Skip to content

Latest commit

 

History

History
73 lines (54 loc) · 1.5 KB

README.org

File metadata and controls

73 lines (54 loc) · 1.5 KB
Raw

SpeechState starter project

Getting started

  1. You need Node.js (LTS version) and Yarn package manager.
  2. Clone SpeechState (https://github.com/vladmaraev/speechstate) into a sibling directory.
  3. Switch to no-react branch.
  4. Install the dependencies and build SpeechState: 
    yarn
npx tsc
  5. Clone this project.
  6. Link SpeechState (from this project): 
    yarn link ../speechstate -A -p -r
  7. Add Azure key in src/index.ts
  8. Install the dependencies and start the application: 
    yarn
yarn start

Some considerations:

  1. Yarn (update to last version: yarn set version stable, see .yarnrc.yml)
  2. Cleaning parcel cache: rm -rf .parcel-cache

Assignments

Form-based dialogue management

Tutorial

Recap

  • [ ] states
  • [ ] transition
  • [ ] actions
  • [ ] conditions (guards)
  • [ ] context
  • [ ] timers/delays (see SpeechState timers)
  • [ ] history
  • [ ] invoke

example of assign

assign({
  feedback: ({ context, event }) => context.feedback + " " + event.feedback,
})
// Some feedback Some other feedback

Xstate 5 - what’s new?

https://stately.ai/docs/xstate-v5/migration

  • [ ] Actor model
  • [ ] Spawning

Project structure

…and actor model

Parallel states

(GUI example)

Form-filling

https://www.w3.org/TR/voicexml20/#dml2.1.6