feat: Add comprehensive Azure MSTTS support with automatic namespace injection

[willwade]

Overview

This PR adds comprehensive Azure MSTTS (Microsoft Text-to-Speech) support to Speech Markdown, including automatic namespace injection, 33 express-as styles with intensity control, and language switching support.

Key Features

1. Automatic Azure SSML Namespace Injection

• Automatic detection: The formatter detects when MSTTS tags (<mstts:express-as>) are present in generated SSML
• Conditional injection: Only adds xmlns:mstts="https://www.w3.org/2001/mstts" namespace when needed
• Seamless integration: Users don't need to manually edit SSML or worry about namespace declarations

2. Complete Express-As Style Support (33 styles)

3. Style Degree (Intensity Control)

• Numeric intensity control from 0.01 to 2.0 (default 1.0)
• Example: (text)[excited:"1.5"] generates <mstts:express-as style="excited" styledegree="1.5">
• Works with all express-as styles

4. Language Switching Support

• Full support for lang modifier: (Paris)[lang:"fr-FR"]
• Generates <lang xml:lang="fr-FR">Paris</lang>
• Works at both text and section levels

5. Docs

• Updated docs/platforms/azure.md with all 33 styles
• Documented multi-speaker dialog (mstts:dialog and mstts:turn) with raw SSML examples
• Documented role attributes (Girl, Boy, YoungAdultFemale, etc.) with raw SSML examples
• Feature comparison with Amazon Alexa and Google Assistant

Examples

Basic Express-As Style

(Hello!)[excited]

Generates:

<speak xmlns:mstts="https://www.w3.org/2001/mstts">
<mstts:express-as style="excited">Hello!</mstts:express-as>
</speak>

Style with Intensity

(This is very exciting!)[excited:"1.8"]

Generates:

<speak xmlns:mstts="https://www.w3.org/2001/mstts">
<mstts:express-as style="excited" styledegree="1.8">This is very exciting!</mstts:express-as>
</speak>

Language Switching

In Paris, they pronounce it (Paris)[lang:"fr-FR"].

Generates:

<speak>
In Paris, they pronounce it <lang xml:lang="fr-FR">Paris</lang>.
</speak>

Section-Level Style

#[excited]
This entire section is excited!
Multiple sentences work too.
#[excited]

Generates:

<speak xmlns:mstts="https://www.w3.org/2001/mstts">
<mstts:express-as style="excited">
This entire section is excited!
Multiple sentences work too.
</mstts:express-as>
</speak>

---

Ready for review! This PR brings Azure MSTTS support from 27 to 33 styles, adds language switching, and provides comprehensive documentation for all Azure-specific features.

[willwade]

Note. WIP. Let me check after Claude done a ton of this

[willwade]

Ready for review now @arjan - this gives way more feature support to azure tts and its various intricacies.. I'd say its more feature packed now than any others..

[willwade]

So note.. We've fixed and tested quite a bit in this - we've added tests directy using SSML snippets from azure and google cloud docs.. We've also done quite a bit to add langs to the voice lists and voice-id so a user can use either id or name and we replace it correctly in the ssml with the id

[arjan]

This looks pretty "comprehensive" indeed Will! Nice work.
I must say, I don't use SSML / Azure that much atm so I will proceed to merge this without further verification, if that's OK with you?

[willwade]

TY @arjan - I'll just keep an eye on a release.. (NB: ive been working on a PR for the editor - I'll hold off doing much more on that till released.. I have grand plans for that but will prolly keep the "grand" plans for a completely seperate PR you may not want to release! second NB: https://github.com/willwade/js-tts-wrapper - is a wrapper supporting speechmarkdown across as many TTS systems as possible.. we can give live preview of the output..)

[arjan]

Release done ✅