@SmokeDetector Not FP, due to repeated characters, and poorly formed HTML; along with NAA. --- The "preview" and "text" tab views show that the post doesn't display correctly, the "Markdown" tab view shows that it's parsed incorrectly by our software; triggering these weights incorrectly:
Few unique characters in answer (88), Mostly punctuation marks in answer (37), Repeating characters in answer (80) --- the first isn't correct, the second is borderline, the third is understandable user error; though unnecessary.
@Rob It's not parsed incorrectly by our software. SD deals with the HTML, as delivered by SE. The detections were accurate for the HTML which was delivered by SE.
Could we try to detect that the post's author incorrectly formatted their Markdown and their dump of HTML code was stripped by SE, due to not complying with the requirements for HTML in SE posts, yes, we could, at least to an extent. However, there's considerable difficulty in accurately discriminating between a post with incorrectly formatted code and a spammer/troll trying to post HTML with elements and/or attributes which contain things SE doesn't permit and/or which are unsafe.
@Makyen Thanks for your reply, true it's a lot of trouble to parse; and determine if it's bypassing SE. --- I still think it is correct that it was caught, it's not a FP, and that the ❌ is incorrect.
