Behaviour for strong emphasis with an asterisk inside

Spec
1

Hey folks,

I’ve read through CommonMark Spec but would love to get a bit of clarification on some edge-case behaviour.

a **\*** b parses as I’d expect:

    <text>a </text>
    <strong>
      <text>*</text>
    </strong>
    <text> b</text>

But as soon as either space is removed (a**\*** b) it no longer parses the emphasis, but still as separate blocks:

    <text>a</text>
    <text>**</text>
    <text>*</text>
    <text>**</text>
    <text> b</text>

Dingus demo

What part of the spec causes this to no longer be treated as correct emphasis?

And relatedly, is there a way to use Markdown syntax (without falling back to just using HTML <strong>) to apply strong emphasis to an asterisk (interestingly a__\*__ b behaves the same way, not producing emphasis)?

2

This is because of flankingness. When you remove the space, the
first ** becomes right-flanking and hence cannot start
emphasis.

Unfortunately the current rules don’t really allow you to express
what you’re going for, without resorting to raw HTML.

3

A workaround could be using a numerical entity instead of a.
This would change the flankingness so that ** is both left
and right flanking.

4

See also this recent GitHub issue submitted against the CommonMark spec: Invalid em + strong rendering in JavaScript implementation

5

Ty for the pointers both!

I wonder if this is worth adding an example/note to the spec for, as it seems like a complex edge-case and feels very similar to Example 415.

6

Almost, but with obvious side effects:

a[**\***]()b
a[**\***][]b
a[**\***]b

[**\***]: <>