commit 54c63b93403cef7c75a246340e55443879e7634b
Author: Tanmai Khanna <khanna.tanmai@gmail.com>
Date:   Sun Jun 14 16:36:15 2020 +0530

    New feature:exclude-tags in pattern matching

diff --git a/samples/apertium-eng-spa.spa-eng.arx b/samples/apertium-eng-spa.spa-eng.arx
index 76023a1..38f4671 100644
--- a/samples/apertium-eng-spa.spa-eng.arx
+++ b/samples/apertium-eng-spa.spa-eng.arx
@@ -13,7 +13,7 @@
 			</anaphor>
 
 			<antecedent>
-				<parameter-item has-tags="n"/>
+				<parameter-item has-tags="n" exclude-tags="test"/>
 				<parameter-item has-tags="np"/>
 			</antecedent>
 		</def-parameter>
diff --git a/samples/test_exclude.txt b/samples/test_exclude.txt
new file mode 100644
index 0000000..2ea24fb
--- /dev/null
+++ b/samples/test_exclude.txt
@@ -0,0 +1,9 @@
+^El<det><def><m><pl>/The<det><def><m><pl>$ ^grup<n><m><pl><test>/group<n><pl>$ ^de<pr>/of<pr>/from<pr>$ 
+^el<det><def><m><sg>/the<det><def><m><sg>$ ^Parlament<n><m><sg>/Parliament<n><sg>$ 
+^haver<vbhaver><pri><p3><pl>/have<vbhaver><pri><p3><pl>$ 
+^mostrar<vblex><pp><m><sg>/show<vblex><pp><m><sg>/display<vblex><pp><m><sg>$ 
+^aquest<det><dem><m><sg>/this<det><dem><m><sg>$ ^dimarts<n><m><sp>/Tuesday<n><ND>$ 
+^.<sent>/.<sent>$ ^Tanmai<n><m><sg>/Tanmai<n><m><sg>$ ^es<cop><test>/is<cop>$ 
+^el seu<det><pos><m><sg>/his<det><pos><m><sg>$ ^suport<n><m><sg>/support<n><sg>$ 
+^a<pr>/at<pr>/in<pr>/to<pr>$ ^el<det><def><m><sg>/the<det><def><m><sg>$ 
+^*batle/*batle$ ^de<pr>/of<pr>/from<pr>$ ^el seu<det><pos><m><sg>/his<det><pos><m><sg>$ ^*Alaró/*Alaró$^.<sent>/.<sent>$
diff --git a/src/parse_arx.cc b/src/parse_arx.cc
index 4cc03b0..943512c 100644
--- a/src/parse_arx.cc
+++ b/src/parse_arx.cc
@@ -78,24 +78,27 @@ void ParseArx::parseParameterItem (xmlDocPtr doc, xmlNodePtr cur, wstring parame
 	xmlChar *Attr;
 	cur = cur->xmlChildrenNode;
 
-	vector <wstring> temp_tags_list;
+	pair< vector <wstring>, vector <wstring> > temp_tags_list;
 
 	while (cur != NULL)
 	{
 	    if ((!xmlStrcmp(cur->name, (const xmlChar *)"parameter-item")))
 	    {
 	    	Attr = xmlGetProp(cur, (const xmlChar *)"has-tags");
-
-	    	//fprintf(stderr, "ParameterItem: ");
-
-    		temp_tags_list = parseTags(XMLParseUtil::towstring(Attr));
+        temp_tags_list.first = parseTags(XMLParseUtil::towstring(Attr));
+        
+        Attr = xmlGetProp(cur, (const xmlChar *)"exclude-tags");
+        if (Attr)
+        {
+          temp_tags_list.second = parseTags(XMLParseUtil::towstring(Attr));
+        }
+        
     		parameters[parameter_type][parameter_name].push_back(temp_tags_list);
 
-	    	temp_tags_list.clear();
+        temp_tags_list.first.clear();
+        temp_tags_list.second.clear();
 
 		    xmlFree(Attr);
-
-		    //key = xmlNodeListGetString(doc, cur->xmlChildrenNode, 1);
  	    }
 
 		cur = cur->next;
@@ -163,23 +166,27 @@ void ParseArx::parseCatItem (xmlDocPtr doc, xmlNodePtr cur, wstring cat_name)
 	xmlChar *Attr;
 	cur = cur->xmlChildrenNode;
 
-	vector <wstring> temp_tags_list;
+	pair< vector <wstring>, vector <wstring> > temp_tags_list;
 
 	while (cur != NULL)
 	{
 	    if ((!xmlStrcmp(cur->name, (const xmlChar *)"cat-item")))
 	    {
-	    	Attr = xmlGetProp(cur, (const xmlChar *)"has-tags");
-	    	//fprintf(stderr, "catItem: ");
-
-	    	temp_tags_list = parseTags(XMLParseUtil::towstring(Attr));
+        Attr = xmlGetProp(cur, (const xmlChar *)"has-tags");
+        temp_tags_list.first = parseTags(XMLParseUtil::towstring(Attr));
+        
+        Attr = xmlGetProp(cur, (const xmlChar *)"exclude-tags");
+        if (Attr)
+        {
+          temp_tags_list.second = parseTags(XMLParseUtil::towstring(Attr));
+        }
 		    cats[cat_name].push_back(temp_tags_list);
 
-		    temp_tags_list.clear();
+        temp_tags_list.first.clear();
+        temp_tags_list.second.clear();
 
 		    xmlFree(Attr);
 
-		    //key = xmlNodeListGetString(doc, cur->xmlChildrenNode, 1);
  	    }
 
 		cur = cur->next;
@@ -279,13 +286,11 @@ void ParseArx::parsePatterns (xmlDocPtr doc, xmlNodePtr cur, wstring markable_na
 
 	    	if (parameter_name)
 	    	{
-	    		//cerr << "HELLO!";
 	    		wstring parameter_name_ws = XMLParseUtil::towstring(parameter_name);
 	    		parameter_markables_score[parameter_name_ws][markable_name] = score_int;
 	    	}
 	    	else
 	    	{
-	    		//cerr << "FUCK!";
 	    		all_markables_score[markable_name] = score_int;
 	    	}
 	    }
diff --git a/src/parse_arx.h b/src/parse_arx.h
index a43190d..90a39ab 100644
--- a/src/parse_arx.h
+++ b/src/parse_arx.h
@@ -30,7 +30,7 @@
 
 using namespace std;
 
-typedef vector< vector<wstring> > acceptable_tags;
+typedef vector< pair < vector<wstring>, vector<wstring> > > acceptable_tags; //a vector of pairs of tags to match and exclude
 
 struct markable_pattern
 {
@@ -78,4 +78,4 @@ public:
 	unordered_map<wstring, int> get_parameter_markables_score(wstring parameter_name);
 };
 
-#endif
\ No newline at end of file
+#endif
diff --git a/src/pattern_arx.cc b/src/pattern_arx.cc
index 5c86f8f..a1d8a72 100644
--- a/src/pattern_arx.cc
+++ b/src/pattern_arx.cc
@@ -67,31 +67,33 @@ int check_acceptable_tags(vector<wstring> input_tags, acceptable_tags check_tags
 
 		int flag_contains_all = 1;
 
-		vector<wstring> temp_tags = *i;
-
-		for(std::vector<wstring>::iterator j = temp_tags.begin(); j != temp_tags.end(); ++j)
+    vector<wstring> temp_tags = i->first;
+    vector<wstring> temp_exclude_tags = i->second;
+    
+		for(std::vector<wstring>::iterator j = temp_tags.begin(); j != temp_tags.end(); ++j) //check for the tags in has-tags
 		{
-
 			if(*j == L"*") //ignore * in the tags list
 				continue;
 
-			if(!contains(input_tags, *j)) //if the required tag is NOT in the input LU tags
+			if(!contains(input_tags, *j)) //if the has-tag is NOT in the input LU tags
 			{
 				flag_contains_all = 0;
 				break;
 			}
-			/*
-			else
-			{
-				cerr << "FoundTag:";
-				wcerr << *j;
-				cerr <<"\n";
-			}
-			*/
 		}
-
-		if(flag_contains_all == 1) //if any tag list fully matched
-			return 1; //else continue to next tag list
+    
+    for(std::vector<wstring>::iterator j = temp_exclude_tags.begin(); j != temp_exclude_tags.end(); ++j) //check for the tags in exclude-tags
+    {
+      if(contains(input_tags, *j)) //if the exclude-tag IS in the input LU tags
+      {
+        flag_contains_all = 0;
+        break;
+      }
+    }
+
+		if(flag_contains_all == 1) //if any tag list fully matched (i.e. has-tags present, exclude-tags absent)
+			return 1;
+    //else continue to next tag list
 	}
 
 	return 0; //if it didn't return 1 then no tag list was fully matched