@RyanM Matched by (?:zero|one|two|three|four|five|six|seven|eight|nine|ten)\b(?!.{0,25}three\W++four\W++five\W++six\W++seven\W++eight\W++nine)(?:\W++(?:p\W++)*+(?:zero|one|two|three|four|five|six|seven|eight|nine|ten)){8,}+(?!(?:[^<]|<(?!\/?code>))*+<\/code>) on line 24038 of watched_keywords.txt
Also, as the person who wrote the current version of that regex, I'm sorry ;-)
(it's intended to catch spammers who spell out phone numbers, but mostly avoid catching people printing "one two three four five six seven eight nine" in a loop)
@tripleee was this intentionally not exactly the same format as the original watch?
@JeffSchaller The regex requires that there's at least one non-word character separating the numbers when written written out as words. In this case, they are on separate lines in the Markdown and HTML. When displayed in the post, the newlines are shown as a space. However, in all regex based detections for the why data we strip any newlines which the regex captures. It's currently unclear to me why we strip out the newlines rather than substituting spaces.
@SmokeDetector fp- link is possibly spam, but was associated with a specific user. This user seems to have legit contributions, albeit all within the last day. Giving benefit of doubt.
