeset/markdown-thematic-break-ends-list.md
new file mode 100644
index 0000000000..f18801f33e
--- /dev/null
+++ b/.changeset/markdown-thematic-break-ends-list.md
@@ -0,0 +1,8 @@
+---
+'@astryxdesign/core': patch
+---
+
+[fix] Markdown: end a list at a thematic break
+@cixzhang
+
+`Markdown` now reads a line such as `* * *` or `- - -` after a list item as a thematic break that ends the list, as CommonMark specifies, rather than as another item holding a nested list.
diff --git a/packages/core/src/Markdown/listThematicBreak.test.tsx b/packages/core/src/Markdown/listThematicBreak.test.tsx
new file mode 100644
index 0000000000..0c9f5bdee8
--- /dev/null
+++ b/packages/core/src/Markdown/listThematicBreak.test.tsx
@@ -0,0 +1,49 @@
+// Copyright (c) Meta Platforms, Inc. and affiliates.
+
+import {describe, expect, it} from 'vitest';
+import {parseMarkdownAst} from './parser';
+
+type Json = {type: string; value?: string; children?: Json[]};
+
+/** The document's blocks as compact markup. */
+function blocks(markdown: string): string {
+ const show = (node: Json): string =>
+ node.type === 'text'
+ ? JSON.stringify(node.value)
+ : node.children == null
+ ? node.type
+ : `${node.type}[${node.children.map(show).join(', ')}]`;
+ return (parseMarkdownAst(markdown).children as unknown as Json[])
+ .map(show)
+ .join(' | ');
+}
+
+describe('a thematic break after a list item (CommonMark 0.31 §4.1)', () => {
+ it.each([
+ ['`* * *` after a `-` item', '- item\n* * *'],
+ ['`- - -` after a `-` item', '- item\n- - -'],
+ ['`* * *` after a `*` item', '* item\n* * *'],
+ ['a break after a blank line', '- item\n\n* * *'],
+ ['a break after an ordered item', '1. item\n* * *'],
+ ])('ends the list: %s', (_, markdown) => {
+ expect(blocks(markdown)).toMatch(
+ /^list\[listItem\[paragraph\["item"\]\]\] \| thematicBreak$/,
+ );
+ });
+
+ it.each([
+ [
+ 'a break indented into the item',
+ '- a\n * * *\n b',
+ 'list[listItem[paragraph["a"], thematicBreak, paragraph["b"]]]',
+ ],
+ ['an item holding a break', '- * * *', 'list[listItem[thematicBreak]]'],
+ [
+ 'two items',
+ '- a\n- b',
+ 'list[listItem[paragraph["a"]], listItem[paragraph["b"]]]',
+ ],
+ ])('leaves %s as it was', (_, markdown, expected) => {
+ expect(blocks(markdown)).toBe(expected);
+ });
+});
diff --git a/packages/core/src/Markdown/parser.ts b/packages/core/src/Markdown/parser.ts
index bc5cf3f9e1..3335a16272 100644
--- a/packages/core/src/Markdown/parser.ts
+++ b/packages/core/src/Markdown/parser.ts
@@ -3097,7 +3097,11 @@ function parseList(
let loose = false;
let index = startIndex;
- while (index < lines.length && itemPattern.test(lines[index])) {
+ // A line that is a thematic break is one, not an item, even where an item
+ // could start (CommonMark 0.31 §4.1): `* * *` after `- item` ends the list.
+ const startsItem = (line: string) =>
+ itemPattern.test(line) && !isHorizontalRule(line);
+ while (index < lines.length && startsItem(lines[index])) {
const content = ordered
? lines[index].replace(new RegExp(`^ *\\d+${escDelim} `), '')
: lines[index].replace(/^ *[-*+] /, '');
@@ -3213,7 +3217,7 @@ function parseList(
if (
lookahead > index &&
lookahead < lines.length &&
- itemPattern.test(lines[lookahead])
+ startsItem(lines[lookahead])
) {
loose = true;
index = lookahead;
diff --git a/packages/richtext/src/markdownListThematicBreak.test.ts b/packages/richtext/src/markdownListThematicBreak.test.ts
new file mode 100644
index 0000000000..5b2919488a
--- /dev/null
+++ b/packages/richtext/src/markdownListThematicBreak.test.ts
@@ -0,0 +1,31 @@
+// Copyright (c) Meta Platforms, Inc. and affiliates.
+
+import {describe, expect, it} from 'vitest';
+import {parseMarkdownAst} from '@astryxdesign/core/Markdown/parser';
+import {markdownToEditorStateJSON} from './markdownSerializers';
+
+type Json = {type: string; children?: Json[]};
+
+const CORE_TO_RICH: Record<string, string> = {
+ list: 'list',
+ thematicBreak: 'horizontalrule',
+ paragraph: 'paragraph',
+};
+
+describe('a thematic break after a list item, on both surfaces', () => {
+ it.each([
+ '- item\n* * *',
+ '- item\n- - -',
+ '* item\n* * *',
+ '- item\n\n* * *',
+ ])('reads %j as a list and then a break, as core does', markdown => {
+ const rich = (
+ JSON.parse(markdownToEditorStateJSON(markdown)) as {root: Json}
+ ).root.children?.map(node => node.type);
+ const core = (parseMarkdownAst(markdown).children as unknown as Json[]).map(
+ node => CORE_TO_RICH[node.type] ?? node.type,
+ );
+ expect(rich).toEqual(['list', 'horizontalrule']);
+ expect(core).toEqual(rich);
+ });
+});