Build fails for PR update if Force Coverage Improvement

[yardstick17]

Describe the context

• Extension: BuildQualityChecks
• Environment: Azure DevOps Services (cloud)
  • Server version: NA
• Agent type: Microsoft-hosted
  • Agent version: NA
• Pipeline type: yaml pipeline

Describe the problem and expected behavior I am intending to use Force Coverage Improvement to improve coverage over time. This coverage improvement is based on subsequent builds while I want the coverage improvement against a branch ( e.g., develop). If I use this feature enabled - the build works fine If I added new unit tests. In practice you update the PR addressing reviewers comment. Once you do that: coverage is compared against your last build, and if new units are not added in each update - pipeline will fail.

[ReneSchumacher]

Hi @yardstick17,

please look at the docs at https://github.com/MicrosoftPremier/VstsExtensions/blob/master/BuildQualityChecks/en-US/PullRequests.md#pull-request-policy. I believe they describe the scenario you would like to achieve.

The task needs a baseline build for the comparison and this can either be the latest successful build that ran for the same branch on which the task is currently running (e.g., you run a build as part of PR 21; BQC looks for the latest successful build that ran for branch refs/pull/21/merge) or the latest successful build for a specific branch. While the first option is the default, the second one can be configured by setting the baseBranchRef parameter to a fixed value (e.g., refs/heads/develop). For PR scenarios, it is useful to set baseBranchRef to the variable $(System.PullRequest.TargetBranch) to ensure that policies always compare with the latest merged in changes to the PR target branch. If you use the variable outside a PR, the task automatically falls back to the default behavior of comparing with a build for the current branch.

[yardstick17]

Thanks @ReneSchumacher for your prompt reply. baseBranchRef is exactly what we wanted.