在 Azure DevOps 中拉取请求后挂钩或修改文件的任务

Question

I would like to modify some files in my protected master-branch after a pull-request from a feature-branch to the master-branch is done in Azure DevOps Server.

One file needs to be renamed, another file needs to be created and a in a third file there is a string to replace.

Is this possible through server side git hooks or an Azure Pipeline or Task (sorry, I´m new to this topic).

Update

The hook or pipeline should do following:

1. check if a file current.sql has any content (if yes, go to next step)
2. update an integer-variable (DbVersion) in a C#-Class (add 1 to it)
3. rename the current.sql file to db_update_[new DbVersion].sql
4. create a new empty current.sql file

Answer 1

Your scenario description is a bit vague, but I will try to write some answer based on what I understand you want to do:

Generally, it is not good practice to modify your branch in your build pipeline. As mentioned before, you should do that before merging the pull request in to your branch.

For what you describe it might be good to know, that Git handles versioning automatically. Each commit creates a new version of the changed file, which you can find in the commit history in Azure DevOps. It sounds to me as if you basically just don't want to loose the edits you made in your .sql file.

If you, though, require the new version of the file to be part of the build artifact which your pipeline creates, it might make sense to write a short shell script, which you can run via a shell script task . For string replacement, use the "Replace token" task .

Another solution might be a pre-push githook, which would execute a shell script creating the new file? You can read more about githooks here .

I hope you found something useful in this, otherwise it would be good to know what you want to do with the versions that you create. Are you going to use the versions lateron? Maybe it might be better to reconsider a different branching model, which supports multiple environments.