1. The reader should be configurable to highlight and read each word, phrase, paragraph or sentence by the app.
2. The Screen Reader does not need Text-to-Speech. This will be supplied by the main app. The Screen Reader should highlight and then pass the current word, phrase, sentence or paragraph to the main app via a hook function.
3. The screen reader should be configurable to display some kind of media control (play, pause, stop) or to be controlled from outside the webview environment via hook functions.
4. The user should have a way of setting a new start point for speech, using using touch.
I have provided an partially working example html page and script that demonstrates some of what I want.
You may fix and enhance the example code or rewrite completely.