From: Vsevolod Stakhov <vsevolod@rambler-co.ru>
Date: Tue, 19 Jul 2011 13:51:28 +0000 (+0400)
Subject: * Add classifiers pre-selection script
X-Git-Tag: 0.4.0~9
X-Git-Url: https://source.dussan.org/?a=commitdiff_plain;h=4335e5a8ce4e23abd9c2977df7f1b13665b80202;p=rspamd.git

* Add classifiers pre-selection script
---

diff --git a/conf/lua/rspamd.classifiers.lua b/conf/lua/rspamd.classifiers.lua
new file mode 100644
index 000000000..e158a29a6
--- /dev/null
+++ b/conf/lua/rspamd.classifiers.lua
@@ -0,0 +1,48 @@
+-- Detect language of message and selects appropriate statfiles for it
+
+classifiers['bayes'] = function(classifier, task, is_learn, is_spam)
+	-- Subfunction for detection of message's language
+	local detect_language = function(task)
+		local parts = task:get_text_parts()
+		for _,p in ipairs(parts) do
+			local l = p:get_language()
+			if l then
+				return l
+			end
+		end
+		return nil
+	end
+
+	-- Main procedure
+	language = detect_language(task)
+	if language then
+		-- Find statfiles with specified language
+		local selected = {}
+		for _,st in pairs(classifier:get_statfiles()) do
+			local st_l = st:get_param('language')
+			if st_l and st_l == language then
+			    -- Insert statfile with specified language    
+			    table.insert(selected, st)
+			end
+		end
+		if table.maxn(selected) > 1 then
+			return selected
+		end
+	else
+		-- Language not detected
+		local selected = {}
+		for _,st in ipairs(classifier:get_statfiles()) do
+			local st_l = st:get_param('language')
+			-- Insert only statfiles without language
+			if not st_l then
+				table.insert(selected, st)
+			end
+		end
+		if table.maxn(selected) > 1 then
+			return selected
+		end
+	end
+
+	return nil
+end
+
diff --git a/conf/lua/rspamd.lua b/conf/lua/rspamd.lua
index 203862776..822f6447f 100644
--- a/conf/lua/rspamd.lua
+++ b/conf/lua/rspamd.lua
@@ -82,3 +82,7 @@ end
 if file_exists('rspamd.local.lua') then
 	dofile('rspamd.local.lua')
 end
+
+if file_exists('rspamd.classifiers.lua') then
+	dofile('rspamd.classifiers.lua')
+end