How does this work?

A scenario is defined in JSON, for example "resolve"

{
  "name": "resolve",
  "input": {
    "did": "did:example:123",
    "did-resolution-input-metadata": {}
  },
  "output": {
    "did-document": {
      "id": "did:example:123"
    },
    "did-document-metadata": {},
    "did-resolution-metadata": {}
  }
}

A JSON object representing the scenario is posted to the server,

A small javascript function processes the request and returns a response:

const assertions = {
  "did should match did-document id": (scenario) => {
    return scenario.input.did === scenario.output["did-document"].id;
  },
};

const testResolve = (scenario) => {
  let assertion_results = {};

  Object.keys(assertions).forEach((assertion) => {
    assertion_results = {
      ...assertion_results,
      [assertion]: assertions[assertion](scenario),
    };
  });

  const test_result = Object.values(assertion_results).every((result) => {
    return result === true;
  });

  return {
    scenario: scenario.name,
    test: test_result ? "PASS" : "FAIL",
    assertion_results,
  };
};

For example:

{
  "scenario": "resolve",
  "test": "PASS",
  "assertion_results": { "did should match did-document id": true }
}

Multiple scenarios can be submitted at once:

curl -s -X POST http://localhost:8080/test \
-H "Content-Type: application/json" \
-d @./packages/did-core-test-vectors/did-method/did-key/test-all.json \
| jq '.'

And all results are produced as json:

❯ curl -s -X POST http://localhost:8080/test \
-H "Content-Type: application/json" \
-d @./packages/did-core-test-vectors/did-method/did-key/test-all.json \
| jq '.'
{
  "test": "PASS",
  "scenarios": [
    {
      "scenario": "resolve",
      "test": "PASS",
      "assertion_results": {
        "did should match did-document id": true
      }
    }
  ]
}

Designing Scenarios

Scenarios can be simple or complex, but they require a vendor to be honest about the inputs and outputs of their system with respect to the did core data model.

A vendor might write extra software to be able to autoamtically generate scenarios for validation.

If a concept can't be represented via a JSON data model and some small javascript program that validates it, it cannot be tested.

Presentation of Test Results

With more time, test results might be better structured for being consumed in 3rd party software.

Run Docker Image

docker run --publish 3000:3000 --detach --name dcts or13/did-core-test-server:1.0

Run Server Locally

npm run docker:up