In my [PR 224](https://github.com/LukeSmithxyz/mutt-wizard/pull/224), I
introduced a bug.  While I was catching SIGINT with the trap statement,
I did not actually exit the script and returned into the shell.

Fixes https://github.com/LukeSmithxyz/mutt-wizard/issues/234

The terminal text color  wasn't reset when the user  aborted the wizard
while being prompted for input.
This resolves https://github.com/LukeSmithxyz/mutt-wizard/issues/224

The current test  did not terminate  the script in cases  when the user
already has a private GPG key,  but they'd not initialized the password
store.
This leads to an infinite loop in the `getpass()` function.

The commit fixes https://github.com/LukeSmithxyz/mutt-wizard/issues/204

This allows mailsync to be called like this `bindsym $mod+e exec --no-startup-id mailsync; exec $term -e neomutt && pkill -RTMIN+12 i3blocks` such that it refreshes your mail when you open mutt without it needlessly syncing if a sync is already running.

Resolves https://github.com/LukeSmithxyz/mutt-wizard/issues/197.

Checking for a secret key with format `<foo@bar.com>` may not work for
people that store key IDs in .gpg-id. Removing the `<` and `>` makes it
work for both beople that use emails and people that use IDs.

If -a is passed along with "$@", mailsync will sync all mailboxes even
if only a single channel is passed in with "$@". This breaks the
semantics of the `o` macro in mutt which should only sync the current
mailbox that is currently being viewed in mutt.