Changeset 29167 for main/trunk/greenstone2

Timestamp:

2014-07-31T14:00:03+12:00 (10 years ago)

Author:

ak19

Message:

Fix to nullpointer exception when there are no term docs, such as for search term bla. Also some comments on how to get all the matching docs in lucene 4.7.2

File:

• main/trunk/greenstone2/common-src/indexers/lucene-gs/src/org/greenstone/LuceneWrapper4/GS2LuceneQuery.java (modified) (2 diffs)

main/trunk/greenstone2/common-src/indexers/lucene-gs/src/org/greenstone/LuceneWrapper4/GS2LuceneQuery.java

@@ -266 +266 @@
         int term_freq = 0;
         int match_docs = 0;
-        
-        int docID = -1;
-        while((docID = term_docs.nextDoc()) != DocsEnum.NO_MORE_DOCS) {//while (term_docs.next())   
-            if (term_docs.freq() != 0)
-            {
-            term_freq += term_docs.freq();
-            match_docs++;
-
-            // Calculate the document-level term frequency as well
-            Integer lucene_doc_num_obj = new Integer(term_docs.docID());
-            int doc_term_freq = 0;
-                        if (doc_term_freq_map.containsKey(lucene_doc_num_obj))
-            {
-                doc_term_freq = ((Integer) doc_term_freq_map.get(lucene_doc_num_obj)).intValue();
-            }
-            doc_term_freq += term_docs.freq();
-
-            doc_term_freq_map.put(lucene_doc_num_obj, new Integer(doc_term_freq));
-            }
+
+        if(term_docs != null) {
+            int docID = -1;
+            while((docID = term_docs.nextDoc()) != DocsEnum.NO_MORE_DOCS) {//while (term_docs.next())   
+            if (term_docs.freq() != 0)
+                {
+                term_freq += term_docs.freq();
+                match_docs++;
+                
+                // Calculate the document-level term frequency as well
+                Integer lucene_doc_num_obj = new Integer(term_docs.docID());
+                int doc_term_freq = 0;
+                if (doc_term_freq_map.containsKey(lucene_doc_num_obj))
+                    {
+                    doc_term_freq = ((Integer) doc_term_freq_map.get(lucene_doc_num_obj)).intValue();
+                    }
+                doc_term_freq += term_docs.freq();
+                
+                doc_term_freq_map.put(lucene_doc_num_obj, new Integer(doc_term_freq));
+                }
+            }
+        } else {
+            System.err.println("@@@ GS2LuceneQuery.java: term_docs is null for term " + term.text());
         }
 
@@ -302 +306 @@
         }
         
-        // do the query
+        // Extracting all documents for a given search - http://www.gossamer-threads.com/lists/lucene/java-user/134873
+        // http://lucene.apache.org/core/3_4_0/api/core/org/apache/lucene/search/TotalHitCountCollector.html
+        // http://lucene.apache.org/core/4_7_2/core/index.html?org/apache/lucene/search/TopFieldDocs.html
+
+        // 1. Figure out how many results there will be.
+        //TotalHitCountCollecter countCollector = new TotalHitCountCollector();
+        //searcher.search(query, filter, collector);
+        //int hitCount = collector.count;
+
+        // Actually do the query
         // Simple case for getting all the matching documents
         if (end_results == Integer.MAX_VALUE) {