there's no `loose_item_end` token

[notslang]

Just wanted to mention that I found this pretty odd since we have a loose_item_start. Any particular reason for this?

[notslang]

Actually, it's probably more accurate to say that loose_item_start is terminated by list_item_end, which is equally weird. Right now I'm using this code as a workaround:

  # there is no `loose_item_end` token, so we need to make it. thankfully, it's
  # pretty easy to determine where they go, by finding `list_item_end` wherever
  # there's an unclosed `loose_item`
  openLooseItem = false
  ast = ast.map (token) ->
    if token.type is 'loose_item_start'
      openLooseItem = true
    else if token.type is 'list_item_end' and openLooseItem
      openLooseItem = false
      # replace the list_item_end with loose_item_end
      token = type: 'loose_item_end'
    return token

[UziTech]

@slang800 Is there any reason there needs to be a loose_item_end? Can you give an example of where it would be helpful to know at the end of a list item whether it was loose or not?

[notslang]

It's been awhile since I worked with marked, so I don't remember exactly what I was doing, but I'm pretty sure I had a function that looked for every token that was a *_start or *_end and matched them together to form a simple syntax tree.

The fact that loose_item_start isn't terminated by a loose_item_end adds extra complexity to those types of functions. Rather than just looking for tokenType + "_start" or tokenType + "_end", I need to add a special case that handles loose_item_* tokens differently.