Description
I think this was explained by @pshao25 in another issue.
But basically, during all migrations, the original swagger files are deleted as they have been migrated, and the new swagger file is generated. So, at the end, we have a pull request containing only the new swagger file.
Right now, using the tool under the workflow is not very user-friendly. My expectation is that LHS can reference a remote location (repository link to the target branch) and we can run the tool. Of course, the tool can just follow the link, take the files, save them in an internal folder under .gitignore, and work with those swagger files.
However, the fact that people need to keep the files in an additional folder, run the tool, and ensure they never mess up pushing those files makes it feel like the tool can't be easily run with the common-neutral scenario, which is that the original swagger files are no longer there.
Additionally, supporting this scenario will help us in the future if we want to integrate the tool to the pipeline.