5.2.3 SpeechSynthesisUtterance Attributes

At 5.2.3 SpeechSynthesisUtterance Attributes text attribute
<https://w3c.github.io/speech-api/speechapi.html#utterance-attributes>
>
> This may be either plain text or a complete, well-formed SSML document.



How is a "well-formed SSML document" expected to be parsed when set at
the *text
*attribute of a *single* instance of SpeechSynthesisUtterance?

The purpose for the need for clarification is relevant to the present
attempt to implement SSML parsing for Web Speech API, see
https://github.com/guest271314/SpeechSynthesisSSMLParser.