Rogue post author badge

[SeanTOSCD]

Something's happening in the html5_comment() method of Themedd_Walker_Comment class. The post author badge appended to the avatar loses its position-relative parent anchor tag in certain cases. My comment/post author has a website saved to the user profile, so not sure what's happening... didn't look super deep.

Standard display:

Missing badge (it made its way up to the top of the post content):

Note: my admin account is the author of the post and the comment. Same behavior logged in or out.

Might be beneficial to wrap the avatar and badge in their own relative-positioned container no matter what, and let the anchor tag wrap all of it.

[SeanTOSCD]

That's v1.1, by the way. Unable to check master at the moment.

[amdrew]

@SDavisMedia 👍

Can I take a look at this issue on the EDD staging site once ready? Can't replicate locally.

[SeanTOSCD]

@amdrew It's fixed already and was only replicated locally. So I probably won't be able to replicate on staging.

The problem, from what I can tell, is with the first and the last line here: https://github.com/easydigitaldownloads/themedd/blob/release/1.1/includes/class-themedd-walker-comment.php#L73-#L81

The avatar and post author badge are sibling elements:
$avatar = $avatar . $post_author_badge;

Then in the condition below that code, based on whether or not the comment author URL is empty (this is the part I couldn't seem to control for replication), an anchor element wraps the aforementioned avatar and post author badge (here's the opening anchor tag):
printf( '<a href="%s" rel="external nofollow" class="url comment-author-avatar mr-3 position-relative">', $comment_author_url );

Notice that anchor tag has the position-relative class. But based on the entire conditional, there's some sort of case where the anchor tag may not exist. That's what happened to me... I just can't figure out how because the comment author had a URL as far as I can tell.

So what I'm suggesting is that no matter what, the avatar/badge siblings get their own dedicated relative positioned wrap. Even without replication of the issue, just from a markup standpoint, those two elements need a consistent wrap with relative positioning that can never be taken away by an unrelated condition.