Allow an Azure DevOps pipeline to be run against a specific commit

[ValdisThomann]

Background

We run end-to-end tests in an Azure DevOps pipeline at the end of an Octopus deployment. We need to target the specific git commit of the end-to-end tests that align with the packages that have been deployed by the pipeline.

Results

Allows the user to optionally provide a git commit hash/id against which to run the Azure DevOps pipeline. If this parameter is left blank it will default to the latest commit.

Before

After

Pre-requisites

• Id should be a GUID that is not 00000000-0000-0000-0000-000000000000
  • NOTE If you are modifying an existing step template, please make sure that you do not modify the Id property (updating the Id will break the Library sync functionality in Octopus).
• Version should be incremented, otherwise the integration with Octopus won't update the step template correctly
• Parameter names should not start with $
• Step template parameter names (the ones declared in the JSON, not the script body) should be prefixed with a namespace so that they are less likely to clash with other user-defined variables in Octopus (see this issue). For example, use an abbreviated name of the step template or the category of the step template).
• LastModifiedBy field must be present, and (optionally) updated with the correct author
• The best practices documented here have been applied

(Not applicable)

• If a new Category has been created:
  • An image with the name {categoryname}.png must be present under the step-templates/logos folder
  • The switch in the humanize function in gulpfile.babel.js must have a case statement corresponding to it

[CLAassistant]

All committers have signed the CLA.

[github-actions]

Start Hyponome locally

docker pull ghcr.io/hnrkndrssn/hyponome:main
docker run --rm -p 8000:8080 -it ghcr.io/hnrkndrssn/hyponome:main

Review in Hyponome

[hnrkndrssn]

Hi @ValdisThomann, thank you for your contribution.

The change looks good, I just have a question I'd like to confirm.

When passing in both branch and commit, the run will trigger for the commit even if the commit does not exist on the branch passed in, is this something you took into consideration as part of this change, could this create confusion if someone were to specify both branch and commit in their step?