vmario/docs
1
0
Fork 0
mirror of https://github.com/sismics/docs.git synced 2024-11-25 23:27:57 +01:00
Code Issues Packages Projects Releases Wiki Activity

#289: better search parsing (including wildcard and fuzzy)

Browse Source
This commit is contained in:
Benjamin Gamard 2019-01-24 17:26:46 +01:00
parent 10d5c4334b
commit 6e56a0f568
2 changed files with 29 additions and 19 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

47
docs-core/src/main/java/com/sismics/docs/core/util/indexing/LuceneIndexingHandler.java
View File

@ -25,8 +25,7 @@ import org.apache.lucene.document.Field;
import org.apache.lucene.document.StringField; import org.apache.lucene.document.StringField;
import org.apache.lucene.document.TextField; import org.apache.lucene.document.TextField;
import org.apache.lucene.index.*; import org.apache.lucene.index.*;
import org.apache.lucene.queryparser.flexible.standard.QueryParserUtil; import org.apache.lucene.queryparser.simple.SimpleQueryParser;
import org.apache.lucene.queryparser.flexible.standard.StandardQueryParser;
import org.apache.lucene.search.*; import org.apache.lucene.search.*;
import org.apache.lucene.search.highlight.Highlighter; import org.apache.lucene.search.highlight.Highlighter;
import org.apache.lucene.search.highlight.QueryScorer; import org.apache.lucene.search.highlight.QueryScorer;
@ -371,29 +370,26 @@ public class LuceneIndexingHandler implements IndexingHandler {
* @throws Exception e * @throws Exception e
*/ */
private Map<String, String> search(String searchQuery, String fullSearchQuery) throws Exception { private Map<String, String> search(String searchQuery, String fullSearchQuery) throws Exception {
// Escape query and add quotes so QueryParser generate a PhraseQuery // The fulltext query searches in all fields
String escapedSearchQuery = "\"" + QueryParserUtil.escape(searchQuery + " " + fullSearchQuery) + "\""; searchQuery = searchQuery + " " + fullSearchQuery;
String escapedFullSearchQuery = "\"" + QueryParserUtil.escape(fullSearchQuery) + "\"";
// Build search query // Build search query
Analyzer analyzer = new StandardAnalyzer(); Analyzer analyzer = new StandardAnalyzer();
StandardQueryParser qpHelper = new StandardQueryParser(analyzer);
qpHelper.setPhraseSlop(100); // PhraseQuery add terms
// Search on documents and files // Search on documents and files
BooleanQuery query = new BooleanQuery.Builder() BooleanQuery query = new BooleanQuery.Builder()
.add(qpHelper.parse(escapedSearchQuery, "title"), BooleanClause.Occur.SHOULD) .add(buildQueryParser(analyzer, "title").parse(searchQuery), BooleanClause.Occur.SHOULD)
.add(qpHelper.parse(escapedSearchQuery, "description"), BooleanClause.Occur.SHOULD) .add(buildQueryParser(analyzer, "description").parse(searchQuery), BooleanClause.Occur.SHOULD)
.add(qpHelper.parse(escapedSearchQuery, "subject"), BooleanClause.Occur.SHOULD) .add(buildQueryParser(analyzer, "subject").parse(searchQuery), BooleanClause.Occur.SHOULD)
.add(qpHelper.parse(escapedSearchQuery, "identifier"), BooleanClause.Occur.SHOULD) .add(buildQueryParser(analyzer, "identifier").parse(searchQuery), BooleanClause.Occur.SHOULD)
.add(qpHelper.parse(escapedSearchQuery, "publisher"), BooleanClause.Occur.SHOULD) .add(buildQueryParser(analyzer, "publisher").parse(searchQuery), BooleanClause.Occur.SHOULD)
.add(qpHelper.parse(escapedSearchQuery, "format"), BooleanClause.Occur.SHOULD) .add(buildQueryParser(analyzer, "format").parse(searchQuery), BooleanClause.Occur.SHOULD)
.add(qpHelper.parse(escapedSearchQuery, "source"), BooleanClause.Occur.SHOULD) .add(buildQueryParser(analyzer, "source").parse(searchQuery), BooleanClause.Occur.SHOULD)
.add(qpHelper.parse(escapedSearchQuery, "type"), BooleanClause.Occur.SHOULD) .add(buildQueryParser(analyzer, "type").parse(searchQuery), BooleanClause.Occur.SHOULD)
.add(qpHelper.parse(escapedSearchQuery, "coverage"), BooleanClause.Occur.SHOULD) .add(buildQueryParser(analyzer, "coverage").parse(searchQuery), BooleanClause.Occur.SHOULD)
.add(qpHelper.parse(escapedSearchQuery, "rights"), BooleanClause.Occur.SHOULD) .add(buildQueryParser(analyzer, "rights").parse(searchQuery), BooleanClause.Occur.SHOULD)
.add(qpHelper.parse(escapedSearchQuery, "filename"), BooleanClause.Occur.SHOULD) .add(buildQueryParser(analyzer, "filename").parse(searchQuery), BooleanClause.Occur.SHOULD)
.add(qpHelper.parse(escapedFullSearchQuery, "content"), BooleanClause.Occur.SHOULD) .add(buildQueryParser(analyzer, "content").parse(fullSearchQuery), BooleanClause.Occur.SHOULD)
.build(); .build();
// Search // Search
@ -435,6 +431,19 @@ public class LuceneIndexingHandler implements IndexingHandler {
return documentMap; return documentMap;
} }
/**
* Build a query parser for searching.
*
* @param analyzer Analyzer
* @param field Field
* @return Query parser
*/
private SimpleQueryParser buildQueryParser(Analyzer analyzer, String field) {
SimpleQueryParser simpleQueryParser = new SimpleQueryParser(analyzer, field);
simpleQueryParser.setDefaultOperator(BooleanClause.Occur.MUST); // AND all the terms
return simpleQueryParser;
}
/** /**
* Build Lucene document from database document. * Build Lucene document from database document.
* *

1
docs-web/src/test/java/com/sismics/docs/rest/TestDocumentResource.java
View File

@ -184,6 +184,7 @@ public class TestDocumentResource extends BaseJerseyTest {
// Search documents // Search documents
Assert.assertEquals(1, searchDocuments("full:uranium full:einstein", document1Token)); Assert.assertEquals(1, searchDocuments("full:uranium full:einstein", document1Token));
Assert.assertEquals(2, searchDocuments("tit*", document1Token));
Assert.assertEquals(2, searchDocuments("full:title", document1Token)); Assert.assertEquals(2, searchDocuments("full:title", document1Token));
Assert.assertEquals(2, searchDocuments("title", document1Token)); Assert.assertEquals(2, searchDocuments("title", document1Token));
Assert.assertEquals(1, searchDocuments("super description", document1Token)); Assert.assertEquals(1, searchDocuments("super description", document1Token));