Is it possible to get TODO list item parsing for a custom language plugin outside of Comments.

I have tried everything to allow TODO parsing outside of comments but the information in the BaseFilterLexer derived class seems to be ignored. No matter what I try, only TODO's in comments are accepted:

public void advance() {
    scanWordsInToken(UsageSearchContext.ANY, false, false);
    advanceTodoItemCountsInToken();
    myDelegate.advance();
}

I even tried using a simple plain text lexer, which returns each line in the source as a COMMENT element type, for my TodoIndexer but saw no difference. Only comment TODO items are recognized.

The reason I would want TODO outside of comments is that Markdown does not really have comments defined. You can use <!--   --> as an inline or HTML block element but this is somewhat counter-intuitive and lots of extra typing.

What am I missing or is this functionality hard-coded and ignores the parameters passed to scanWordsInToken?