Visualisation text layer missing in text mode

viz/text-missing

WCAG success criterion: 1.1.1 (level A)
Outcome class: Automatic. An automated check that can report a confirmed violation.
Area: Accessibility, Content & Alternatives
Default severity: high
axe-core rule: display/text-not-visible

Why we flag this

In text mode the statement of the visualisation is lost.

[data-viz-text] missing, empty or not visible with html[data-display=text].

What the rule checks

Scope. Every figure[data-viz] (opt-in convention).

Expectation. A visualisation without a [data-viz-text] layer, or whose layers contain only whitespace, has nothing for text mode and no text alternative (1.1.1): violation at the figure (best-practice tag). One word of text passes. A figure that is not presented (inside a hidden panel) carries no graphic either and is not judged in that state (#725).

Source. Display-mode convention v0 (casoon/barrierlab#22) hung on WCAG 2.2 SC 1.1.1, 1.3.1, 2.2.2; HTML §4.4.12-13 figure/figcaption

Detection: shared; needs dom.

Failing example

The check reports a violation here. Corpus case content_viz_broken, element figure#loans-chart. The page may contain further findings for other rules.

content_viz_broken.htmlHTML

<!DOCTYPE html>
<html lang="en">
<head>
  <meta charset="utf-8">
  <meta name="viewport" content="width=device-width, initial-scale=1">
  <title>Library in numbers: broken figures</title>
  <script>document.documentElement.dataset.display = 'visual';</script>
  <style>
    html:not([data-display="text"]) [data-viz-text] {
      position: absolute; width: 1px; height: 1px; overflow: hidden;
      clip-path: inset(50%); white-space: nowrap;
    }
    body > a[href="#main"], #display-toggle { display: inline-block; min-width: 44px; min-height: 44px; }
  </style>
</head>
<body>
  <a href="#main">Skip to main content</a>
  <header>
    <nav aria-label="Display">
      <label for="display-toggle">Display</label>
      <select id="display-toggle" data-display-toggle>
        <option value="visual" selected>Visual</option>
        <option value="calm">Calm</option>
        <option value="text">Text</option>
      </select>
    </nav>
  </header>
  <main id="main">
    <h1>Library in numbers</h1>
    <figure id="loans-chart" data-viz="chart">
      <svg data-viz-static aria-hidden="true" width="120" height="30" viewBox="0 0 120 30"><path d="M0 25 L120 5" stroke="#333" fill="none"></path></svg>
    </figure>
    <figure id="blank-text" data-viz="map">
      <svg data-viz-static aria-hidden="true" width="120" height="30" viewBox="0 0 120 30"><path d="M0 25 L120 5" stroke="#333" fill="none"></path></svg>
      <div data-viz-text>   </div>
      <figcaption>Branches on the map</figcaption>
    </figure>
    <figure id="globe" data-viz="3d">
      <canvas data-viz-live width="200" height="100" aria-hidden="true"></canvas>
      <div data-viz-text><p>A globe of partner libraries; drag to turn it.</p></div>
      <figcaption>Partner libraries</figcaption>
    </figure>
    <figure id="floor-explorer" data-viz="Interactive">
      <canvas data-viz-live width="200" height="100" aria-hidden="true"></canvas>
      <div data-viz-text><p>Floor plan; select a room to see its seats.</p></div>
      <div><figcaption>Floor explorer</figcaption></div>
    </figure>
    <viz-card></viz-card>
    <div id="late-viz"></div>
    <script>
      customElements.define('viz-card', class extends HTMLElement {
        connectedCallback() {
          this.attachShadow({ mode: 'open' }).innerHTML =
            '<figure id="shadow-chart" data-viz="chart"><div data-viz-text><p>Visits rose from 200 to 310.</p></div></figure>';
        }
      });
      document.getElementById('late-viz').innerHTML =
        '<figure id="late-chart" data-viz="chart"><div data-viz-text><p>Members rose from 4,100 to 4,800.</p></div></figure>';
    </script>
  </main>
  <footer><p>City Library, 1 Market Square</p></footer>
</body>
</html>

Passing example

The check passes here and reports nothing. Corpus case content_viz_hidden_panel. The page may contain further findings for other rules.

content_viz_hidden_panel.htmlHTML

<!DOCTYPE html>
<html lang="en">
<head>
  <meta charset="utf-8">
  <meta name="viewport" content="width=device-width, initial-scale=1">
  <title>Library in numbers: archived panel</title>
  <script>document.documentElement.dataset.display = 'visual';</script>
  <style>
    html:not([data-display="text"]) [data-viz-text] {
      position: absolute; width: 1px; height: 1px; overflow: hidden;
      clip-path: inset(50%); white-space: nowrap;
    }
    body > a[href="#main"], #display-toggle { display: inline-block; min-width: 44px; min-height: 44px; }
  </style>
</head>
<body>
  <a href="#main">Skip to main content</a>
  <header>
    <nav aria-label="Display">
      <label for="display-toggle">Display</label>
      <select id="display-toggle" data-display-toggle>
        <option value="visual" selected>Visual</option>
        <option value="calm">Calm</option>
        <option value="text">Text</option>
      </select>
    </nav>
  </header>
  <main id="main">
    <h1>Library in numbers</h1>
    <section aria-labelledby="this-year">
      <h2 id="this-year">This year</h2>
      <p>Figures for this year follow in spring.</p>
    </section>
    <section aria-labelledby="last-year" hidden>
      <h2 id="last-year">Last year</h2>
      <figure id="archived-chart" data-viz="interactive">
        <canvas data-viz-live width="200" height="100" aria-hidden="true"></canvas>
        <div data-viz-text><p>Loans fell slightly to 14,200.</p></div>
      </figure>
    </section>
  </main>
  <footer><p>City Library, 1 Market Square</p></footer>
</body>
</html>

This page is generated from the rule catalogue and the rule contracts; see therule overview for how the outcome classes are derived.