diff --git a/lib/html.js b/lib/html.js
index 10027413f..3ed942b04 100644
--- a/lib/html.js
+++ b/lib/html.js
@@ -78,11 +78,12 @@ const defaultHtmlOpts = {
textElements: ['label', 'h1', 'h2'],
allowedAttrs: ['id', 'for', 'class', 'name', 'type', 'value', 'tabindex', 'aria-labelledby', 'aria-label', 'label', 'placeholder', 'title', 'alt', 'src', 'role'],
allowedRoles: ['button', 'checkbox', 'search', 'textbox', 'tab'],
+ keepText: false,
}
function removeNonInteractiveElements(html, opts = {}) {
opts = { ...defaultHtmlOpts, ...opts }
- const { interactiveElements, textElements, allowedAttrs, allowedRoles } = opts
+ const { interactiveElements, textElements, allowedAttrs, allowedRoles, keepText } = opts
// Parse the HTML into a document tree
const document = parse(html)
@@ -111,8 +112,14 @@ function removeNonInteractiveElements(html, opts = {}) {
return false
}
+ function hasVisibleText(node) {
+ if (node.nodeName === '#text') return !!node.value.trim()
+ return (node.childNodes || []).some(hasVisibleText)
+ }
+
function hasMeaningfulText(node) {
if (textElements.includes(node.nodeName)) return true
+ if (keepText && hasVisibleText(node)) return true
return false
}
@@ -294,7 +301,7 @@ function splitByChunks(text, chunkSize) {
function simplifyHtmlElement(html, maxLength = 300) {
try {
- html = removeNonInteractiveElements(html)
+ html = removeNonInteractiveElements(html, { keepText: true })
html = html.replace(/(?:.*?<\/head>)?(.*)<\/body><\/html>/s, '$1').trim()
} catch (e) {
// keep raw html if minification fails
diff --git a/test/unit/html_test.js b/test/unit/html_test.js
index 1de0d444d..17e5dd577 100644
--- a/test/unit/html_test.js
+++ b/test/unit/html_test.js
@@ -3,7 +3,7 @@ import path from 'path'
import { expect } from 'chai'
import { fileURLToPath } from 'url'
import * as cheerio from 'cheerio'
-import { scanForErrorMessages, removeNonInteractiveElements, minifyHtml, splitByChunks, cleanHtml, formatHtml, isTrashClass } from '../../lib/html.js'
+import { scanForErrorMessages, removeNonInteractiveElements, minifyHtml, splitByChunks, cleanHtml, formatHtml, isTrashClass, simplifyHtmlElement } from '../../lib/html.js'
const __filename = fileURLToPath(import.meta.url)
const __dirname = path.dirname(__filename)
@@ -224,4 +224,35 @@ describe('HTML module', () => {
expect(out).to.include('hi')
})
})
+
+ describe('#simplifyHtmlElement', () => {
+ const button = label =>
+ '`
+
+ it('keeps the visible label when it is nested in non-interactive elements', () => {
+ const out = simplifyHtmlElement(button('New test'))
+ expect(out).to.include('New test')
+ expect(out).to.include('