4 or more asterisks causes differing results

[erusev]

4 or more asterisks causes differing results

[erusev]

What is you understanding based on? The spec says nothing about the matter and the original parser would parse the text the same way Parsedown would - ****test**** to <p><strong><em>*test</em>*</strong></p>. The same goes for the parser at StackOverflow.

[gitdexby]

Sorry for the delay. The 4-asterisk issue could possibly be dealt with, but the 5 asterisk issue results in invalid HTML - it's actually html in the first example with 4 asterisks, but I mistakenly put it there instead of on the 5 asterisks. If you take the HTML generated from the 5 asterisk example and put it into a simple document to be validated at w3c.org, you'll see the failures. At the very least, the parsing should result in valid HTML.

<p><strong><em></strong>test<strong></em></strong></p>

I came across this issue when I was working on an HTML parser in order to convert some formatting over to PDFs via parsing the stream of HTML. When the nesting and tag matching is invalid as described with the 5 asterisks, then parsing the stream becomes unreliable and fails.

I'd like to use Parsedown because it is so much faster, but without valid HTML as the result, it becomes unpredictable so I had to switch over to michelf's library which just results in:

*_test_*

[erusev]

I'll see what I can do.

[erusev]

Should be resolved in the latest release latest release.

[hkdobrev]

The related commit has introduced a regression and the Travis CI build failed: https://travis-ci.org/erusev/parsedown/builds/14410969

[hkdobrev]

The related commit has introduced a regression

OK, it is not a regression, but just the new tests for this fix are not passing.

[erusev]

There was a test that wasn't passing. The latest release resolves this.