Tweaks to unit test

tests/unit/mocks/mockoon.json

@@ -17,7 +17,7 @@
       "responses": [
         {
           "uuid": "eb55d0af-aa59-4d0f-8cea-58ba9c68fa55",
-          "body": "{\"status\":\"success\",\"errors\":null,\"data\":{\"daedalus\":\"externally mocked daedalus model version\",\"daedalus.api\":\"externally mocked R API version\"}}",
+          "body": "{\"status\":\"success\",\"errors\":null,\"data\":{\"daedalus\":\"1.2.3.4.5.6.7.8\",\"daedalus.api\":\"8.7.6.5.4.3.2.1\"}}",
           "latency": 0,
           "statusCode": 200,
           "label": "",

tests/unit/server/api/README.md

@@ -12,10 +12,20 @@ The @nuxt/test-utils/e2e module just mentioned, which has Vitest run the test se
 
 ## Mockoon
 
-To run that separate service, [MSW](https://mswjs.io/) would be a common choice, but it doesn't seem to be compatible with Nuxt (at least server-side): [here](https://github.com/nuxt/test-utils/issues/775) is an example of a GitHub issue for this. So instead we are using [Mockoon](https://mockoon.com/), whose canned responses are stored in /tests/unit/mocks/mockoon.json. To configure the responses, you can either directly edit that file, or use the convenient UI of the Mockoon desktop app.
+To run that separate service, [MSW](https://mswjs.io/) would be a common choice, but it doesn't seem to be compatible with Nuxt (at least server-side): [here](https://github.com/nuxt/test-utils/issues/775) is an example of a GitHub issue for this. So instead we are using [Mockoon](https://mockoon.com/), whose canned responses are stored in /tests/unit/mocks/mockoon.json. To configure the responses, you can either directly edit that file (before re-starting the mock server), or use the convenient UI of the Mockoon desktop app.
 
 To save time writing mock responses, Mockoon is able to record responses from another service, if you follow the steps in [this GitHub comment](https://github.com/mockoon/mockoon/issues/21#issuecomment-1900049410). These recorded responses can then be played back in tests, and edited.
 
 This mocking service has to be up and running for unit tests to pass, and it does not interfere with uses of `registerEndpoint` (which, anyway, works by registering mock endpoints from within the web app, so can't interfere with things served on external ports). By contrast, when we run our end-to-end suite (with Playwright), we want to test that the whole collection of services works properly in a realistic simulation, and for that we must use the real R API service. Thus, we include a check in the end-to-end tests to verify that the mock R API server is not running. We do this by sending a request to a smoke endpoint `/mock-smoke` that should never be implemented by the genuine R API. The same smoke endpoint should be used by tests that do rely on the mocking service, to verify that they are being run against the mocked API.
 
 We should be able to switch away from Mockoon to another solution if desired, by [exporting](https://mockoon.com/docs/latest/openapi/import-export-openapi-format/) the canned responses to OpenAPI format.
+
+### To run the mock server
+
+#### From CLI:
+
+npx mockoon-cli start --data ./tests/unit/mocks/mockoon.json
+
+#### From desktop app:
+
+Make sure the correct API config is selected in the left sidebar (not the demo config). Then click the green "play" button.

tests/unit/server/api/versions.spec.ts

@@ -12,7 +12,7 @@ const nodeFetch = fetch; // Normal 'fetch' from Node
 
 beforeAll(async () => {
   // Verify that the user of the test suite has started the mock server
-  // by checking that the server is listening on localhost:8001
+  // by checking that the server is listening on localhost:8001/mock-smoke
 
   let response;
   try {
@@ -28,15 +28,14 @@ beforeAll(async () => {
 });
 
 describe("api/versions", async () => {
-  // Run the setup function to start the Nuxt server
-  await setup();
+  await setup(); // Start the Nuxt server
 
   it("returns the expected version data", async () => {
     const response = await nuxtTestUtilsFetch("/api/versions");
     const json = await response.json();
 
-    expect(json.daedalusModel).toBe("externally mocked daedalus model version");
-    expect(json.daedalusApi).toBe("externally mocked R API version");
+    expect(json.daedalusModel).toBe("1.2.3.4.5.6.7.8");
+    expect(json.daedalusApi).toBe("8.7.6.5.4.3.2.1");
     expect(json.daedalusWebApp).toMatch(/(\d+\.)?(\d+\.)?(\*|\d+)/);
   });
 });