SWC-6531: refactor SWC e2e helpers to use exposed SRC methods and remove duplicate code #5178
Conversation
…ove duplicate code
| export enum BackendDestinationEnum { | ||
| REPO_ENDPOINT, | ||
| PORTAL_ENDPOINT, |
There was a problem hiding this comment.
Replicate BackendDestinationEnum, so that the endpoint can be passed to doGet, etc... without evaluating the page to access SRC.
| export async function waitForSrcEndpointConfig(page: Page) { | ||
| // window only available after page has initially loaded | ||
| await navigateToHomepageIfPageHasNotBeenLoaded(page) | ||
| // ensure that endpoint config is set, | ||
| // ...so API calls point to the correct stack | ||
| await expect(async () => { | ||
| const response = await page.evaluate('window.SRC.OVERRIDE_ENDPOINT_CONFIG') | ||
| expect(response).not.toBeUndefined() | ||
| }).toPass() |
There was a problem hiding this comment.
Method moved from e2e/helpers/verification.ts, so it is colocated with other methods that rely on evaluating the page to access SRC
| const srcEndpoint = await SRC.SynapseEnums.BackendDestinationEnum[ | ||
| endpoint | ||
| ] | ||
| // @ts-expect-error: Cannot find name 'SRC' | ||
| return await SRC.HttpClient.doPost( | ||
| url, | ||
| requestJsonObject, | ||
| accessToken, | ||
| srcEndpoint, | ||
| additionalOptions, | ||
| ) |
There was a problem hiding this comment.
Maintain a doPost method that evaluates the page to call the SRC.HttpClient.doPost method to reduce code duplication in API calls. Plan to remove this method (and doGet and doDelete) once SynapseClient is its own package that can be used directly instead
| function getUserIdFromJwt(token: string) { | ||
| const payload = JSON.parse( | ||
| Buffer.from(token.split('.')[1], BASE64_ENCODING).toString(), | ||
| ) | ||
| return payload.sub | ||
| } |
There was a problem hiding this comment.
Moved from e2e/helpers/http.ts, since method is only used in this file
| accessToken?: string, | ||
| userName?: string, | ||
| apiKey?: string, | ||
| page: Page, |
There was a problem hiding this comment.
Instead of passing a Page around, you could consider configuring the test process to point to the correct stack using an environment variable, probably by reading the Maven settings.xml so there's just one source of truth. Then I think a lot of this code could run without requiring the browser and loaded page.
You wouldn't have confidence that the SRC.OVERRIDE_ENDPOINT_CONFIG is properly set, but I don't think that's such a big deal.
What do you think?
There was a problem hiding this comment.
I like that idea for looking up the correct stack! Since we'll still need to evaluate the page to use SRC's HttpClient and SynapseClient, I might hold off on implementing that for now. However, I've opened a ticket to track the work and capture this idea!
No description provided.