Add a test for non-ascii char input escaped with C-v

[tompng]

There was a mystery of 0x16 in Reline. There is no test and no source code comment.
0x16 seems to be inserted in some case (what case?) in some way (how?) but nobody knows how and why.

(c == 0x16 && @input.tty? && @input.raw(min: 0, time: 0, &:getbyte)) || c

It turned out that this is 0x16 came from a configuration of terminal emulator Escape non-ASCII input with Control-V
(ohmyzsh/ohmyzsh#7412)

With this configuration, every non-ascii byte will be escaped with \C-v

irb> STDIN.raw{STDIN.readpartial(100)} # input あ = "\xE3\x81\x82"
=> "\x16\xE3\x16\x81\x16\x82"

# Bracketed paste also includes \x16
irb> print "\e[?2004h"; STDIN.raw{STDIN.readpartial(100)+STDIN.readpartial(100)} # paste あ
=> "\e[200~\x16\xE3\x16\x81\x16\x82\e[201~"

Now we can add a test and a comment

[tompng]

Ideally, I think changing yamatanooroti's write to not do this conversion is better.
But if we do that, there will be no way to input meta key to yamatanooroti-windows and yamatanooroti-vterm in a common way.