There are a set of functions within GHPM that pose problems for testing. All of these functions are modifying the targeted repo in some way via a mutation graphql query. Here are the set of functions:
create_labels()
delete_all_labels()
add_pull_request_reviewers()
create_pull_request()
create_projectboard()
clone_projectboard()
Testing these causes problems as running them on repos like pkgr or rbabylon might get messy. Even considering the fact that a "cleanup" process is initiated to undo the modifications, if a test fails mid modification or after cleanup, it could leave a lot of dangling elements.
The return response for all of these functions is usually a TRUE/FALSE or a list output of the specific data that was modified and any associated values (repo name, user responsible for modification, url, etc...).
There are a set of functions within GHPM that pose problems for testing. All of these functions are modifying the targeted repo in some way via a mutation graphql query. Here are the set of functions:
Testing these causes problems as running them on repos like pkgr or rbabylon might get messy. Even considering the fact that a "cleanup" process is initiated to undo the modifications, if a test fails mid modification or after cleanup, it could leave a lot of dangling elements.
The return response for all of these functions is usually a TRUE/FALSE or a list output of the specific data that was modified and any associated values (repo name, user responsible for modification, url, etc...).