Add /api/speech endpoint

[cecheta]

Closes #501, required by #101

Purpose

This PR adds a new /api/speech endpoint, which uses the Speech Key to return a short-lived token that the frontend can use to call the speech service.
The frontend calls the endpoint every time the microphone button is pressed.
When using RBAC, the speech key is fetched directly from Azure as it is not present on any env var.

https://github.com/Azure-Samples/cognitive-services-speech-sdk/blob/master/samples/js/browser/README.md#token-exchange-process

The PR also refactors some of the tests/mocking in test_app.py ,so the mocking of the env_helper is done in one place.

Does this introduce a breaking change?

[ ] Yes
[x] No

Pull Request Type

What kind of change does this Pull Request introduce?

[ ] Bugfix
[x] Feature
[ ] Code style update (formatting, local variables)
[ ] Refactoring (no functional changes, no api changes)
[ ] Documentation content changes
[ ] Other... Please describe:

How to Test

• Deploy application with Keys + with RBAC

What to Check

Verify that the following are valid

• Test that speech-to-text is working

[github-actions]

Coverage Report •

File	Stmts	Miss	Cover	Missing
code
create_app.py	148	3	97%	199, 204, 327
code/backend/batch/utilities/helpers
EnvHelper.py	114	9	92%	188, 193–194, 197–199, 206–208
TOTAL	1724	833	51%

Tests	Skipped	Failures	Errors	Time
72	0 💤	0 ❌	0 🔥	10.305s ⏱️

[ross-p-smith]

I think you've just made the same comment - but maybe some co-ordination with #660. It feels like maybe this goes in first and shows how to get config? What are your thoughts?

[cecheta]

I think you've just made the same comment - but maybe some co-ordination with #660. It feels like maybe this goes in first and shows how to get config? What are your thoughts?

Actually, this PR removes the ability to fetch config from the /api/config endpoint, but now I'm thinking it would actually be better to leave that in, so it can be used in #660

Will update

[superhindupur]

@cecheta so we're on the same page:

• I'm now convinced that injecting the env variable at buildtime into the frontend typescript app in Make the speech recognizer's supported languages configurable. Required by #317 #660 is overly complicating the build-time dependencies.
• I agree that /api/config is a good place to add the speech languages
• Should I go ahead and remove the azureSpeechKey from the /api/config endpoint in my PR so that we don't return the key from there?

[cecheta]

@cecheta so we're on the same page:

• I'm now convinced that injecting the env variable at buildtime into the frontend typescript app in Bhindupur/speech as input #660 is overly complicating the build-time dependencies.
• I agree that /api/config is a good place to add the speech languages
• Should I go ahead and remove the azureSpeechKey from the /api/config endpoint in my PR so that we don't return the key from there?

Sounds good, although I was going to try to remove the secret from the config response in #683 , however I don't mind if you remove it in yours since I haven't started yet

[superhindupur]

@cecheta and I chatted offline - the conclusion was that we will rename the /api/config to /api/speech and return the token from there. Then I will add the languages to it under #660

[adamdougal]

[ross-p-smith]

😎