@ThomasWard The strange CI issues with the .count commit were some random CI issues. I re-ran the CI workflow and the rubocop issues are only the ones left from an earlier commit (i.e. they are in the master branch). The now-failures on the rails test is a typical random failure.
in theory it might... the flag condition posts query uses it to filter posts, but that's a filtered join not an indexed field so it probably doesn't make much difference
Come to think of it, that might have been a constant 1 for a reason... it means it's a lower bound for the accuracy rather than possibly overestimating and overflagging... again, unlikely, limited impact, but that might not be laziness
so... I'm gonna close this one, @user12986714, I don't think it's the right approach
FTR, the code you would've wanted ispost.reasons.count (or post.reasons.size, either works), CI was complaining about other things
@ArtOfCode The things in there that had caught my attention were the loops over conditions.each do |condition|. In both of those, the loop then calls condition.validate! for the same post over every existing/active condition.
The validate! function then calls post.reasons.pluck(:weight).reduce(:+) every single time that it's called (i.e. each iteration through the loop), but that's a value which will not change for the same post, and each loop only tests one post, so it's only one value for all iterations of the conditions.each do |condition| loop.
It seems like it would be much more efficient to compute that value once for the entire conditions.each do |condition| loop, and then use that value to test against every condition.
However, I don't really know how expensive post.reasons.pluck(:weight).reduce(:+) is, but I understand it to be a database access, which means it's not going to be as quick as just using a single pre-computed value, and it's called a few/several hundred times per post.
@ArtOfCode I'd probably rename validate! to validate_post! and then create something like validate_by_values!, which just takes straight values as arguments for weight, poster_rep, and reason_count. The validate_post! function could then just call validate_by_values! with the values from the post. The loops could compute the values for the weight, poster_rep, and reason_count once outside the loop and just call validate_by_values! directly in each iteration.
Hash maps in memory are volatile, get dropped when the app quits
Databases aren't
Could use redis but then it's another data store to keep in sync; can't just not use the database at all because it needs to be in there for the web app
Could cache it; again, you've got to keep it in sync and invalidate dead data
Database ops aren't that expensive when properly optimized, and often a hell of a lot faster than code for specific kinds of tasks
@user12986714 There are three reports for the post with different versions. You will need to go to MS to make sure the feedback on each report is what you intend.
@Makyen That pattern looks like it's already caught by Potentially bad keyword in answer and Potentially bad keyword in body; append -force if you really want to do that.
[ SmokeDetector | MS ] Link at end of answer, potentially bad ip for hostname in answer, potentially bad ns for domain in answer, potentially bad keyword in answer (63): Circuit breaker trip timing by liubo Zi on electronics.SE
@ArtOfCode @Undo @ThomasWard We've got autoflagging back! It's much faster! This answer [MS] was reported to MS 6 seconds after it was created. The first autoflag was raised 7 seconds after the report was passed to MS. Thank you, everyone, for the time and effort to get us to where MS is autoflagging rapidly.
@SmokeDetector Edited by a moderator to remove the spam link.
@SmokeDetector Also edited by a moderator to remove the link.
[ SmokeDetector | MS ] Bad keyword in body, bad keyword in title, pattern-matching product name in body, potentially bad ip for hostname in body, potentially bad ns for domain in body, +1 more (294): How To Use Vitrexotin Male Enhancement? by vitrexotin05 on workplace.SE
[ SmokeDetector | MS ] Bad ns for domain in body, bad keyword in body, bad keyword in title, blacklisted website in body, link at beginning of body, +6 more (731): Weight Loss Pills 2020 by MaryS Bechtel on askubuntu.com
MS is hosted by Thomas now and there's good chance it's on Ubuntu Bionic, which ships 8.10 in stock. However both Debian Buster and Ubuntu Focal ships Node 10.x
@Makyen I cannot use FIRE to mark as Spam, it gives me this message FIRE requires your metasmoke account to be write-authenticated with Stack Exchange in order to submit spam flags. Your metasmoke account doesn't appear to be write-authenticated. Please open this page to authenticate with Stack Exchange. I followed the link (metasmoke.erwaysoftware.com/authentication/status), but situation mantains the same. Any ideas?
