feat: remove the 10,000-character article body limit and make the New article body optional

web/src/lib/__tests__/bodyCounts.spec.ts

@@ -0,0 +1,33 @@
+import { describe, expect, it } from 'vitest'
+import { bodyCounts, isLargeBody, LARGE_BODY_CHARACTERS } from '../bodyCounts'
+
+describe('bodyCounts', () => {
+	it('counts words across block boundaries and characters as the string length', () => {
+		const html =
+			'<h2>Price</h2><ul><li>One</li><li>Two</li></ul><p>A <strong>bold</strong> end.</p>'
+
+		expect(bodyCounts(html)).toEqual({ words: 6, characters: html.length })
+	})
+
+	it('splits adjacent blocks and line breaks into separate words', () => {
+		expect(bodyCounts('<p>one</p><p>two</p>').words).toBe(2)
+		expect(bodyCounts('<p>one<br>two</p>').words).toBe(2)
+		expect(bodyCounts('<p>  one   two  </p>').words).toBe(2)
+	})
+
+	it('counts no words in markup without text', () => {
+		expect(bodyCounts('<p> </p>')).toEqual({ words: 0, characters: 8 })
+	})
+
+	it('reads an empty document as zero', () => {
+		expect(bodyCounts('')).toEqual({ words: 0, characters: 0 })
+		expect(bodyCounts('<p></p>')).toEqual({ words: 0, characters: 0 })
+	})
+})
+
+describe('isLargeBody', () => {
+	it('flips exactly at the threshold', () => {
+		expect(isLargeBody(LARGE_BODY_CHARACTERS - 1)).toBe(false)
+		expect(isLargeBody(LARGE_BODY_CHARACTERS)).toBe(true)
+	})
+})

web/src/lib/bodyCounts.ts

@@ -0,0 +1,23 @@
+const HTML_MIME = 'text/html'
+const WHITESPACE = /\s+/
+const EMPTY_DOCUMENT = '<p></p>'
+const BLOCK_END = /<\/(?:p|h[1-6]|li|blockquote|pre|div)>|<br\s*\/?>/gi
+
+export const LARGE_BODY_CHARACTERS = 50000
+
+export function isLargeBody(characters: number) {
+	return characters >= LARGE_BODY_CHARACTERS
+}
+
+export type BodyCounts = { words: number; characters: number }
+
+export function bodyCounts(html: string): BodyCounts {
+	if (html.length === 0 || html === EMPTY_DOCUMENT) {
+		return { words: 0, characters: 0 }
+	}
+	const text =
+		new DOMParser().parseFromString(html.replace(BLOCK_END, '$& '), HTML_MIME).body
+			.textContent ?? ''
+	const words = text.split(WHITESPACE).filter(Boolean).length
+	return { words, characters: html.length }
+}