innerHTML parser interprets / in unquoted element attributes incorrectly

[maxmilton]

Describe the bug
When setting el.innerHTML, the resulting DOM is incorrect when an element's attribute value is unquoted and contains a forward slash. Rather than the slash becoming part of the value, it's removed and the remaining value is tokenized into a separate attribute.

To Reproduce
Steps to reproduce the behavior:

1. Create a test file; index.test.ts:

...

test('renders data-x attribute', () => {
  const container = document.createElement('div');

  container.innerHTML = '<div data-x=a/b></div>';

  expect(container.innerHTML).toBe('<div data-x="a/b"></div>');
});

test('renders href attribute', () => {
  const container = document.createElement('div');

  container.innerHTML = '<a href=https://example.com/foo/bar>link</a>';

  expect(container.innerHTML).toBe('<a href="https://example.com/foo/bar">link</a>');
});

2. Run the tests.

Expected behavior
In the repro, both tests should pass. Forward slashes should be included as part of the attribute value.

Screenshots

Failing test output:

...

error: expect(received).toBe(expected)

Expected: "<div data-x=\"a/b\"></div>"
Received: "<div data-x=\"a\" b=\"\"></div>"

...

error: expect(received).toBe(expected)

Expected: "<a href=\"https://example.com/foo/bar\">link</a>"
Received: "<a href=\"https:\" example.com=\"\" foo=\"\" bar=\"\">link</a>"

Device:

• OS: Linux 6.4.4-arch1-1 x86_64
• happy-dom version: 10.5.2

Additional context
Since under the hood XMLParser is used to parse when setting innerHTML, there will be many other characters which exhibit the same behavior. However, forward slashes are the biggest impact for me as I'm running into issues with setting URLs in anchor href tags.