Add SSML parser and speakSsml function (#15699)
Preparations for #12945
Summary of the issue:
The way you can output speech using the nvda controller is pretty limited. It is not possible to control prosody parameters, for example. Also you can't easily insert pauses in speech.
Construction of speech sequences can be pretty complex. It would if we'd have some standardized way of providing speech attributes in text.
Description of user facing changes
None for normal users, only developers.
Description of development approach
Documented in changelog entries. Basically:
Added an SSML Parser to generate speech sequences
Added speech.speakSsml to generate speech output using SSML.
Support in the NVDA Controller Client will be added in a follow up.
