A new GitHub feature landed where the API client can specify the desired
name of the new fork. This avoids the necessity of subsequently having
to rename the forked repo after the fork operation has created one.
For backwards compatibility, the renaming logic is still here, but
activates only if the resulting repo name is not the desired name.
The IsPrivate field of "internal" repositories is always true, but those
repositories aren't truly private and absolutely can be forked. We
shouldn't be checking for platform permissions in the client anyway, so
let's just drop this check and have the platform decide whether this is
a valid operation.
We used to do the equivalent of `rootCmd.SetOut(os.Stdout)` because we
thought that Cobra's "Out" stream represents standard output. However,
upon closer inspection it turns out that this is Cobra's stream for
usage errors and deprecation warnings, and those we want written to
stderr as well. It is not clear to me why Cobra maintains a distinction
between "Out" and "Err" streams since both seem to go to sdterr by
default.
This change also ceases our usage of `command.Print()` functions in
favor of explicitly writing to `IOStreams.Out/ErrOut`.
This completely rewrites the PR lookup mechanism so that the caller
must specify the GraphQL fields to query for each PR. Additionally, this
fixes some export problems with `pr view --json`.
Features:
- Each pr command now gets assigned a concept of a Finder. This makes it
easier to stub the PR in tests without having to stub the underlying
HTTP calls or git invocations.
- `pr view --web` is much faster since it only fetches the "url" field.
- `pr diff 123` now skips a whole API call where a whole PR was
unnecessarily preloaded just to access its diff in a subsequent call.
- PullRequestGraphQL query builder is now used to construct queries.
- A bunch of individual commands are now freed of having to know about
concepts such as BaseRepo, Branch, Config, or Remotes.
This removes sensitivity to the BROWSER environment variable in tests
and makes it easier to verify the URL that the browser was invoked with
without having to stub sub-processes.
Either InterruptErr or SilentErr will be present when the user has
chosen "Cancel" or pressed Ctrl-C in prompts. We don't want the recovery
mechanism to kick in these cases because the cancellation was likely
willingly initiated by the user.
