Archives/gnosygnu_xowa
1
0
Fork 0
mirror of https://github.com/gnosygnu/xowa.git synced 2026-03-02 03:49:30 +00:00
Code Issues Projects Releases Wiki Activity

Full-text search: Change to lucene 6.4.2

Browse Source
This commit is contained in:
gnosygnu
2017-03-15 19:51:58 -04:00
parent 8f8e414c80
commit 8c31c8fd62
14 changed files with 136 additions and 149 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
gplx.gflucene/.classpath
View File

@@ -3,10 +3,10 @@
<classpathentry kind="src" path="src"/>
<classpathentry combineaccessrules="false" kind="src" path="/100_core"/>
<classpathentry kind="con" path="org.eclipse.jdt.launching.JRE_CONTAINER"/>
<classpathentry kind="lib" path="lib/lucene-analyzers-common-6.4.2.jar"/>
<classpathentry kind="lib" path="lib/lucene-core-6.4.2.jar"/>
<classpathentry kind="lib" path="lib/lucene-highlighter-6.4.2.jar"/>
<classpathentry kind="lib" path="lib/lucene-memory-6.4.2.jar"/>
<classpathentry kind="lib" path="lib/lucene-queryparser-6.4.2.jar"/>
<classpathentry kind="lib" path="lib/6.4.2/lucene-analyzers-common-6.4.2.jar"/>
<classpathentry kind="lib" path="lib/6.4.2/lucene-core-6.4.2.jar"/>
<classpathentry kind="lib" path="lib/6.4.2/lucene-highlighter-6.4.2.jar"/>
<classpathentry kind="lib" path="lib/6.4.2/lucene-memory-6.4.2.jar"/>
<classpathentry kind="lib" path="lib/6.4.2/lucene-queryparser-6.4.2.jar"/>
<classpathentry kind="output" path="bin"/>
</classpath>

16
gplx.gflucene/src/gplx/gflucene/highlighters/Gflucene_highlighter_mgr.java
View File

@@ -32,10 +32,13 @@ import org.apache.lucene.search.IndexSearcher;
import org.apache.lucene.search.Query;
import org.apache.lucene.search.TopDocs;
import org.apache.lucene.search.highlight.Formatter;
import org.apache.lucene.search.highlight.Fragmenter;
import org.apache.lucene.search.highlight.Highlighter;
import org.apache.lucene.search.highlight.InvalidTokenOffsetsException;
import org.apache.lucene.search.highlight.QueryScorer;
import org.apache.lucene.search.highlight.SimpleFragmenter;
import org.apache.lucene.search.highlight.SimpleHTMLFormatter;
import org.apache.lucene.search.highlight.SimpleSpanFragmenter;
import org.apache.lucene.search.highlight.TextFragment;
import org.apache.lucene.search.highlight.TokenSources;
import org.apache.lucene.store.FSDirectory;
@@ -60,16 +63,21 @@ public class Gflucene_highlighter_mgr {
// create highlighter
SimpleHTMLFormatter htmlFormatter = new SimpleHTMLFormatter("<span class='snip_highlight'>", "</span>");
Highlighter highlighter = new Highlighter(htmlFormatter, new QueryScorer(query));
QueryScorer scorer = new QueryScorer(query);
scorer.setExpandMultiTermQuery(false);
Highlighter highlighter = new Highlighter(htmlFormatter, scorer);
SimpleFragmenter fragmenter = new SimpleFragmenter(100);
highlighter.setTextFragmenter(fragmenter);
// get token stream
String text = doc_data.body;
TokenStream tokenStream = analyzer.tokenStream("body", text);
// get fragments from stream
String[] frags;
TextFragment[] frags;
try {
frags = highlighter.getBestFragments(tokenStream, text, 10);
// frags = highlighter.getBestTextFragments(tokenStream, text, false, 1000);
frags = highlighter.getBestTextFragments(tokenStream, text, true, 10);
} catch (IOException e) {
throw Err_.new_exc(e, "lucene_index", "failed to get best", "query", qry_data.query);
} catch (InvalidTokenOffsetsException e) {
@@ -80,7 +88,7 @@ public class Gflucene_highlighter_mgr {
int frags_len = frags.length;
Gflucene_highlighter_item[] array = new Gflucene_highlighter_item[frags_len];
for (int i = 0; i < frags_len; i++) {
String frag = frags[i];
String frag = frags[i].toString();
array[i] = new Gflucene_highlighter_item(i, frag);
}
return array;