@iBug hmmm? if you attempted to provide feedback on the posts you submitted yourself, please don't -- we generally recommend reporting and reviewing to be done by different people
[ SmokeDetector | MS ] URL-only title, bad NS for domain in body, bad NS for domain in title, bad keyword in body, bad keyword in title, +6 more: supplementlab.org/primal-growth/ by crvfkgcuh on askubuntu.com (@ThomasWard @quartata @TetsuyaYamamoto)
we went through upgrading the project last year, it was less tortured than we imagined, though I wasn't directly involved in the actual porting or testing effort
@micsthepick the regular expression library in the Python standard library is called re but we use a replacement with more features which is hopefully going to replace re one day
@iBug yeah so at some point somebody who was still on Python 2 did pip install regex and you have it in your site-packages under /lib/something/python2/something but this has not been done for python3 so you don't have it in /lib/something/python3/something
@micsthepick long story, the one I'm thinking of is the ability to say L<list> to have a Python list interpolated as (?:list member 1|list member 2|etc)
@iBug yeah but if you are in a Python3 virtualenv that's still just pip
... he says, and finds himself more or less forced to reboot in order to finish upgrading to High Sierra even though I'd really have better things to do right now
I've been meaning to look into that, should the blacklisted websites regex be better anchored to not just match on substrings? but some of our patterns probably rely on this now
[ SmokeDetector | MS ] URL-only title, bad keyword in body, bad keyword in title, bad keyword in username, blacklisted website in body, +4 more: healthsupplementzone.com/testro-t3/ by Testro T3 on askubuntu.com (@ThomasWard @quartata @TetsuyaYamamoto)
[ SmokeDetector | MS ] URL-only title, bad NS for domain in body, bad NS for domain in title, bad keyword in body, bad keyword in title, +6 more: supplementlab.org/primal-growth/ by merdikukki on askubuntu.com (@ThomasWard @quartata @TetsuyaYamamoto)
Title - Position 58-68: maxx-boost Title - Position 37-57: wellness4healthy.com Title - wellness4healthy.com NS suspicious dns2.namecheaphosting.com.,dns1.namecheaphosting.com. Title - Position 26-57: http://www.wellness4healthy.com Body - Position 4-14: Maxx Boost, Position 59-69: Maxx Boost, Position 581-591: maxx-boost, Position 652-662: maxx-boost Body - Position 560-580: wellness4healthy.com, Position 631-651: wellness4healthy.com Post - wellness4healthy.com NS suspicious dns2.namecheaphosting.com.,dns1.namecheaphosting.com.
@micsthepick That pattern looks like it's already caught by Potentially bad keyword in answer, Potentially bad keyword in body, Bad ns for domain in body, and Bad ns for domain in answer; append -force if you really want to do that.
@micsthepick That pattern looks like it's already caught by Bad ns for domain in body and Bad ns for domain in answer; append -force if you really want to do that.
Even then this is a dubious, hackish solution which should come with a big fat ugly comment with a warning message telling you how this will bite your shapely lower abdomen, and another saying "I told you so" for each time this happened. — tripleee39 mins ago
I'm expecting a literary award of some sort
phew, that High Sierra upgrade really was a tiresome experience, is my wifi this slow or did it download a terabyte?
@micsthepick if something is acceptable by the criteria and Smoke Detector refuses because some stupid reason then yes, but in the general case, understand why it refuses
[ SmokeDetector | MS ] Bad NS for domain in body, bad keyword in body, bad keyword in title, blacklisted website in body, link at end of body, +2 more: Alpha Prime Elite If you find yourself by Evelyncravend on askubuntu.com (@ThomasWard @quartata @TetsuyaYamamoto)
@micsthepick sorry, that's a side effect whenever we move messages, I don't know of a way to avoid or control that (nor have I ever seen any justification for this behavior, let alone documentation, though I suppose it makes sense for moving discussions between rooms generally)
[ SmokeDetector | MS ] Bad keyword in body, bad keyword in title, blacklisted website in body, link at end of body, pattern-matching website in body, +1 more: Rapid Lift Fx - Review by ariaampert on meta.SE (@iBug @TetsuyaYamamoto)
[ SmokeDetector | MS ] URL-only title, bad keyword in body, blacklisted website in body, blacklisted website in title, link at end of body, +3 more: brainfireadvice.com/enhance-mind-iq/ by susnbaker on askubuntu.com (@ThomasWard @quartata @TetsuyaYamamoto)
Conflicting feedback across revisions: current, #1
Conflicting feedback across revisions: current, #1
[ SmokeDetector | MS ] Bad NS for domain in body, bad keyword in body, bad keyword in title, blacklisted website in body, pattern-matching website in body, +1 more: KetoBoost Forskolin is works? by lelakilelo on astronomy.SE
[ SmokeDetector | MS ] Bad keyword in body, bad keyword in title, blacklisted website in body, link at end of body, pattern-matching website in body, +1 more: What is TestoSup Xtreme by testoxtreme on superuser.com (@TetsuyaYamamoto @iBug)
[ SmokeDetector | MS ] URL-only title, bad NS for domain in body, bad NS for domain in title, bad keyword in body, bad keyword in title, +4 more: x4up.org/primal-growth/ by xuqsjpfuj on askubuntu.com (@ThomasWard @quartata @TetsuyaYamamoto)
at some point in the past Metasmoke mysteriously acquired improved regex skills but nobody seemed to know how that happened or how to repeat it so when it was moved to a new host, we lost that
@micsthepick there is a separate rude/abusive bot which might be a better target for this sort of pattern maybe? look at Fire Alarm (though I guess it's Stack Overflow only, not network-wide at the mo)
but at the very least its developers might have more data than we do
There are two versions of FireAlarm. One version is written in C, and the other is written in Swift. This post is about the Swift version.
FireAlarm is a chatbot Ashish Ahuja and I have been working on. It attempts to detect low-quality questions using a naive Bayes filter.
How accura...
@Magisch that doesn't look like what we had, it seemed that it just magically enabled things like lookarounds and \d \b \s \w etc but of course I'm not too familiar with the surrounding code in Metasmoke, I can only report my experiences from using the search page
my vague recollection is that at some point around maybe a year ago I started using regex search more often and concluded that the regex it supported was pretty limited, but then a few months later noticed that somehow somebody had posted a search URL here which had stuff I thought wasn't supported, and discovered much not only to my own surprise apparently that magic had somehow been sprinkled
but then when Metasmoke had to be shut down and moved to a new host that magic sprinkle was lost
that might explain how the magic was sprinkled but the Metasmoke instructions say MySQL and the search results I get seem to confirm that that's what we have now
[ SmokeDetector | MS ] Bad keyword in body, blacklisted website in body, link at end of body, pattern-matching website in body: how skin make be perfect by user778933 on askubuntu.com (@ThomasWard @quartata @TetsuyaYamamoto)
@tripleee There's a bad catch for "phone number". It's a unix timestamp not far away from now. Shall we implement such a filter over phone number detection?
@iBug the blacklisted_websites.txt regexes are applied in a context where no anchors are in place so a substring match is possible ... didn't we discuss this already?
@iBug that could work, but it won't fix FPs on old posts
Regarding this answer, is it forbidden to suggest using a web-page as a solution, or is it spamming?
There are two versions of FireAlarm. One version is written in C, and the other is written in Swift. This post is about the Swift version.
FireAlarm is a chatbot Ashish Ahuja and I have been working on. It attempts to detect low-quality questions using a naive Bayes filter.
How accura...
