@SmokeDetector This MS post also advertises this same service. It also uses a shortened URL to provide a link to the site. However, it uses a different URL shortening service.
Perhaps we need a detection which fetches the URL from the link shortening service and applies our URL rules (e.g. including blacklist and watchlist) against that URL.
@SmokeDetector This MS post also advertises this same service. It also uses a shortened URL to provide a link to the site. However, it uses a different URL shortening service.
Unfortunately, I replied to the wrong SD report for my previous message about this.
@SmokeDetector However, the other one uses a different parameter for hop (which is, presumably, being used to track clicks). This one is gretat, while the other is misard.
@Machavity Unless they post it in another answer, to me it looks more like someone trying to be helpful, rather than spam. You could edit it out, if you feel it's an issue.
[ SmokeDetector ] SmokeDetector started at [rev b98a50e (SmokeDetector: Auto watch of ^lb[qg]t\W by tripleee --autopull)](//github.com/Charcoal-SE/SmokeDetector/commit/b98a50e) (running on teward/Solar Flare)
@tripleee I think you wanted this to be LGBT, not LBGT. This was also an issue with your recent !!/watch. If you're wanting that to be a watch of only usernames, it's probably good to also limit the overall length of the string. Thus, the watch could be ^(?=.{0,40}$)l[qg]bt (i.e. the string must be <= 40 characters, which is the max for a username). The \W is inherent in the bookending \b, unless you don't want to match just ^lgbt$, or ^lgbt\W\W+.
@tripleee Probably. While my impression is that bodies begin with an HTML tag, title's don't. Some titles will still match ^lgbt. Limiting it to <40 characters prevents a reasonably significant number of FP on titles. That's based on what we've seen for the "TV" watch, which is intended for usernames, but which also gets FP on titles.
@tripleee Your watch doesn't say exactly 5. Exactly 5 would be ^lgbt\W$. What you watched was ^l[qg]bt\W, which is >=5. Saying >=5 and <=40 isn't redundant.
@Makyen no, it says "the first character must be l, the second must be g, the third must be b, the fourth must be t, the fifth must be non-word"; there is no restriction on how long the string can be, but the regex will only ever examine the first five characters (and most often only the first)
I see now that you are saying "the remainder of the string can be max 35 characters" as an additional condition but really, that seems like an unnecessary extra fence
@tripleee In this case, it's probably not going to save us that much wrt. FP, given that there are very few SE posts with that string anywhere in the title.
Poor grammar and formatting, but it says "here's a static canonical link <canonical link to his website> and here's a dynamic one <php to generate canonical link for any website>"
@Makyen That pattern looks like it's already caught by Pattern-matching website in body and Pattern-matching website in answer; append -force if you really want to do that.
@tripleee That pattern looks like it's already caught by Bad ns for domain in answer and Bad ns for domain in body; append -force if you really want to do that.
@tripleee Blacklisted 2conv\.xyz
CI on fded697 succeeded. Message contains 'autopull', pulling...
[ SmokeDetector ] SmokeDetector started at [rev 1348909 (SmokeDetector: Auto blacklist of muh[ae]mm[ae]d\Wthe\Wp[ae]dophile by Zoe --autopull)](//github.com/Charcoal-SE/SmokeDetector/commit/1348909) (running on teward/Solar Flare)
I have been on StackOverflow for a few months by now, and yesterday I bumped upon this question: click
Now I get that the question of this user is about iframes, but they are referencing twice to a site which seems to be a streaming site. The linking itself seems unnecessary to me, so to me it c...
@Makyen Added [a-z_]*(?:1_*)?844[\W_]*617[\W_]*9531[a-z_]* to watchlist
[ SmokeDetector ] SmokeDetector started at [rev da2217f (SmokeDetector: Auto watch of [a-z_]*(?:1_)?844[\W_]*617[\W_]*9531[a-z_]* by Makyen --autopull*)](//github.com/Charcoal-SE/SmokeDetector/commit/da2217f) (running on teward/Solar Flare)
I just stumbled upon this answer:
Southeast Asian Hokkien Is Spoken In Southeast Asia Where There Are Fujianese Migrants In The 17th Century It Is Distinct From Taiwanese Hokkien (台湾福建語) And Singaporean Hokkien (シンガポル福建語) And Also Penang Hokkien (ペナン口福建語) These Hokkien Variants Are Similar Bu...
@SmokeDetector Does not attempt to answer the question. Talking about solar here is only marginally related to the question. It's clearly spam (undisclosed association; unsolicited promotion) that has just been tossed on as an answer to a question that the spammer felt was remotely related (it's not).
@tripleee That pattern looks like it's already caught by Bad ns for domain in answer, Pattern-matching website in body, Pattern-matching website in answer, and Bad ns for domain in body; append -force if you really want to do that.