Test results summary: fix JavaScript console error when 'showLink' HTML element is missing

[jayaddison]

The showLink HTML anchor element is only rendered server-side when fewer tests are displayed on the page than exist in the job result.

So if the job passes (zero failures displayed, zero failures exist), or if there are up to ten failures (up-to-ten failures displayed, up-to-ten failures exist), then the showLink HTML element won't be present on the page.

The associated show-failures.js adjunct code attempts to add an onclick event handler to a showLink-ID'd element on the page after the HTML DOM loads:

junit-plugin/src/main/resources/hudson/tasks/test/AbstractTestResultAction/show-failures.js

Lines 29 to 31 in 19a5192

	const showFailuresLink = document.getElementById("showLink");
	showFailuresLink.onclick = (_) => showFailures();

This results in an error appearing in the browser's JavaScript console when the showLink element is not present in the HTML.

The fix suggested here is to only include the relevant JavaScript adjunct when the link itself is also placed on the page.

• Make sure you are opening from a topic/feature/bugfix branch (right side) and not your main branch!
• Ensure that the pull request title represents the desired changelog entry
• Please describe what you did
• Link to relevant issues in GitHub or Jira
• Link to relevant pull requests, esp. upstream and downstream changes
• Ensure you have provided tests - that demonstrates feature works or fixes the issue
  • Not done, however I have reproduced and confirmed the issue locally
    • Checked that the error occurred before the fix was applied
    • Checked that the error is resolved after the fix is applied

Related issues

• Resolves JS Error on build page when there are no test failures #480
• Resolves JavaScript browser console error: showFailuresLink is null #513
• ⚠️ Affects a similar part of the code to [JENKINS-69658] - CSP compatibility for failed-test.jelly #451 -- but does not conflict with it, I don't think (cc @Artmorse)

[timja]

Seems reasonable, thanks!

I assume you manually tested this?

[jayaddison]

You're welcome - and yep, tested manually locally with a sample JUnit XML report.

[timja]

(Triggering build)

[jglick]

@timja this apparently produced a release (jenkinsci/bom#2059) but it is not in Releases?

[jglick]

Hmm jenkinsci/bom#2058 https://github.com/jenkinsci/workflow-api-plugin/releases https://www.githubstatus.com/incidents/pr3498h3qkfy I guess.

[timja]

Hmm jenkinsci/bom#2058 jenkinsci/workflow-api-plugin/releases githubstatus.com/incidents/pr3498h3qkfy I guess.

Yeah

https://github.com/jenkinsci/junit-plugin/actions/runs/4941356327/jobs/8833887199

• gh api -F tag=1202.v79a_986785076 -F message=1202.v79a_986785076 -F object=79a986785076d8abd07e686de315faa77fdb4047 -F type=commit /repos/jenkinsci/junit-plugin/git/tags
  unexpected end of JSON input

---

Manually created now: https://github.com/jenkinsci/junit-plugin/releases/tag/1202.v79a_986785076next