Turns out the "standard" way of wrapping errors in Go is via
`fmt.Errorf("%w")`, which doesn't require an external package and also allows a
finer control of error sentence formatting.
Also define a handful of utility methods:
- `New(owner, repo)`
- `FullName`: the name slash owner pair
- `FromFullName`: parse the name slash owner pair
- `FromURL`: parse a GitHub.com URL
- `IsSame(r1, r2)`: compare two repositories
- The local git remotes are scanned and resolved to GitHub repositories
- The "base" repo is the first result resolved to its parent repo (if a fork)
- The name of the default branch is read from the base repo
- The "head" repo is the first repo that has push access
If multiple templates are found, the user is prompted to select one.
The templates are searched for, in order of preference:
- issues:
1. `.github/ISSUE_TEMPLATE/*.md`
2. `.github/ISSUE_TEMPLATE.md`
3. `ISSUE_TEMPLATE/*.md`
4. `ISSUE_TEMPLATE.md`
5. `docs/ISSUE_TEMPLATE/*.md`
6. `docs/ISSUE_TEMPLATE.md`
- pull requests:
1. `.github/PULL_REQUEST_TEMPLATE/*.md`
2. `.github/PULL_REQUEST_TEMPLATE.md`
3. `PULL_REQUEST_TEMPLATE/*.md`
4. `PULL_REQUEST_TEMPLATE.md`
5. `docs/PULL_REQUEST_TEMPLATE/*.md`
6. `docs/PULL_REQUEST_TEMPLATE.md`
The filename matches are case-insensitive.
This stubs stderr separately from stdout in command tests (before those
streams were combined) and improves test assertions around output.
Additionally, no longer use the `cmd.Print*()` family of Cobra functions
because their name sounds like the text will go to stdout, but they
write to stderr instead. Use the more explicit `cmd.ErrOrStderr()` as
output destination instead.
