text-to-speech

Python text-to-speech

This sample adapts PyWinRT's text-to-speech example to generated dynwinrt bindings. It demonstrates:

• native Python asyncio over a WinRT asynchronous operation;
• structural interface typing that passes SpeechSynthesisStream directly where IRandomAccessStream is required;
• a typed MediaPlayer.MediaEnded callback; and
• deterministic COM/WinRT cleanup.

Prerequisites

• Windows 10 or 11 with a Windows SDK installed;
• dynwinrt installed in the selected Python interpreter; and
• dynwinrt-codegen on PATH, or passed to generate.ps1 with -Codegen.

Generate the bindings:

.\generate.ps1 -Codegen ..\..\..\target\release\dynwinrt-codegen.exe

Play synthesized speech:

.\run.ps1 -Python C:\path\to\python.exe -Text "Hello from dynwinrt."

For non-interactive validation, synthesize the stream without playing it:

.\run.ps1 -Python C:\path\to\python.exe -Smoke