feat: Multi-Speaker Dialogue, 28 Emotional Styles

[huan-zz3]

🙏 First PR Notice

🌟 This is my first Pull Request ever! 🌟

I'm incredibly excited (and a bit nervous) to be contributing to this project. While I've done my best to ensure everything is perfect, I understand there might be areas that need improvement.

If you find anything:

• Incomplete translations
• Awkward phrasing
• Missing edge cases
• Better ways to structure things
• Any issues whatsoever

Please don't hesitate to let me know! I'm committed to making this PR as good as it needs to be. I'll respond quickly to any feedback and make changes until you're completely satisfied with the result.

Your guidance and review would mean a lot to me as I learn the contribution process. Thank you for your time and consideration! 🙏

✨ New Features Added

1. Multi-Speaker Dialogue Support

Build multi-speaker conversations effortlessly with chainable API:

import { DialogueBuilder, buildDialogueSSML } from "msedge-tts";

// Method 1: Chained builder
const dialogue = new DialogueBuilder()
  .addTurn({
    voice: "zh-CN-XiaoxiaoNeural",
    text: "Hello everyone!",
    style: "cheerful"
  })
  .addTurn({
    voice: "en-US-AndrewNeural",
    text: "Welcome to our podcast!",
    lang: "en-US"
  })
  .build();

// Method 2: Functional API
const ssml = buildDialogueSSML([
  { voice: "zh-CN-YunxiNeural", text: "Today we discuss AI", style: "documentary-narration" },
  { voice: "en-US-AriaNeural", text: "That's right!", style: "excited", lang: "en-US" }
]);

2. 28 Emotional Styles

Full support for Microsoft Azure's official emotional styles:

const ssml = buildDialogueSSML([
  { voice: "zh-CN-XiaomoNeural", text: "I'm so happy!", style: "cheerful" },
  { voice: "zh-CN-XiaoxiaoNeural", text: "Welcome to customer service", style: "customerservice" },
  { voice: "en-US-JennyNeural", text: "Breaking news!", style: "newscast-formal", lang: "en-US" }
]);

Supported Styles:

• advertisement_upbeat, affectionate, angry, assistant
• calm, chat, cheerful, customerservice
• depressed, documentary-narration, empathetic, excited
• fearful, friendly, gentle, hopeful
• lyrical, narration-professional, narration-relaxed
• newscast, newscast-casual, newscast-formal
• poetry-reading, sad, serious, shouting
• sports_commentary, sports_commentary_excited
• terrified, unfriendly, whispering

3. Style Degree Control

Fine-tune emotional intensity from 0.01 to 2.0:

const ssml = buildDialogueSSML([
  { 
    voice: "zh-CN-XiaomoNeural", 
    text: "This is normal",
    style: "sad",
    styleDegree: 0.5  // Weaker emotion
  },
  { 
    voice: "zh-CN-XiaomoNeural", 
    text: "This is heartbreaking!",
    style: "sad",
    styleDegree: 2.0  // Strongest emotion
  }
]);

4. Text Substitution

Replace abbreviations and technical terms with full pronunciations:

const ssml = buildDialogueSSML([
  { 
    voice: "zh-CN-XiaoxiaoNeural",
    text: "W3C制定了 Web 标准，API 基于 HTTP 协议",
    substitutions: [
      { text: "W3C", alias: "万维网联盟" },
      { text: "Web", alias: "万维网" },
      { text: "HTTP", alias: "超文本传输协议" }
    ],
    style: "narration-professional"
  }
]);

5. Comprehensive Examples

6 ready-to-run examples demonstrating all features:

Example	Description	File
0	Simple dialogue demo	00-simple-dialogue-demo.ts
1	Multi-speaker (chained)	01-multi-speaker-dialogue-chained.ts
2	Multi-speaker (functional)	02-multi-speaker-dialogue-functional.ts
3	31 emotional styles	03-31-emotional-styles-demo.ts
4	Style degree control	04-style-degree-control-demo.ts
5	Text substitution	05-text-substitution-demo.ts

[Migushthe2nd]

I like the idea. I could maybe add a utils feature that could provide this.
Please however

• remove ai docs
• remove ai agent files
• use an xml escape library, not a custom-made function (this library could provide a automatic escaping in the non-raw functions)
• review your code, examples and config, to remove any reliances on and mentions of your "https://ttspro.cn/" website, and do not use a config like this in the first place. (See test scripts for examples)
• please verify for me that all your voice types, parameters, and ssml speech tags work. I feel like AI just assumed all of them work (while some are blocked - see readme.md)

[huan-zz3]

thanks for teaching, I will make changes as your requirment.