Incorrect Display of Test Failures in Merge Requests

Summary:

According to the GitLab documentation (https://docs.gitlab.com/ee/ci/testing/unit_test_reports.html), the Test Summary panel in a Merge Request should differentiate between new and existing test failures by comparing JUnit report XML files from the base and head branches. However, the comparison feature appears to malfunction, causing all test failures to be marked as new, even when they exist in the base branch.

Steps to Reproduce:

• Create a main branch with a junit.xml file containing 2 test failures.
• Run the pipeline on the main branch and ensure artifacts (including the junit.xml file) are saved.
• Create a feature branch from main and modify the junit.xml to include the same 2 failures plus an additional new failure.
• Run the pipeline on the feature branch and open a merge request to main.

Expected Results:

The Test Summary panel in the MR should display: 2 existing failures (test cases that failed on both the base and the head branches). 1 newly failed test (test case that passed on the base branch and failed on the head branch).

Actual Results:

The Test Summary panel incorrectly displays all 3 failures as new, suggesting that the comparison between the base and head branches is either not occurring or not functioning correctly.

This issue may lead to confusion and misinterpretation of test results, impacting the development workflow and the ability to accurately track the stability and health of the codebase.