Project model only

[kounelisagis]

We just need one model because:

• Less complex source code (+ easier to implement new functionalities)
• We have one to one correspondence between a project and a repo. Thus, Double information is stored without particular reason. All Repository’s struct fields (ID, CreatedAt, UpdatedAt, RepositoryUrl, Project, ProjectID, IssueCount, Issues, LastParsed, Tags) are also fields of the Project’s struct (maybe with another name) or links to the corresponding project.
• User is only able to see projects (repositories are not visible)

So we need one, but which one?

We use the Project Model and not the Repository Model because: Open Source Contributors know them as “projects” Better compatibility with the current front-end (only /api/projects endpoint available)

Endpoints

• We need no information from:

developer.github.com/v3/projects docs.gitlab.com/ee/api/repositories.html

• And we got everything we need from:

developer.github.com/v3/repos docs.gitlab.com/ee/api/projects.html

Conclusion: GitLab: what we want is called “project”, so we are ok. GitHub: we use repos endpoint. We get the information from there and we just name it as “project”.

[alexdor]

@kounelisagis you have to resolve the merge conflicts by rebasing your changes on top of the current master