From afda0bc563cdfdf8f22da318b5deffd121618269 Mon Sep 17 00:00:00 2001
From: anttil <anttil>
Date: Fri, 5 Feb 2016 11:29:17 +0000
Subject: [PATCH] SSDM-3147: Global search: XML stripping with wildcard mode
showed empty matches in result
SVN: 35642
---
.../dataaccess/db/HibernateSearchDAO.java | 53 +++++++++++++++++--
1 file changed, 49 insertions(+), 4 deletions(-)
diff --git a/openbis/source/java/ch/systemsx/cisd/openbis/generic/server/dataaccess/db/HibernateSearchDAO.java b/openbis/source/java/ch/systemsx/cisd/openbis/generic/server/dataaccess/db/HibernateSearchDAO.java
index 9f9ca664138..cc9d375da80 100644
--- a/openbis/source/java/ch/systemsx/cisd/openbis/generic/server/dataaccess/db/HibernateSearchDAO.java
+++ b/openbis/source/java/ch/systemsx/cisd/openbis/generic/server/dataaccess/db/HibernateSearchDAO.java
@@ -17,6 +17,7 @@
package ch.systemsx.cisd.openbis.generic.server.dataaccess.db;
import java.io.IOException;
+import java.io.StringReader;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Collections;
@@ -29,6 +30,10 @@ import java.util.Set;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
+import javax.xml.stream.XMLInputFactory;
+import javax.xml.stream.XMLStreamConstants;
+import javax.xml.stream.XMLStreamReader;
+
import org.apache.commons.lang.StringUtils;
import org.apache.log4j.Logger;
import org.apache.lucene.analysis.Analyzer;
@@ -73,6 +78,8 @@ import org.springframework.jdbc.support.JdbcAccessor;
import org.springframework.orm.hibernate4.HibernateCallback;
import org.springframework.orm.hibernate4.support.HibernateDaoSupport;
+import ch.systemsx.cisd.common.collection.CollectionUtils;
+import ch.systemsx.cisd.common.collection.CollectionUtils.ICollectionFilter;
import ch.systemsx.cisd.common.exceptions.UserFailureException;
import ch.systemsx.cisd.common.logging.LogCategory;
import ch.systemsx.cisd.common.logging.LogFactory;
@@ -113,6 +120,8 @@ final class HibernateSearchDAO extends HibernateDaoSupport implements IHibernate
private Map<String, DocValuesType> fieldTypesCache;
+ private final XMLInputFactory xif = XMLInputFactory.newFactory();
+
HibernateSearchDAO(final SessionFactory sessionFactory,
HibernateSearchContext hibernateSearchContext)
{
@@ -378,17 +387,18 @@ final class HibernateSearchDAO extends HibernateDaoSupport implements IHibernate
if (useWildcardSearchMode)
{
Pattern pattern = Pattern.compile("(?s)(^|\\s)" + q.toLowerCase().replace("*", ".*").replace("?", ".?") + "($|\\s)");
- Matcher matcher = pattern.matcher(content[i].toLowerCase());
+ String cont = stripXml(content[i].toLowerCase());
+ Matcher matcher = pattern.matcher(cont);
while (matcher.find())
{
int start = matcher.start();
int end = matcher.end();
- score += getScore(content[i].toLowerCase(), start, end, fields[i], useWildcardSearchMode);
+ score += getScore(cont, start, end, fields[i], useWildcardSearchMode);
}
} else
{
- String rest = content[i].toLowerCase();
+ String rest = stripXml(content[i].toLowerCase());
while (rest.length() > 0)
{
int start = rest.indexOf(q.toLowerCase());
@@ -483,6 +493,34 @@ final class HibernateSearchDAO extends HibernateDaoSupport implements IHibernate
return createMatchingEntity((Document) tuple[1], text, score);
}
+ private String stripXml(String cont)
+ {
+ if (cont.startsWith("<") && cont.endsWith(">"))
+ {
+
+ StringBuffer value = new StringBuffer();
+ try
+ {
+ XMLStreamReader xsr = xif.createXMLStreamReader(new StringReader(cont));
+ while (xsr.hasNext())
+ {
+ int x = xsr.next();
+ if (x == XMLStreamConstants.CHARACTERS)
+ {
+ value.append(xsr.getText() + " ");
+ }
+ }
+ } catch (Exception e)
+ {
+ value = new StringBuffer(cont);
+ }
+ return value.toString();
+ } else
+ {
+ return cont;
+ }
+ }
+
private double getScore(String text, int start, int end, String field, boolean wildcard)
{
boolean fullmatch = (start == 0 || !StringUtils.isAlphanumeric(text.substring(start - 1, start))) &&
@@ -600,8 +638,15 @@ final class HibernateSearchDAO extends HibernateDaoSupport implements IHibernate
List<?> list = hibernateQuery.list();
List<MatchingEntity> result = AbstractDAO.cast(list);
+ return CollectionUtils.filter(result, new ICollectionFilter<MatchingEntity>()
+ {
+ @Override
+ public boolean isPresent(MatchingEntity element)
+ {
+ return element != null && element.getScore() > 0.0;
+ }
+ });
- return filterNulls(result);
}
// detailed search
--
GitLab