[ SmokeDetector | MS ] Bad keyword in body, bad keyword in title, blacklisted website in body, pattern-matching website in body (392): Stamimax review by careymoore on android.SE
@iBug Well, probably, it just means it's not fully managing its room presence, or not cleaning up after a room leaves rooms.yml. I could just kick it from the room.
anyone who has high reputation (so he can see deleted messages on SO) and can look at this deleted question stackoverflow.com/questions/51884623/… and tell me if it had a solution, or if it got moved/merged to another one?
@ThomasWard I can tell you my idea: Instead of using exit value (limited to 0-255), Smokey can create a file containing exit information. Then nocrash will read it and perform actions accordingly.
I have a blueprint in my mind, just need some free time to convert it into Python.
Windows seems to allow a return value ranging from 0 to 4294967295 (max of uint32)
@iBug either way it'd still pull data in, but the question is why add in an 'extra file' for exit tasks when we can just call them with exit codes as we always have done (because we can control that within loops with try/except, we can't control errors on execution of commands if we execute separate files and individual commands)
@iBug i don't have time during the day to handle this right now, too busy, but I'mma probably punch holes in your suggestions though with 'handling execution' problems.
if it's LITERALLY just a pip upgrade that's just two commands - not sure we really NEED an extra 'file' to handle the commands to execute, just a thought
@iBug again, point missed, we already do this with exit codes and logging in nocrash.py
yeah i guess, did read that one already, my problem is that i issue a logout and then a close for the bot client of discord.py and after that (asnyc btw) i call my routine which shuts down my programm and all i get is a pretty nasty exit
Task exception was never retrieved
future: <Task finished coro=<Client._run_event() done, defined at /home/michael/.local/lib/python3.6/site-packages/discord/client.py:304> exception=SystemExit(0,)>
Traceback (most recent call last):
@SmokeDetector This could be an actual question, but it looks much more like spam to me. In particular given the title and the link have the same text.
Hey, Teams Admins (@ArtOfCode and @Undo, I think), what's the Teams exported data look like?
> Not totally related, but also potentially relevant -- as a SO for Teams admin, you can download Teams data from the admin dashboard in JSON format at any time.
@Makyen That pattern looks like it's already caught by Pattern-matching website in body, Pattern-matching website in answer, Bad keyword in body, and Bad keyword in answer; append -force if you really want to do that.
OK, that didn't work. Smokey was reporting this URL as potentially being incorrectly flagged: stackoverflow.com/questions/5593775/… . It was, in fact, spam and I just destroyed the user behind it.
@SmokeDetector While the question is an off-topic request for external resource. The answer does give a link to something that purports to be what the OP asked for. Voted to close the question.
!!/watch-force- ^\s*(?:<p>\s*)?<a href="(?!(?:[a-z]+:)?//(?:[^/.]+\.)*(?:(?:apple|google|github|i\.stack\.imgur|stackexchange|stackoverflow|serverfault|superuser|askubuntu|stackapps)\.com|mathoverflow\.net|github\.io|edu|wikipedia\.org)[/"])\W*(?#Post begins with link; test for potential new detection; at watch 12650/12410/175/73)
@ThomasWard For this issue, on Circle CI, it appears to have a test case that it considers false which matches the regex I used. I'd argue that the test case is wrong, but I'm fine with changing the regex not to match.
!!/unwatch !!/watch-force- ^\s*(?:<p>\s*)?<a href="(?!(?:[a-z]+:)?//(?:[^/.]+\.)*(?:(?:apple|google|github|i\.stack\.imgur|stackexchange|stackoverflow|serverfault|superuser|askubuntu|stackapps)\.com|mathoverflow\.net|github\.io|edu|wikipedia\.org)[/"])\W*(?#Post begins with link; test for potential new detection; at watch 12650/12410/175/73)
@Makyen No such item !!/watch-force- ^\s*(?:<p>\s*)?<a href="(?!(?:[a-z]+:)?//(?:[^/.]+\.)*(?:(?:apple|google|github|i\.stack\.imgur|stackexchange|stackoverflow|serverfault|superuser|askubuntu|stackapps)\.com|mathoverflow\.net|github\.io|edu|wikipedia\.org)[/"])\W*(?#Post begins with link; test for potential new detection; at watch 12650/12410/175/73) in watchlist.
!!/unwatch ^\s*(?:<p>\s*)?<a href="(?!(?:[a-z]+:)?//(?:[^/.]+\.)*(?:(?:apple|google|github|i\.stack\.imgur|stackexchange|stackoverflow|serverfault|superuser|askubuntu|stackapps)\.com|mathoverflow\.net|github\.io|edu|wikipedia\.org)[/"])\W*(?#Post begins with link; test for potential new detection; at watch 12650/12410/175/73)
@ArtOfCode Thanks. It's actually easier, from my POV, to just change the regex, as I'd just exclude another domain, example.com. It's a reasonable domain to exclude, but it has no hits in MS data matching the regex, so I didn't already put it in the list of excludes.
!!/watch-force ^\s*(?:<p>\s*)?<a href="(?!(?:[a-z]+:)?//(?:[^/.]+\.)*(?:(?:example|apple|google|github|i\.stack\.imgur|stackexchange|stackoverflow|serverfault|superuser|askubuntu|stackapps)\.com|mathoverflow\.net|github\.io|edu|wikipedia\.org)[/"])\W*(?#Post begins with link; test for potential new detection; at watch 12650/12410/175/73)
!!/unwatch ^\s*(?:<p>\s*)?<a href="(?!(?:[a-z]+:)?//(?:[^/.]+\.)*(?:(?:example|apple|google|github|i\.stack\.imgur|stackexchange|stackoverflow|serverfault|superuser|askubuntu|stackapps)\.com|mathoverflow\.net|github\.io|edu|wikipedia\.org)[/"])\W*(?#Post begins with link; test for potential new detection; at watch 12650/12410/175/73)
!!/watch-force ^\s*<p>\s*?<a href="(?!(?:[a-z]+:)?//(?:[^/.]+\.)*(?:(?:example|apple|google|github|i\.stack\.imgur|stackexchange|stackoverflow|serverfault|superuser|askubuntu|stackapps)\.com|mathoverflow\.net|github\.io|edu|wikipedia\.org)[/"])\W*(?#Post begins with link; test for potential new detection; at watch 12649/12409/175/73)
Prior to watching (i.e. just based on the biased data in MS), that's at 98.1% TP. There's another like it, but which has a lower %TP rate (57.7% TP). The lower %TP one should either stay as a watch or be a separate detection. That assumes the %TP rates stay about what they are once some real-world data is available.
!!/watch-force ^\s*<p>\s*<a href="(?:(?:[a-z]+:)?//(?:[^/.]+\.)*(?:(?:apple|google|github|i\.stack\.imgur|stackexchange|stackoverflow|serverfault|superuser|askubuntu|stackapps)\.com|mathoverflow\.net|github\.io|edu|wikipedia\.org)[/"])\W*(?#Post begins with link: lower %TP domains; at watch 776/448/281/60)
While MS shows a decent TP rate for i.stack.imgur.com in that spot, the reality appears to be that it's way too common.
Let's try that without imgur.com.
!!/watch-force ^\s*<p>\s*<a href="(?:(?:[a-z]+:)?//(?:[^/.]+\.)*(?:(?:apple|google|github|stackexchange|stackoverflow|serverfault|superuser|askubuntu|stackapps)\.com|mathoverflow\.net|github\.io|edu|wikipedia\.org)[/"])\W*(?#Post begins with link: lower %TP domains; no imgur; at watch 210/146/44/24)
@Makyen Based on the data that was in MS at the time of the watch being added, i.stack.imgur.com was at 52.5% TP. Given that there were 5 FP right off the bat, the real rate is almost certainly way below that.
