From: Vsevolod Stakhov <vsevolod@rspamd.com>
Date: Sun, 2 Jul 2023 16:33:28 +0000 (+0100)
Subject: [Minor] Lua_magic: Try to reduce fp rate for html heuristic
X-Git-Tag: 3.6~46
X-Git-Url: http://git.ipfire.org/?a=commitdiff_plain;h=42873dc11090b0ce0c6f49333a73123fa5de05cb;p=thirdparty%2Frspamd.git

[Minor] Lua_magic: Try to reduce fp rate for html heuristic
---

diff --git a/lualib/lua_magic/heuristics.lua b/lualib/lua_magic/heuristics.lua
index ea8f08e081..98cfb0eee9 100644
--- a/lualib/lua_magic/heuristics.lua
+++ b/lualib/lua_magic/heuristics.lua
@@ -63,13 +63,13 @@ local zip_patterns = {
 local txt_trie
 local txt_patterns = {
   html = {
-    {[[(?i)<html\b]], 32},
+    {[=[(?i)<html[\s>]]=], 32},
     {[[(?i)<script\b]], 20}, -- Commonly used by spammers
     {[[<script\s+type="text\/javascript">]], 31}, -- Another spammy pattern
     {[[(?i)<\!DOCTYPE HTML\b]], 33},
     {[[(?i)<body\b]], 20},
     {[[(?i)<table\b]], 20},
-    {[[(?i)<a\b]], 10},
+    {[[(?i)<a\s]], 10},
     {[[(?i)<p\b]], 10},
     {[[(?i)<div\b]], 10},
     {[[(?i)<span\b]], 10},