Your checklist for this pull request

🚨Please review the guidelines for contributing to this repository.

• Ensure all commits include DCO sign-off.
• Ensure that your contributions pass unit testing.
• Ensure that your contributions contain documentation if applicable.

Description

In preparation for refactoring how Git commands are executed by this gem (see #617), add two new classes: Git::FailedError and Git::SignaledError to indicate different failure conditions when running the Git CLI.

• Git::FailedError is raised when the git command returns a non-zero exitstatus.
• Git::Signaled is raised when the git command fails because of an uncaught signal (like a kill signal).

Both of these new errors derive from Git::GitExecuteError so existing code that was rescuing this error will still work as before.

Both of these new errors expect to be initialized with an instance of another new class added in this PR: Git::CommandLineResult. Objects of this class capture:

• the git command executed
• the resulting Process::Status
• the command's stdout
• the command's stderr

The intent is to give callers better context as to the type of error (non-zero exitstatus vs. uncaught signal) as well as provide better context as to the command executed, the ending status of the process (including exitstatus or signal #), and the captured command line output.

Git::FailedError has been fully integrated into Git::Lib#command. Any situation where a Git::GitExecuteError would have been raised before now raises a Git::FailedError. As mentioned before, since Git::FailedError is a subclass of Git::GitExecuteError, no change is required to client code.

Git::SignaledError has not yet be integrated into Git::Lib#command.