Implemented parser.rs (Semi-fixes #161)

[jD91mZM2]

This parser isn't yet hooked up to anything.
It also isn't fully compatible with bash.

It is - however - pretty damn similar.
All the tests are confirmed working, A.K.A.

echo "\"HELLO\", WORLD"lol $TEST. ${LOL!}lol
cat < in > out
echo '<'

All of these work as you'd expect them to.

note: The first one says ${LOL!}lol, which converts to Token::Env("LOL!"), Token::Text("lol").
This isn't actually how bash would do it probably, since that doesn't actually allow non-[a-zA-Z0-9] characters even if you specify them inside { and }. This one does.

note 2: This does not throw any errors (yet). Unquoted strings aren't handled. That could easily be done though, if you let me know how you want to handle them.

[jD91mZM2]

Because GitHub didn't recognize #161 in the title, I'm just going to mention it here so that issue gets a "mentioned in #217" automatic message.

EDIT: There we go

[casey]

Thanks for this! But closing, since the sh dependency will need to be replaced by something with the same features.

[jD91mZM2]

I thought the pull request could be expanded over-time... Whatever