package userutil import ( "regexp" "strings" "sync" "unicode" ) // slurPatterns contains regexes that match slurs and their common obfuscations, // including unicode lookalikes. Ported from tranquil-pds moderation/mod.rs. // // CONTENT WARNING: this file exists solely to block hateful language from // appearing in subdomains. The patterns below do not reflect the views of // this project or its contributors. var ( slurPatterns []*regexp.Regexp slurPatternsOnce sync.Once ) func getSlurPatterns() []*regexp.Regexp { slurPatternsOnce.Do(func() { raw := []string{ `(?i)\b[cĆćĈĉČčĊċÇçḈḉȻȼꞒꞓꟄꞔƇƈɕ][hĤĥȞȟḦḧḢḣḨḩḤḥḪḫH̱ẖĦħⱧⱨꞪɦꞕΗНн][iÍíi̇́Ììi̇̀ĬĭÎîǏǐÏïḮḯĨĩi̇̃ĮįĮ́į̇́Į̃į̇̃ĪīĪ̀ī̀ỈỉȈȉI̋i̋ȊȋỊịꞼꞽḬḭƗɨᶖİiIıIi1lĺľļḷḹl̃ḽḻłŀƚꝉⱡɫɬꞎꬷꬸꬹᶅɭȴLl][nŃńǸǹŇňÑñṄṅŅņṆṇṊṋṈṉN̈n̈ƝɲŊŋꞐꞑꞤꞥᵰᶇɳȵꬻꬼИиПпNn][kḰḱǨǩĶķḲḳḴḵƘƙⱩⱪᶄꝀꝁꝂꝃꝄꝅꞢꞣ][sŚśṤṥŜŝŠšṦṧṠṡŞşṢṣṨṩȘșS̩s̩ꞨꞩⱾȿꟅʂᶊᵴ]?\b`, `(?i)\b[cĆćĈĉČčĊċÇçḈḉȻȼꞒꞓꟄꞔƇƈɕ][ÓóÒòŎŏÔôỐốỒồỖỗỔổǑǒÖöȪȫŐőÕõṌṍṎṏȬȭȮȯO͘o͘ȰȱØøǾǿǪǫǬǭŌōṒṓṐṑỎỏȌȍȎȏƠơỚớỜờỠỡỞởỢợỌọỘộO̩o̩Ò̩ò̩Ó̩ó̩ƟɵꝊꝋꝌꝍⱺOo0]{2}[nŃńǸǹŇňÑñṄṅŅņṆṇṊṋṈṉN̈n̈ƝɲŊŋꞐꞑꞤꞥᵰᶇɳȵꬻꬼИиПпNn][sŚśṤṥŜŝŠšṦṧṠṡŞşṢṣṨṩȘșS̩s̩ꞨꞩⱾȿꟅʂᶊᵴ]?\b`, `(?i)\b[fḞḟƑƒꞘꞙᵮᶂ][aÁáÀàĂăẮắẰằẴẵẲẳÂâẤấẦầẪẫẨẩǍǎÅåǺǻÄäǞǟÃãȦȧǠǡĄąĄ́ą́Ą̃ą̃ĀāĀ̀ā̀ẢảȀȁA̋a̋ȂȃẠạẶặẬậḀḁȺⱥꞺꞻᶏẚAa@4][gǴǵĞğĜĝǦǧĠġG̃g̃ĢģḠḡǤǥꞠꞡƓɠᶃꬶGg]{1,2}([ÓóÒòŎŏÔôỐốỒồỖỗỔổǑǒÖöȪȫŐőÕõṌṍṎṏȬȭȮȯO͘o͘ȰȱØøǾǿǪǫǬǭŌōṒṓṐṑỎỏȌȍȎȏƠơỚớỜờỠỡỞởỢợỌọỘộO̩o̩Ò̩ò̩Ó̩ó̩ƟɵꝊꝋꝌꝍⱺOo0e3ЄєЕеÉéÈèĔĕÊêẾếỀềỄễỂểÊ̄ê̄Ê̌ê̌ĚěËëẼẽĖėĖ́ė́Ė̃ė̃ȨȩḜḝĘęĘ́ę́Ę̃ę̃ĒēḖḗḔḕẺẻȄȅE̋e̋ȆȇẸẹỆệḘḙḚḛɆɇE̩e̩È̩è̩É̩é̩ᶒⱸꬴꬳEeiÍíi̇́Ììi̇̀ĬĭÎîǏǐÏïḮḯĨĩi̇̃ĮįĮ́į̇́Į̃į̇̃ĪīĪ̀ī̀ỈỉȈȉI̋i̋ȊȋỊịꞼꞽḬḭƗɨᶖİiIıIi1lĺľļḷḹl̃ḽḻłŀƚꝉⱡɫɬꞎꬷꬸꬹᶅɭȴLl][tŤťṪṫŢţṬṭȚțṰṱṮṯŦŧȾⱦƬƭƮʈT̈ẗᵵƫȶ]{1,2}([rŔŕŘřṘṙŖŗȐȑȒȓṚṛṜṝṞṟR̃r̃ɌɍꞦꞧⱤɽᵲᶉꭉ][yÝýỲỳŶŷY̊ẙŸÿỸỹẎẏȲȳỶỷỴỵɎɏƳƴỾỿ]|[rŔŕŘřṘṙŖŗȐȑȒȓṚṛṜṝṞṟR̃r̃ɌɍꞦꞧⱤɽᵲᶉꭉ][iÍíi̇́Ììi̇̀ĬĭÎîǏǐÏïḮḯĨĩi̇̃ĮįĮ́į̇́Į̃į̇̃ĪīĪ̀ī̀ỈỉȈȉI̋i̋ȊȋỊịꞼꞽḬḭƗɨᶖİiIıIi1lĺľļḷḹl̃ḽḻłŀƚꝉⱡɫɬꞎꬷꬸꬹᶅɭȴLl][e3ЄєЕеÉéÈèĔĕÊêẾếỀềỄễỂểÊ̄ê̄Ê̌ê̌ĚěËëẼẽĖėĖ́ė́Ė̃ė̃ȨȩḜḝĘęĘ́ę́Ę̃ę̃ĒēḖḗḔḕẺẻȄȅE̋e̋ȆȇẸẹỆệḘḙḚḛɆɇE̩e̩È̩è̩É̩é̩ᶒⱸꬴꬳEe])?)?[sŚśṤṥŜŝŠšṦṧṠṡŞşṢṣṨṩȘșS̩s̩ꞨꞩⱾȿꟅʂᶊᵴ]?\b`, `(?i)\b[kḰḱǨǩĶķḲḳḴḵƘƙⱩⱪᶄꝀꝁꝂꝃꝄꝅꞢꞣ][iÍíi̇́Ììi̇̀ĬĭÎîǏǐÏïḮḯĨĩi̇̃ĮįĮ́į̇́Į̃į̇̃ĪīĪ̀ī̀ỈỉȈȉI̋i̋ȊȋỊịꞼꞽḬḭƗɨᶖİiIıIi1lĺľļḷḹl̃ḽḻłŀƚꝉⱡɫɬꞎꬷꬸꬹᶅɭȴLlyÝýỲỳŶŷY̊ẙŸÿỸỹẎẏȲȳỶỷỴỵɎɏƳƴỾỿ][kḰḱǨǩĶķḲḳḴḵƘƙⱩⱪᶄꝀꝁꝂꝃꝄꝅꞢꞣ][e3ЄєЕеÉéÈèĔĕÊêẾếỀềỄễỂểÊ̄ê̄Ê̌ê̌ĚěËëẼẽĖėĖ́ė́Ė̃ė̃ȨȩḜḝĘęĘ́ę́Ę̃ę̃ĒēḖḗḔḕẺẻȄȅE̋e̋ȆȇẸẹỆệḘḙḚḛɆɇE̩e̩È̩è̩É̩é̩ᶒⱸꬴꬳEe]([rŔŕŘřṘṙŖŗȐȑȒȓṚṛṜṝṞṟR̃r̃ɌɍꞦꞧⱤɽᵲᶉꭉ][yÝýỲỳŶŷY̊ẙŸÿỸỹẎẏȲȳỶỷỴỵɎɏƳƴỾỿ]|[rŔŕŘřṘṙŖŗȐȑȒȓṚṛṜṝṞṟR̃r̃ɌɍꞦꞧⱤɽᵲᶉꭉ][iÍíi̇́Ììi̇̀ĬĭÎîǏǐÏïḮḯĨĩi̇̃ĮįĮ́į̇́Į̃į̇̃ĪīĪ̀ī̀ỈỉȈȉI̋i̋ȊȋỊịꞼꞽḬḭƗɨᶖİiIıIi1lĺľļḷḹl̃ḽḻłŀƚꝉⱡɫɬꞎꬷꬸꬹᶅɭȴLl][e3ЄєЕеÉéÈèĔĕÊêẾếỀềỄễỂểÊ̄ê̄Ê̌ê̌ĚěËëẼẽĖėĖ́ė́Ė̃ė̃ȨȩḜḝĘęĘ́ę́Ę̃ę̃ĒēḖḗḔḕẺẻȄȅE̋e̋ȆȇẸẹỆệḘḙḚḛɆɇE̩e̩È̩è̩É̩é̩ᶒⱸꬴꬳEe])?[sŚśṤṥŜŝŠšṦṧṠṡŞşṢṣṨṩȘșS̩s̩ꞨꞩⱾȿꟅʂᶊᵴ]*\b`, `(?i)\b[nŃńǸǹŇňÑñṄṅŅņṆṇṊṋṈṉN̈n̈ƝɲŊŋꞐꞑꞤꞥᵰᶇɳȵꬻꬼИиПпNn][iÍíi̇́Ììi̇̀ĬĭÎîǏǐÏïḮḯĨĩi̇̃ĮįĮ́į̇́Į̃į̇̃ĪīĪ̀ī̀ỈỉȈȉI̋i̋ȊȋỊịꞼꞽḬḭƗɨᶖİiIıIi1lĺľļḷḹl̃ḽḻłŀƚꝉⱡɫɬꞎꬷꬸꬹᶅɭȴLloÓóÒòŎŏÔôỐốỒồỖỗỔổǑǒÖöȪȫŐőÕõṌṍṎṏȬȭȮȯO͘o͘ȰȱØøǾǿǪǫǬǭŌōṒṓṐṑỎỏȌȍȎȏƠơỚớỜờỠỡỞởỢợỌọỘộO̩o̩Ò̩ò̩Ó̩ó̩ƟɵꝊꝋꝌꝍⱺOoІіa4ÁáÀàĂăẮắẰằẴẵẲẳÂâẤấẦầẪẫẨẩǍǎÅåǺǻÄäǞǟÃãȦȧǠǡĄąĄ́ą́Ą̃ą̃ĀāĀ̀ā̀ẢảȀȁA̋a̋ȂȃẠạẶặẬậḀḁȺⱥꞺꞻᶏẚAa][gǴǵĞğĜĝǦǧĠġG̃g̃ĢģḠḡǤǥꞠꞡƓɠᶃꬶGgqꝖꝗꝘꝙɋʠ]{2}(l[e3ЄєЕеÉéÈèĔĕÊêẾếỀềỄễỂểÊ̄ê̄Ê̌ê̌ĚěËëẼẽĖėĖ́ė́Ė̃ė̃ȨȩḜḝĘęĘ́ę́Ę̃ę̃ĒēḖḗḔḕẺẻȄȅE̋e̋ȆȇẸẹỆệḘḙḚḛɆɇE̩e̩È̩è̩É̩é̩ᶒⱸꬴꬳEe]t|[e3ЄєЕеÉéÈèĔĕÊêẾếỀềỄễỂểÊ̄ê̄Ê̌ê̌ĚěËëẼẽĖėĖ́ė́Ė̃ė̃ȨȩḜḝĘęĘ́ę́Ę̃ę̃ĒēḖḗḔḕẺẻȄȅE̋e̋ȆȇẸẹỆệḘḙḚḛɆɇE̩e̩È̩è̩É̩é̩ᶒⱸꬴꬳEeaÁáÀàĂăẮắẰằẴẵẲẳÂâẤấẦầẪẫẨẩǍǎÅåǺǻÄäǞǟÃãȦȧǠǡĄąĄ́ą́Ą̃ą̃ĀāĀ̀ā̀ẢảȀȁA̋a̋ȂȃẠạẶặẬậḀḁȺⱥꞺꞻᶏẚAa][rŔŕŘřṘṙŖŗȐȑȒȓṚṛṜṝṞṟR̃r̃ɌɍꞦꞧⱤɽᵲᶉꭉ]?|n[ÓóÒòŎŏÔôỐốỒồỖỗỔổǑǒÖöȪȫŐőÕõṌṍṎṏȬȭȮȯO͘o͘ȰȱØøǾǿǪǫǬǭŌōṒṓṐṑỎỏȌȍȎȏƠơỚớỜờỠỡỞởỢợỌọỘộO̩o̩Ò̩ò̩Ó̩ó̩ƟɵꝊꝋꝌꝍⱺOo0][gǴǵĞğĜĝǦǧĠġG̃g̃ĢģḠḡǤǥꞠꞡƓɠᶃꬶGgqꝖꝗꝘꝙɋʠ]|[a4ÁáÀàĂăẮắẰằẴẵẲẳÂâẤấẦầẪẫẨẩǍǎÅåǺǻÄäǞǟÃãȦȧǠǡĄąĄ́ą́Ą̃ą̃ĀāĀ̀ā̀ẢảȀȁA̋a̋ȂȃẠạẶặẬậḀḁȺⱥꞺꞻᶏẚAa]?)?[sŚśṤṥŜŝŠšṦṧṠṡŞşṢṣṨṩȘșS̩s̩ꞨꞩⱾȿꟅʂᶊᵴ]?\b`, `(?i)[nŃńǸǹŇňÑñṄṅŅņṆṇṊṋṈṉN̈n̈ƝɲŊŋꞐꞑꞤꞥᵰᶇɳȵꬻꬼИиПпNn][iÍíi̇́Ììi̇̀ĬĭÎîǏǐÏïḮḯĨĩi̇̃ĮįĮ́į̇́Į̃į̇̃ĪīĪ̀ī̀ỈỉȈȉI̋i̋ȊȋỊịꞼꞽḬḭƗɨᶖİiIıIi1lĺľļḷḹl̃ḽḻłŀƚꝉⱡɫɬꞎꬷꬸꬹᶅɭȴLloÓóÒòŎŏÔôỐốỒồỖỗỔổǑǒÖöȪȫŐőÕõṌṍṎṏȬȭȮȯO͘o͘ȰȱØøǾǿǪǫǬǭŌōṒṓṐṑỎỏȌȍȎȏƠơỚớỜờỠỡỞởỢợỌọỘộO̩o̩Ò̩ò̩Ó̩ó̩ƟɵꝊꝋꝌꝍⱺOoІіa4ÁáÀàĂăẮắẰằẴẵẲẳÂâẤấẦầẪẫẨẩǍǎÅåǺǻÄäǞǟÃãȦȧǠǡĄąĄ́ą́Ą̃ą̃ĀāĀ̀ā̀ẢảȀȁA̋a̋ȂȃẠạẶặẬậḀḁȺⱥꞺꞻᶏẚAa][gǴǵĞğĜĝǦǧĠġG̃g̃ĢģḠḡǤǥꞠꞡƓɠᶃꬶGgqꝖꝗꝘꝙɋʠ]{2}(l[e3ЄєЕеÉéÈèĔĕÊêẾếỀềỄễỂểÊ̄ê̄Ê̌ê̌ĚěËëẼẽĖėĖ́ė́Ė̃ė̃ȨȩḜḝĘęĘ́ę́Ę̃ę̃ĒēḖḗḔḕẺẻȄȅE̋e̋ȆȇẸẹỆệḘḙḚḛɆɇE̩e̩È̩è̩É̩é̩ᶒⱸꬴꬳEe]t|[e3ЄєЕеÉéÈèĔĕÊêẾếỀềỄễỂểÊ̄ê̄Ê̌ê̌ĚěËëẼẽĖėĖ́ė́Ė̃ė̃ȨȩḜḝĘęĘ́ę́Ę̃ę̃ĒēḖḗḔḕẺẻȄȅE̋e̋ȆȇẸẹỆệḘḙḚḛɆɇE̩e̩È̩è̩É̩é̩ᶒⱸꬴꬳEe][rŔŕŘřṘṙŖŗȐȑȒȓṚṛṜṝṞṟR̃r̃ɌɍꞦꞧⱤɽᵲᶉꭉ])[sŚśṤṥŜŝŠšṦṧṠṡŞşṢṣṨṩȘșS̩s̩ꞨꞩⱾȿꟅʂᶊᵴ]?`, `(?i)\b[tŤťṪṫŢţṬṭȚțṰṱṮṯŦŧȾⱦƬƭƮʈT̈ẗᵵƫȶ][rŔŕŘřṘṙŖŗȐȑȒȓṚṛṜṝṞṟR̃r̃ɌɍꞦꞧⱤɽᵲᶉꭉ][aÁáÀàĂăẮắẰằẴẵẲẳÂâẤấẦầẪẫẨẩǍǎÅåǺǻÄäǞǟÃãȦȧǠǡĄąĄ́ą́Ą̃ą̃ĀāĀ̀ā̀ẢảȀȁA̋a̋ȂȃẠạẶặẬậḀḁȺⱥꞺꞻᶏẚAa4]+[nŃńǸǹŇňÑñṄṅŅņṆṇṊṋṈṉN̈n̈ƝɲŊŋꞐꞑꞤꞥᵰᶇɳȵꬻꬼИиПпNn]{1,2}([iÍíi̇́Ììi̇̀ĬĭÎîǏǐÏïḮḯĨĩi̇̃ĮįĮ́į̇́Į̃į̇̃ĪīĪ̀ī̀ỈỉȈȉI̋i̋ȊȋỊịꞼꞽḬḭƗɨᶖİiIıIi1lĺľļḷḹl̃ḽḻłŀƚꝉⱡɫɬꞎꬷꬸꬹᶅɭȴLl][e3ЄєЕеÉéÈèĔĕÊêẾếỀềỄễỂểÊ̄ê̄Ê̌ê̌ĚěËëẼẽĖėĖ́ė́Ė̃ė̃ȨȩḜḝĘęĘ́ę́Ę̃ę̃ĒēḖḗḔḕẺẻȄȅE̋e̋ȆȇẸẹỆệḘḙḚḛɆɇE̩e̩È̩è̩É̩é̩ᶒⱸꬴꬳEe]|[yÝýỲỳŶŷY̊ẙŸÿỸỹẎẏȲȳỶỷỴỵɎɏƳƴỾỿ]|[e3ЄєЕеÉéÈèĔĕÊêẾếỀềỄễỂểÊ̄ê̄Ê̌ê̌ĚěËëẼẽĖėĖ́ė́Ė̃ė̃ȨȩḜḝĘęĘ́ę́Ę̃ę̃ĒēḖḗḔḕẺẻȄȅE̋e̋ȆȇẸẹỆệḘḙḚḛɆɇE̩e̩È̩è̩É̩é̩ᶒⱸꬴꬳEe][rŔŕŘřṘṙŖŗȐȑȒȓṚṛṜṝṞṟR̃r̃ɌɍꞦꞧⱤɽᵲᶉꭉ])[sŚśṤṥŜŝŠšṦṧṠṡŞşṢṣṨṩȘșS̩s̩ꞨꞩⱾȿꟅʂᶊᵴ]?\b`, } for _, r := range raw { slurPatterns = append(slurPatterns, regexp.MustCompile(r)) } }) return slurPatterns } // stripTrailingDigits removes trailing ASCII digits from s. func stripTrailingDigits(s string) string { return strings.TrimRightFunc(s, func(r rune) bool { return r >= '0' && r <= '9' }) } // normalizeLeetspeak replaces common leetspeak substitutions with their // standard ASCII equivalents. func normalizeLeetspeak(s string) string { var b strings.Builder b.Grow(len(s)) for _, c := range s { switch c { case '4', '@': b.WriteRune('a') case '3': b.WriteRune('e') case '1', '!', '|': b.WriteRune('i') case '0': b.WriteRune('o') case '5', '$': b.WriteRune('s') case '7': b.WriteRune('t') case '8': b.WriteRune('b') case '9': b.WriteRune('g') default: b.WriteRune(c) } } return b.String() } // foldToASCII strips unicode combining characters and maps each rune to its // lowercase ASCII equivalent where possible, leaving other runes unchanged. func foldToASCII(s string) string { var b strings.Builder b.Grow(len(s)) for _, r := range s { // drop combining/modifier categories cat := unicode.Is(unicode.Mn, r) || unicode.Is(unicode.Mc, r) || unicode.Is(unicode.Me, r) if cat { continue } b.WriteRune(unicode.ToLower(r)) } return b.String() } // subdomainHasSlur returns true if the given subdomain contains a slur, // checking across several normalisation passes to defeat common bypass // attempts: // // 1. raw lowercased input // 2. separators (. - _) stripped // 3. trailing digits stripped // 4. separators stripped + trailing digits stripped // 5. leetspeak normalised variants of all of the above func HasSlur(subdomain string) bool { lower := strings.ToLower(subdomain) normalized := strings.NewReplacer(".", "", "-", "", "_", "").Replace(lower) stripped := stripTrailingDigits(lower) normalizedStripped := stripTrailingDigits(normalized) candidates := []string{ lower, normalized, stripped, normalizedStripped, normalizeLeetspeak(normalized), normalizeLeetspeak(normalizedStripped), } for _, pat := range getSlurPatterns() { for _, c := range candidates { if pat.MatchString(c) { return true } } } return false }