Add customization options for `typeof` formatting

[pineapplemachine]

I prefer to write typeof expressions like typeof(value), rather than typeof value or typeof (value), but if there is a configuration option to allow or enforce the former I wasn't able to find it. dprint wants to reformat places where I wrote typeof(value) as either typeof value or typeof (value), depending on value, but I would prefer to enforce the former as the convention.

[pineapplemachine]

A note in case it's helpful to anyone else:

I've noticed that I can get almost the behavior I want by editing gen_unary_expr in src/generate.rs like so, but I'm not familiar with the code and am so far drawing a blank on how to only add the enclosing paren tokens if the operand didn't already have them (also, of course, this would need to be gated behind a config option, and I think gen_type_query in the same source file would need a corresponding change for consistency):

fn gen_unary_expr<'a>(node: &UnaryExpr<'a>, context: &mut Context<'a>) -> PrintItems {
  let mut items = PrintItems::new();
  items.push_sc(get_operator_text(node.op()));
  if node.op() == UnaryOp::TypeOf { // Newly added
    items.push_sc(sc!("("));
  }
  items.extend(gen_node(node.arg.into(), context));
  if node.op() == UnaryOp::TypeOf { // Newly added
    items.push_sc(sc!(")"));
  }

  if get_should_use_parens(node) {
    items = surround_with_parens(items);
  }

  return items;

  fn get_operator_text(op: UnaryOp) -> &'static StringContainer {
    match op {
      UnaryOp::Void => sc!("void "),
      UnaryOp::TypeOf => sc!("typeof"), // Removed a space
      UnaryOp::Delete => sc!("delete "),
      UnaryOp::Bang => sc!("!"),
      UnaryOp::Plus => sc!("+"),
      UnaryOp::Minus => sc!("-"),
      UnaryOp::Tilde => sc!("~"),
    }
  }

  fn get_should_use_parens(node: &UnaryExpr) -> bool {
    if let Node::BinExpr(parent) = node.parent() {
      return matches!(parent.op(), BinaryOp::In | BinaryOp::InstanceOf);
    }
    false
  }
}

Edit: I since found surround_with_parens as a probably better alternative to the separate if checks, and I think I could use a function similar to should_add_parens_around_expr to check for existing parens? Will continue to poke into this later, and make a PR if I can figure it out.