Merge pull request #2940 from ntimo/task/split-bad-words

[RSPAMD] Split bad words into multiple files per language

data/conf/rspamd/custom/bad_words.map

@@ -1,9 +1,6 @@
-/\ssex\s/i
-/\svagina\s/i
 /\serotic\s/i
 /\serection\s/i
 /\ssexy\s/i
-/\spenis\s/i
 /\sass\s/i
 /\sviagra\s/i
 /\stits\s/i
@@ -17,28 +14,16 @@
 /\spharma\s/i
 /\spharmacy\s/i
 /\sseo\s/i
-/\smarketing\s/i
 /\sjackpot\s/i
-/\slotto\s/i
 /\slottery\s/i
-/pillenversand/i
-/\skredithilfe\s/i
-/\skapital\s/i
-/\skrankenversicherung\s/i
 /bitcoin/i
-/pädophil/i
-/paedophil/i
-/freiberufler/i
-/unternehmer/i
-/masturbieren/i
 /trojaner/i
 /malware/i
 /\sscooter\s/i
-/\sescooter\s/i
-/\se-scooter\s/i
 /testost/i
-/\spotenz\s/i
-/potenzmittel/i
-/rezeptfrei/i
-/apotheke/i
 /web\sdevelopment/i
+/\slottery\s/i
+/\ssex\s/i
+/\svagina\s/i
+/\spenis\s/i
+/\smarketing\s/i

data/conf/rspamd/custom/bad_words_de.map

@@ -0,0 +1,17 @@
+/\slotto\s/i
+/pillenversand/i
+/\skredithilfe\s/i
+/\skapital\s/i
+/\skrankenversicherung\s/i
+/pädophil/i
+/paedophil/i
+/freiberufler/i
+/unternehmer/i
+/masturbieren/i
+/\sescooter\s/i
+/\se-scooter\s/i
+/testost/i
+/\spotenz\s/i
+/potenzmittel/i
+/rezeptfrei/i
+/apotheke/i

data/conf/rspamd/local.d/composites.conf

@@ -27,6 +27,6 @@ OLEFY_MACRO {
   policy = "remove_weight";
 }
 BAD_WORD_BAD_TLD {
-  expression = "FISHY_TLD & BAD_WORDS"
+  expression = "FISHY_TLD & ( BAD_WORDS | BAD_WORDS_DE )"
   score = 10.0;
 }

data/conf/rspamd/local.d/multimap.conf

@@ -119,3 +119,11 @@ BAD_WORDS {
   regexp = true;
   score = 0.1;
 }
+
+BAD_WORDS_DE {
+  type = "content";
+  filter = "text";
+  map = "${LOCAL_CONFDIR}/custom/bad_words_de.map";
+  regexp = true;
+  score = 0.1;
+}