Edit filter configuration

Differences between versions

ItemVersion from 21:41, 17 January 2020 by Suffusion of YellowVersion from 04:40, 17 February 2020 by Ohnoitsjamie
Basic information
Notes:
Regex unit tests at: https://regex101.com/r/YZTMl4/24
Regex unit tests at: https://regex101.com/r/YZTMl4/24


A fairly comprehensive filter for "yolo", "swag", "420 blaze it", "fuck her right in the pussy", "troll" and combinations thereof. "Yolo", "swag" (or "swagger"), and "troll" by themselves are not filtered. There was a filter similar to this a while back that apparently got deleted. I see this vandalism an awful lot, so thought I'd put this regex into a test phase -MusikAnimal
A fairly comprehensive filter for "yolo", "swag", "420 blaze it", "fuck her right in the pussy", "troll" and combinations thereof. "Yolo", "swag" (or "swagger"), and "troll" by themselves are not filtered. There was a filter similar to this a while back that apparently got deleted. I see this vandalism an awful lot, so thought I'd put this regex into a test phase -MusikAnimal


trigger warning and tag -MusikAnimal
trigger warning and tag -MusikAnimal


very few false positives, disallowing -MusikAnimal 2014.07.17
very few false positives, disallowing -MusikAnimal 2014.07.17


Add filter for "swag daddy" and "sweg" instead of "swag" to match new trends, go back to warn/tag temporarily -MusikAnimal 2014.08.27
Add filter for "swag daddy" and "sweg" instead of "swag" to match new trends, go back to warn/tag temporarily -MusikAnimal 2014.08.27


Back to disallow -MusikAnimal 2014.08.28
Back to disallow -MusikAnimal 2014.08.28


increase edit delta to 1000 ~MusikAnimal 2014.09.03
increase edit delta to 1000 ~MusikAnimal 2014.09.03


add 420 blaze it, tag only ~MusikAnimal 2014.09.10
add 420 blaze it, tag only ~MusikAnimal 2014.09.10
and back to disallow ~MusikAnimal 2014.09.11
and back to disallow ~MusikAnimal 2014.09.11


remove edit delta restriction ~MusikAnimal 2014.09.12
remove edit delta restriction ~MusikAnimal 2014.09.12


adding "fuck her right in the pussy", a new trend ~MusikAnimal 2014.09.23
adding "fuck her right in the pussy", a new trend ~MusikAnimal 2014.09.23


adding vandalistic forms of "troll" (trololol, trolllll, etc) ~MusikAnimal 2014.10.13
adding vandalistic forms of "troll" (trololol, trolllll, etc) ~MusikAnimal 2014.10.13


Fix some false positives ~MA 2014.10.29
Fix some false positives ~MA 2014.10.29


Optimize for conditions, replace [l\s*o\s*] with [ol\s] (though something else might be meant), optimize some of the regex, add !removed_lines irlike.
Optimize for conditions, replace [l\s*o\s*] with [ol\s] (though something else might be meant), optimize some of the regex, add !removed_lines irlike.


restore original regex, please use debuggex to test changes; filter runs very fast as-is ~MA 2015.07.08
restore original regex, please use debuggex to test changes; filter runs very fast as-is ~MA 2015.07.08


Moved regex into conditons. If this is compiled when the definition is hit, this will save processing time.  Otherwise no harm is done. RF 2015-07-14
Moved regex into conditons. If this is compiled when the definition is hit, this will save processing time.  Otherwise no harm is done. RF 2015-07-14


+chicken fucker from filter 689 ~MA
+chicken fucker from filter 689 ~MA
+thems the facts from filter 675 ~MA
+thems the facts from filter 675 ~MA


Note "y\s*o\s*[lo\s*]+s\s*w\s*[4ae]+\s*g|y+o+l+o[lo]|s+w+[4ae]+gg[g]|#yolo|#sw[4ae]g|sw[4ae]g\s*(?:yolo|daddy|money|lord|master)|420\s*b+l+a+z+e+\s*i+t|\s*i\s*n\s*t\s*h\s*[ae]\s*p\s*(?:(?:[@uv*]\s*)+(?:[zs$*]\s*)+|[zs$*]{2,})\s*a*y|t+r+o+l(?:o+l+|ll+)|lolo(?:lo)|(?:d[3e](?:[3e]+[sz]+|[sz][sz]*)e?|th[3e][zs$][3e])\s*nu+t[zs$]|(?:f[u\*][c\*]k(?:ing?|ed|s)|sex\s*with?)\s*chickens?|chicken\s*f[u\*]?[c\*]k(?:er|s|ing)?|(?:them\'?s?|dems?|those\s+are)'?\s+(?:th[ea]|da)\s+fa(?:cts?|x)!?"
Note "y\s*o\s*[lo\s*]+s\s*w\s*[4ae]+\s*g|y+o+l+o[lo]|s+w+[4ae]+gg[g]|#yolo|#sw[4ae]g|sw[4ae]g\s*(?:yolo|daddy|money|lord|master)|420\s*b+l+a+z+e+\s*i+t|\s*i\s*n\s*t\s*h\s*[ae]\s*p\s*(?:(?:[@uv*]\s*)+(?:[zs$*]\s*)+|[zs$*]{2,})\s*a*y|t+r+o+l(?:o+l+|ll+)|lolo(?:lo)|(?:d[3e](?:[3e]+[sz]+|[sz][sz]*)e?|th[3e][zs$][3e])\s*nu+t[zs$]|(?:f[u\*][c\*]k(?:ing?|ed|s)|sex\s*with?)\s*chickens?|chicken\s*f[u\*]?[c\*]k(?:er|s|ing)?|(?:them\'?s?|dems?|those\s+are)'?\s+(?:th[ea]|da)\s+fa(?:cts?|x)!?"


is shorter and matches faster:- checks out on debuggex.  RF
is shorter and matches faster:- checks out on debuggex.  RF


FP for "Deez Nuts (band)" - lets assume that if there's no link to the article it's not legit, at least for now. RF
FP for "Deez Nuts (band)" - lets assume that if there's no link to the article it's not legit, at least for now. RF


+"dank meme" per [[Special:Permalink/761029363#Dank Memes]], removing check for "Deez Nuts (band)" for now since it is semi'd ~MA 2016.01.20
+"dank meme" per [[Special:Permalink/761029363#Dank Memes]], removing check for "Deez Nuts (band)" for now since it is semi'd ~MA 2016.01.20


Exempt everything that contains yolo or swag already. - KoH 20170122
Exempt everything that contains yolo or swag already. - KoH 20170122


Adding `+ "|yolo|swag"` apparently broke the filter, but checking against abuseStr by itself should handle the vast majority of false positives ~MA 2017.01.26
Adding `+ "|yolo|swag"` apparently broke the filter, but checking against abuseStr by itself should handle the vast majority of false positives ~MA 2017.01.26


+"subscribe to pewdiepie" and similar -Galobtter 2019.01.01
+"subscribe to pewdiepie" and similar -Galobtter 2019.01.01


+transmania(n), as filter 260 where this was listed has exceptions that this vandalism sneaks in on. -C 6Jan19
+transmania(n), as filter 260 where this was listed has exceptions that this vandalism sneaks in on. -C 6Jan19


\b, false positives with "transmaniacon" (Special:AbuseLog/22928286) -G 2019.01.07
\b, false positives with "transmaniacon" (Special:AbuseLog/22928286) -G 2019.01.07
+"big chungus" -G 2019.01.08
+"big chungus" -G 2019.01.08


+"spicy boi" and "fat biscuit" from filter 787 -G 2019.01.21
+"spicy boi" and "fat biscuit" from filter 787 -G 2019.01.21
Actually remove "fat biscuit" as I don't see any hits in the log of 787 for it -G 2019.01.21
Actually remove "fat biscuit" as I don't see any hits in the log of 787 for it -G 2019.01.21


+"gucci\s*gang[\s,]*gucci\s*gang|\breeeeee|fetus\s*deletus", tested on 953. -G 2019.01.23
+"gucci\s*gang[\s,]*gucci\s*gang|\breeeeee|fetus\s*deletus", tested on 953. -G 2019.01.23
+yeet, tested on 953. -G 2019.01.23
+yeet, tested on 953. -G 2019.01.23
+"\beats?\s*ass\b|ugandan\s*knuckles" tested on 953 -G 2019.01.25
+"\beats?\s*ass\b|ugandan\s*knuckles" tested on 953 -G 2019.01.25
Last batch from 953: "hard\s+(?:pp|peepee)|(?:pp|peepee)\s+hard|\booo+f+\b|\bnibb+a+\b|absolute\s*unit|you'?ve\s*been\s*gnomed". -G 2019.01.31
Last batch from 953: "hard\s+(?:pp|peepee)|(?:pp|peepee)\s+hard|\booo+f+\b|\bnibb+a+\b|absolute\s*unit|you'?ve\s*been\s*gnomed". -G 2019.01.31
+thicc and variants, -spicy boy (stale meme). More variants on subscribe to pewdiepie. Check edit summaries - catches a bit. -G 2019.02.14
+thicc and variants, -spicy boy (stale meme). More variants on subscribe to pewdiepie. Check edit summaries - catches a bit. -G 2019.02.14
Catch variants on transmania(n) per Special:Contributions/ParasolProds and Special:Contributions/OnScreenThat1982 -G 2019.02.16
Catch variants on transmania(n) per Special:Contributions/ParasolProds and Special:Contributions/OnScreenThat1982 -G 2019.02.16
Remove checking of summaries - I did not account for reverts inadvertently introducing the problematic phrases. -G 2019.02.17
Remove checking of summaries - I did not account for reverts inadvertently introducing the problematic phrases. -G 2019.02.17
+"g+aaa+y|\bg+a+yy|\bayyy|\bl+m+a+o|rawr\s*xd|b+o+iii|e+s+k+e+t+i+t|\bh+iii|h+iiii|hehehe|h[au]r[au]mba?e\b|aviation\s*,[\s\S]*?there\s*is\s*no[\s\S]*?bee[\s\S]*?be\s*able\s*to\s*fly"; some terms tested on 953, others (harambe, bee movie script) are merged from 784 and 813. Also check page_title now. -G 2019.02.28
+"g+aaa+y|\bg+a+yy|\bayyy|\bl+m+a+o|rawr\s*xd|b+o+iii|e+s+k+e+t+i+t|\bh+iii|h+iiii|hehehe|h[au]r[au]mba?e\b|aviation\s*,[\s\S]*?there\s*is\s*no[\s\S]*?bee[\s\S]*?be\s*able\s*to\s*fly"; some terms tested on 953, others (harambe, bee movie script) are merged from 784 and 813. Also check page_title now. -G 2019.02.28


Add word boundary to \bg+a+yy, causing FPs on names like "Gayyoom". Would have caused no FNs in last 1000 hits. --Suffusion of Yellow 18:01 16 May 2019
Add word boundary to \bg+a+yy, causing FPs on names like "Gayyoom". Would have caused no FNs in last 1000 hits. --Suffusion of Yellow 18:01 16 May 2019


+"dQw4w9WgXcQ|\buwu\b|\bowo\b|hit\s*or\s*miss[\s,]*I\s*guess", tested on 953. -G 2019-05-17
+"dQw4w9WgXcQ|\buwu\b|\bowo\b|hit\s*or\s*miss[\s,]*I\s*guess", tested on 953. -G 2019-05-17


+"ok(?:ay)? boomer" -G 2019-11-21
+"ok(?:ay)? boomer" -G 2019-11-21


"joe m[oa]m+a", tested on 1013 --Suffusion of Yellow 17:47 26 Nov 2019
"joe m[oa]m+a", tested on 1013 --Suffusion of Yellow 17:47 26 Nov 2019


Merge "#redirect\s*\[\[donald[\s_]trump\]\]" from filter 843 --Suffusion of Yellow 20:56 27 Nov 2019
Merge "#redirect\s*\[\[donald[\s_]trump\]\]" from filter 843 --Suffusion of Yellow 20:56 27 Nov 2019


+"epstein\W+did\s*n.?t\s+kill" (See [[Epstein didn't kill himself]]), tested on 1013 --Suffusion of Yellow 20:46 30 Nov 2019
+"epstein\W+did\s*n.?t\s+kill" (See [[Epstein didn't kill himself]]), tested on 1013 --Suffusion of Yellow 20:46 30 Nov 2019


"g+aaa+y" matched Google Books URLs twice today. Temporarily moving to filter 1014 in case Google changed something in the format. --Suffusion of Yellow 07:13 6 Dec 2019
"g+aaa+y" matched Google Books URLs twice today. Temporarily moving to filter 1014 in case Google changed something in the format. --Suffusion of Yellow 07:13 6 Dec 2019
Leaving out "g+aaa+y" for good. Lots of gbooks urls contain "[A-Z]AAA[A-Z]" and only successfully used for vandalism twice in one week. --Suffusion of Yellow 22:27 14 Dec 2019
Leaving out "g+aaa+y" for good. Lots of gbooks urls contain "[A-Z]AAA[A-Z]" and only successfully used for vandalism twice in one week. --Suffusion of Yellow 22:27 14 Dec 2019


Add misspelling "epstien". --Suffusion of Yellow 21:39 17 Jan 2020
Add misspelling "epstien". --Suffusion of Yellow 21:39 17 Jan 2020
Add "bruh" -- Ohnoitsjamie 16:59 Feb 16 2020
Filter conditions
Conditions:
(documentation)
!("confirmed" in user_groups) &
!("confirmed" in user_groups) &
page_namespace == 0 & (
page_namespace == 0 & (
   abuseStr := "y\s*o\s*[lo\s]+s\s*w\s*[4ae]+\s*g+|\by+o+l+o[lo]+|s+w+[4ae]+gg[g]+|#yolo|#sw[4ae]g|sw[4ae]g\s*(?:yolo|daddy|money|lord|master)|420\s*b+l+a+z+e+\s*i+t+|\s*i\s*n\s*t\s*h\s*[ae]\s*p\s*(?:(?:[@uv*]\s*)+(?:[zs$*]\s*)+|[zs$*]{2,})\s*a*y+|\bt+r+o+l(?:o+l|ll)|lolo(?:lo)+|(?:d[3e](?:[3e]+[sz]+|[sz][sz]*)e*|th[3e][zs$][3e])\s*n+u+t+[zs$]|(?:f[u\*][c\*]k(?:ing?|ed|s)|sex\s*with?)\s*chickens?|chicken\s*f[u\*]?[c\*]k(?:er|s|ing)?|(?:them\'?s?|dems?|those\s+are)'?\s+(?:th[ea]|da)\s+fa(?:cts?|x)!?|hitler\s*did\s*nothing?\s*wrong|bush\s*did\s*9.?11|dank\s*meme|drumpf|\bdat\s* boi|sub(?:scrib(?:e|es|ed|ing))?\s*(?:to|2)\s*(?:p(ew|ud|ue|uw|oo)|te*.?series)|tran?s?.?manian?\b|chung[uea]s\b|gucci\s*gang[\s,]*gucci\s*gang|\br+eeeeee|fetus\s*deletus|(?<!koe\s)\by+ee+t+(?:e+(?:r+|d+))?\b|\beats?\s*ass\b|ugandan\s*knuckles|hard\s+(?:pp|peepee)|(?:pp|peepee)\s+hard|\booo+f+\b|\bnibb+a+\b|absolute\s*unit|you'?ve\s*been\s*gnomed|\bt+\s*h+\s*i+\s*c\s*c|\bg+a+yy(?:y|\b)|\bayyy|\bl+m+a+o|rawr\s*xd|b+o+iii|e+s+k+e+t+i+t|\bh+iii+\b|h+iiiii|hehehe|h[au]r[au]mba?e\b|aviation\s*,[\s\S]*?there\s*is\s*no[\s\S]*?bee[\s\S]*?be\s*able\s*to\s*fly|dQw4w9WgXcQ|\buwu\b|\bowo\b|hit\s*or\s*miss[\s,]*I\s*guess|ok(?:ay)?,? boomer|joe m[oa]m+a|#redirect\s*\[\[donald[\s_]trump\]\]|epst(?:ei|ie)n\W+did\s*n.?t\s+kill";
   abuseStr := "y\s*o\s*[lo\s]+s\s*w\s*[4ae]+\s*g+|\by+o+l+o[lo]+|s+w+[4ae]+gg[g]+|#yolo|#sw[4ae]g|sw[4ae]g\s*(?:yolo|daddy|money|lord|master)|420\s*b+l+a+z+e+\s*i+t+|\s*i\s*n\s*t\s*h\s*[ae]\s*p\s*(?:(?:[@uv*]\s*)+(?:[zs$*]\s*)+|[zs$*]{2,})\s*a*y+|\bt+r+o+l(?:o+l|ll)|lolo(?:lo)+|(?:d[3e](?:[3e]+[sz]+|[sz][sz]*)e*|th[3e][zs$][3e])\s*n+u+t+[zs$]|(?:f[u\*][c\*]k(?:ing?|ed|s)|sex\s*with?)\s*chickens?|chicken\s*f[u\*]?[c\*]k(?:er|s|ing)?|(?:them\'?s?|dems?|those\s+are)'?\s+(?:th[ea]|da)\s+fa(?:cts?|x)!?|hitler\s*did\s*nothing?\s*wrong|bush\s*did\s*9.?11|dank\s*meme|drumpf|\bdat\s* boi|sub(?:scrib(?:e|es|ed|ing))?\s*(?:to|2)\s*(?:p(ew|ud|ue|uw|oo)|te*.?series)|tran?s?.?manian?\b|chung[uea]s\b|gucci\s*gang[\s,]*gucci\s*gang|\br+eeeeee|fetus\s*deletus|(?<!koe\s)\by+ee+t+(?:e+(?:r+|d+))?\b|\beats?\s*ass\b|ugandan\s*knuckles|hard\s+(?:pp|peepee)|(?:pp|peepee)\s+hard|\booo+f+\b|\bnibb+a+\b|absolute\s*unit|you'?ve\s*been\s*gnomed|\bt+\s*h+\s*i+\s*c\s*c|\bg+a+yy(?:y|\b)|\bayyy|\bl+m+a+o|rawr\s*xd|b+o+iii|e+s+k+e+t+i+t|\bh+iii+\b|h+iiiii|hehehe|h[au]r[au]mba?e\b|aviation\s*,[\s\S]*?there\s*is\s*no[\s\S]*?bee[\s\S]*?be\s*able\s*to\s*fly|dQw4w9WgXcQ|\buwu\b|\bowo\b|hit\s*or\s*miss[\s,]*I\s*guess|ok(?:ay)?,? boomer|joe m[oa]m+a|#redirect\s*\[\[donald[\s_]trump\]\]|epst(?:ei|ie)n\W+did\s*n.?t\s+kill|bruh";


   added_lines irlike abuseStr &
   added_lines irlike abuseStr &
   !( (removed_lines + page_title) irlike abuseStr)
   !( (removed_lines + page_title) irlike abuseStr)
)
)