Text to Speech using Cognitive Services

[mcorning]

I'm working on a power app for people with dementia. We believe (without any empirical support, yet) that people with dementia will be more responsive to a voice than to reading the same content in text.

i have a TTS .net console app that can produce an excellent result, but i'm stuck getting this capability any farther than a dos prompt.

A perfect solution is to extend the Cognitive Services connector for Text To Search.

Are there any plans for that now? if not, can i offer to help bring it into existence? your grandmother will thank you...

[mcorning]

I am making progress on this journey to the land of Cognitive Services, though i fear i will eventually come up to a cliff where i will go no further. Until that day, i press on, regardless.

Currently, I've written an Azure Function to handle the REST part of the work, and i've developed a valid swagger file to convert the function into a custom API for PowerApps. But when the custom api creation process gets near the end, it fails with this error:

An error occured while converting swagger file to WADL file. Error: 'Unknown swagger type "array" at JSON path paths./api/CognitiveServicesTts.get.responses.200.headers.Content-Type'. Your API may not work with apps.

so now it looks like we need to understand the WADL spec as well as the Swagger spec. Where are the WADL docs? or are there tools we can use (like swagger.io) to test our files before we go to the trouble of building a custom api with files that won't work?

My technical challenge here is specifying a response object that does what Content-Type: audio/x-wav that Postman returns from the same REST call i'm making in my function.

At this writing, here is my best (wrong) guess at the response header's json:

"responses": {
  "200": {
    "description": "Successful response",
    "schema": {
      "title": "Voice synthesis.",
      "type": "string"
    },
    "headers": {
      "Content-Type": {
        "description": "Text to speech",
        "type": "array",
        "items": {
          "type": "integer",
          "format": "int64"
        }
      }
    }
  }
}