Local token provider for r11s driver

[tanviraumi]

Follow up of #3954. This PR adds a local token provider for r11s drivers to use. Following the model of InsecureUrlResolver, this one is called InsecureTokenProvider. The provider is meant to be used for testing and local running purposes.

InsecureUrlResolver still signs a jwt token. Next PR will remove the signing part from and drivers will only rely on token provider.

[msfluid-bot]

■ @fluidframework/base-host: No change

Metric Name	Baseline Size	Compare Size	Size Diff
main.js	176.35 KB	176.35 KB	■ No change
Total Size	176.35 KB	176.35 KB	■ No change

■ @fluid-example/bundle-size-tests: No change

Metric Name	Baseline Size	Compare Size	Size Diff
container.js	239.58 KB	239.58 KB	■ No change
sharedString.js	225.58 KB	225.58 KB	■ No change
Total Size	465.15 KB	465.15 KB	■ No change

---

Baseline commit: 85293a9

Generated by 🚫 dangerJS against c4ba2ea

[SamBroner]

Yes!

[SamBroner]

Nit: can we reuse the other token provider?

[tanviraumi]

Which one?

[curtisman]

I don't quite like the fact that test-runtime-utils is taking a dependency to routerlicious-driver package. Feels like it belongs to some routerlicious specific thing.

[tanviraumi]

I don't quite like the fact that test-runtime-utils is taking a dependency to routerlicious-driver package. Feels like it belongs to some routerlicious specific thing.

Totally agree. I thought about it but did not find a better place. r11s-driver bundles the interfaces and implementations in one single package. And driver-base seems to be shared between both drivers. Any recommendation on placing the ITokenProvider interface?