Update the on-device speech recognition methods

[evanbliu]

Closes #141

This PR contains the following changes:

• Introduce a new AvailabilityStatus enum to indicate the availability of on-device speech recognition for a given language.
• Update the availableOnDevice method to return a Promise that resolves to an AvailabilityStatus instead of a boolean.
• Rewrite the availableOnDevice and installOnDevice methods in the modern, algorithmic style.

---

Preview | Diff

[evanbliu]

@padenot - Can you please take a look at this PR when you get a chance? Thanks!

[padenot]

Sorry to be picky, but I think it's good to get things right from the get-go to get the ball rolling on the first few PRs. Apart from the "when do we know downloading has finished", my comments are on the form rather than the substance of the PR, but unfortunately that matters in normative writing.

[padenot]

Way better, still some comments.

How does one know that the model has finished downloaded?

[evanbliu]

Way better, still some comments.

How does one know that the model has finished downloaded?

From above:
Right now sites will have to poll availableOnDevice to find out if the download has finished.

What are your thoughts on either:

1. Introducing a new languageinstalled SpeechRecognition event.
2. Introducing a new API that returns a Promise that resolves to true once the language is finished or resolves to false if the installation failed.
3. Updating installOnDevice to resolve to true once on-device speech recognition is installed rather than resolving to true when the installation is initiated.
4. Just require websites to poll availableOnDevice.

[beaufortfrancois]

FYI https://github.com/webmachinelearning/prompt-api?tab=readme-ov-file#download-progress may be of interest for this group.

[evanbliu]

Gentle ping -- @padenot - Do you have any additional comments on this CL, or on how the API for detecting if the download has finished should look like?

[evanbliu]

Gentle ping -- @padenot - Do you have any additional comments on this CL, or on how the API for detecting if the download has finished should look like?

I'd like to propose option 3: update installOnDevice to resolve to true once on-device speech recognition is installed rather than resolving to true when the installation is initiated.

The Promise will resolve to false if the download fails for any reason (i.e. user rejects the download, network issues, unsupported language) and resolve to true once the language is installed and ready. I don't think the delay should be much of an issue because in most cases the download would complete relatively quickly. Of course, there will be some cases in which the download takes much longer, but the current installOnDevice API could also take a long time to resolve to True if the user takes awhile to respond to a prompt, so websites will need to handle this delay already anyways.

[padenot]

Yes, I'd tend to agree. We could imagine having a UA-side progress indicator here.

[evanbliu]

Sounds good, I've updated the spec to reflect this. Let me know if you have any other comments or concerns on this PR!

[padenot]

Looks good, some comments, but nothing too important. If the answer to the last sentence of the 3rd comment is yes, this is good for merging w/ the small adjustment I mentioned.