Changeset 12376

Timestamp:

2006-08-02T15:05:48+12:00 (18 years ago)

Author:

mdewsnip

Message:

Now stores term vectors (for the TX field only), to support query term occurrences. Many thanks to John Thompson and DL Consulting Ltd.

Location:

trunk

Files:

• gsdl/src/java/org/nzdl/gsdl/LuceneWrap/Indexer.java (modified) (1 diff)
• indexers/lucene-gs/src/org/greenstone/LuceneWrapper/Indexer.java (modified) (1 diff)

trunk/gsdl/src/java/org/nzdl/gsdl/LuceneWrap/Indexer.java

@@ -155 +155 @@
     public void endElement(String uri, String localName, String qName) throws SAXException 
     {
-    if (qName.equals(indexable_current_node_)) {
-        current_doc_.add(new Field(qName, current_contents_, Field.Store.NO, Field.Index.TOKENIZED));
-        current_contents_ = "";
-    }
+    if (qName.equals(indexable_current_node_))
+            {
+                current_doc_.add(new Field(qName, current_contents_, Field.Store.NO, Field.Index.TOKENIZED, Field.TermVector.YES));
+                // We only need the term vector for the TX field
+                if (!qName.equals("TX"))
+                    {
+                        current_doc_.add(new Field("by" + qName, current_contents_, Field.Store.NO, Field.Index.UN_TOKENIZED, Field.TermVector.NO));
+                    }
+
+                current_contents_ = "";
+            }
 
     if (qName.equals(doc_tag_level_)) {

trunk/indexers/lucene-gs/src/org/greenstone/LuceneWrapper/Indexer.java

@@ -155 +155 @@
     public void endElement(String uri, String localName, String qName) throws SAXException 
     {
-    if (qName.equals(indexable_current_node_)) {
-        current_doc_.add(new Field(qName, current_contents_, Field.Store.NO, Field.Index.TOKENIZED));
-        current_contents_ = "";
-    }
+    if (qName.equals(indexable_current_node_))
+            {
+                current_doc_.add(new Field(qName, current_contents_, Field.Store.NO, Field.Index.TOKENIZED, Field.TermVector.YES));
+                // We only need the term vector for the TX field
+                if (!qName.equals("TX"))
+                    {
+                        current_doc_.add(new Field("by" + qName, current_contents_, Field.Store.NO, Field.Index.UN_TOKENIZED, Field.TermVector.NO));
+                    }
+
+                current_contents_ = "";
+            }
 
     if (qName.equals(doc_tag_level_)) {