Testing character encoding: non-UTF-8 declaration

<meta charset="ISO-8859-1"> is declared, but this file is actually saved as UTF-8 bytes, a genuine mismatch between what's declared and what's real, and a common real-world authoring mistake (e.g. an editor auto-saving as UTF-8 after the encoding was set to something else in the markup).

Café — naïve piñata costs €5.99, “today” only.

Expected result: the browser should honour the declared (incorrect) ISO-8859-1 encoding and mis-decode the underlying UTF-8 bytes, so the accented letters, the em dash, the euro sign, and the curly quotes above should render as mojibake rather than the intended characters. Assistive technology will read whatever garbled substitute characters result, confirm this directly since decode-mismatch handling can vary slightly by browser.

HTML markup

<head>
  <meta charset="ISO-8859-1">
  ...
</head>
<!-- file itself is saved as UTF-8 -->