Updated the logic for fetching team reviewers in PR edit and create flows. In `pr edit`, team reviewers are always fetched for consistency with existing behavior, with a note to potentially align with `pr create` logic in the future. In `pr create`, team reviewers are now only fetched if a reviewer contains a slash, aligning with behavior before the regression.
Based on PR feedback from @bagtoad, this commit creates a new constant for the Copilot bot name, which is used in the assignee / reviewer selection as well as replacing `@copilot` when going to GitHub.com UI
Thanks to @bagtoad, this commit refactors the argument to `NewCopilotReplacer(bool)` from being where this is used to what it effect is has.
Because there is already precedence for display name being `<login> (<name>)`, I worried there would be confusion for `Copilot (AI)` being display name for assignees and reviewers but `Copilot` when going to GitHub.com UI. Instead, I renamed the argument based on whether the login is returned / replaced.
This commit is primarily focused on fixing the existing `gh issue create` tests with the changes to the underlying queries retrieving the data for resolving metadata.
Additionally, a new test case for `gh issue create --web --assignee @copilot` was added to mirror existing tests. While exercising this capability, I found out the web UI apparently wants `copilot` instead of the `copilot-swe-agent` login, so changes to address that were a part of this commit.
This commit refactors how `gh issue create` and `gh pr create` retrieve information needed to resolve metadata to be more in line with the approach used in `gh issue edit` and `gh pr edit`.
Previously, both commands used `prshared.fillMetadata(...)` function to retrieve assignees, reviewers, labels, and teams outside of `api.RepoMetadata(..)`. Now, these commands will consistently use the same logic and data for resolving metadata.
The `--json` flag accepts a list of GraphQL fields to query for and
output in JSON format. To get the list of available flags, run the
command with a blank value for `--json`. Additional `--jq` and
`--template` flags are available just like in `gh api`.
For metadata types chosen in interactive flow, we fetch all records from
the API in order to be able to display a multi-select interface.
For metadata defined via command-line flags, we resolve records that can
be looked up directly, avoiding fetching the entirety of expensive
datasets (e.g. all members of an organization) if we can.
The new approach ensures efficient fetching when interactive flow is
combined with values from flags.
