Ordered list detection has precedence over title detection

Spec
21

Just to show my face again:

The same client (who came across this issue some time ago and couldn’t figure out why the English looked fine while the German did not) just had the same issue again.

I’m not about to explain escaping characters to non-tech clients. I’ve instructed her to just use the ## style instead, but I have the feeling she had just gotten comfortable with underlining with dashes. It’s a shame this isn’t more intuitive.

1 Like
22

I agree that there are some good reasons to change something here.
BabelMark2 shows three interpretations of your original case:

  1. ordered list, hrule (commonmarks, cheapskate, maruku)
  2. ordered list with header in content (lunamark, parsedown, kramdown)
  3. header with number in content (all the others)

I think 3 is the best interpretation. The question is just how to get the spec to produce it without making the rules complicated or hard to remember.
I’ve already got this on the list of things to figure out before a 1.0 release.

1 Like
23

Agreed. Great; thanks. I look forward to the resolution.

24

It’s not easy to see how to modify the existing spec and parsing strategy to get (3).
I will think about it, but I can’t promise that the resolution will go that way.

25

What about tin-pot’s suggestion above, from Jan 4 and 5? Do they not help in this regard?

26

I think the suggestions were not taking into account that setext headers can be deeply nested. This adds a lot of complexity.

1 Like