Support testing APLA directives

[xavidop]

Description

It is important to support APLA Documents. I saw that you are supporting APL documents.

Steps to Reproduce

Test a resquest of an intent with an APLA directive

Expected behavior: work with APLA documents

Actual behavior: It does not work

Environment

NodeJS-Version: Node.js 10.x

OS: Windows 10

[hoegertn]

Can you provide details on the use case or show a piece of sample code? That would make it easier to follow.

[xavidop]

You can test it adding this piece of code on the file lambda/test.js in the project https://github.com/xavidop/alexa-skill-pokedex

  describe('LaunchRequest', function() {
    alexaTest.test([
      {
        request: new test.LaunchRequestBuilder(skillSettings).build(),
        saysLike: i18n.t('WELCOME_MSG'), repromptsNothing: false, shouldEndSession: false,
      },
    ]);
  });

[s-maheshbabu]

@hoegertn
APL-A support can be added to the existing APLValidator by adding a check for APL Audio documents.

const renderedDocument = response.response.directives.find((value) => value.type === 'Alexa.Presentation.APL.RenderDocument' || value.type === 'Alexa.Presentation.APLA.RenderDocument');

Or we can create a new validator altogether: APLAValidator.ts

What do you think?

Here is an APL-A snippet in case you want to read through.

[xavidop]

Hello team!

any updates of the PR?

Xavi

[habuma]

Was just trying this out and it's a nice start in that it helps me assert that the response includes the expected document and data source. But...and I know that I'm asking for a big thing here...permit me to think aloud...

It would be AMAZING if I could write the test to render the APL-A document into text and then compare that to expected text. Something like this:

renderDocument: {
    token: 'TEST_TOKEN',
    document: (doc) => {

        // render document into string
        // compare rendered document with expected speech

    },
    hasDataSources: {
      ...
    },
},

Of course, the rendered APL-A could include sound files, so...I'd think that the best that could be done is to render the "Speech" elements into flat text...maybe including audio in the resulting text as something like "[audio 'soundbank://soundlibrary/water/nature/nature_08']" so that it can also be checked for from the expected response.

What's missing here is that APL-A renderer. Does such a thing already exist (ideally as a module that can be imported and used easily here)? If not...I wonder how difficult it would be to create such a renderer. (APL-A is relatively simple as compared to APL, but it still could get tricky I suppose.

The ask dialog command kinda does this. It receives the APL-A template and datasource in the response and renders it into text that is emitted in the shell. Maybe that could be reused or duplicated here? (My next step: Go look at the source behind ask dialog to see how it does this.)

[hoegertn]

Happy to look into this if you can provide a PR. Currently, I don't have the bandwidth to implement this.