Fix issues for math span calculation

[toothache]

A follow up fix for #779 . Math span end position should also consider the situation when the math inline is closed with whitespace.

[xoofx]

I'm not sure what exactly the fix for set text heading is trying to address. The following seems to be correct for CommonMark implementations, including markdig.

[toothache]

I'm not sure what exactly the fix for set text heading is trying to address. The following seems to be correct for CommonMark implementations, including markdig.

markdig is parsing the second set text heading with an extra character as heading. But it shouldn't, correct?

As shown below, commonmark and markdig have different behaviors parsing the second set text heading.

...
commonmark-java 0.13.0Java
commonmark.js 0.28.1Javascript
league/commonmark GFM 2.4.0Php
league/commonmark 2.4.0Php
...

<h1>
  Hello
</h1>
<p>
  Test ===n
</p>

Results 1 implementation#4
markdig 0.35.0.0C#

<h1>
  Hello
</h1>
<h1>
  Test
</h1>

[xoofx]

markdig is parsing the second set text heading with an extra character as heading. But it shouldn't, correct?

Sorry, completely misread babelmark results. You are absolutely correct.

[xoofx]

Thanks!

[toothache]

Thanks!

Hi, @xoofx . Is it possible to release a new version containing the fix? Thanks!