Skip to main content

Testing interactions

The Virtual Screen Reader exposes the same broad interaction contract as the real screen reader APIs.

The main APIs to use are:

These methods are meant to be used in a DOM-driven test environment, usually with a virtual.start({ container }) setup.

Activate the current item​

The Virtual Screen Reader supports the same default activation shape as the real screen reader interface:

await virtual.start({ container: document.body });

try {
  await virtual.next();
  await virtual.act();
} finally {
  await virtual.stop();
}

Use act() when a DOM control should be triggered through its default action.

Type and press keys​

Use press() and type() to simulate input-style workflows:

await virtual.press("Control+f");
await virtual.type("checkout");

The input APIs are the Virtual Screen Reader keyboard and text injection points.

Click through a control model​

Use click() when a test needs to drive a pointer behaviour in the Virtual Screen Reader environment:

await virtual.click({ button: "left", clickCount: 1 });

Read the current item text​

itemText() reports the current item label or text under the virtual reader cursor:

const currentItem = await virtual.itemText();
expect(currentItem).toContain("Checkout");

That can be used to confirm which node the Virtual Screen Reader cursor is focused on.