Apache
/
lucene
mirror of https://github.com/apache/lucene.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

SOLR-10264: Fixes multi-term synonym parsing in ManagedSynonymFilterFactory. 

(Jörg Rathlev, Steve Rowe, Christine Poerschke)
This commit is contained in:
Christine Poerschke 2017-04-07 11:11:23 +01:00
parent 206021578e
commit adabfdd9c2
4 changed files with 48 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
solr/CHANGES.txt
View File

@ -163,6 +163,9 @@ Bug Fixes
* SOLR-10323: fix to SpellingQueryConverter to properly strip out colons in field-specific queries.
(Amrit Sarkar via James Dyer)
* SOLR-10264: Fixes multi-term synonym parsing in ManagedSynonymFilterFactory.
(Jörg Rathlev, Steve Rowe, Christine Poerschke)
Other Changes
----------------------

7
solr/core/src/java/org/apache/solr/rest/schema/analysis/ManagedSynonymFilterFactory.java
View File

@ -33,6 +33,7 @@ import org.apache.lucene.analysis.synonym.SynonymFilterFactory;
import org.apache.lucene.analysis.synonym.SynonymMap;
import org.apache.lucene.analysis.util.ResourceLoader;
import org.apache.lucene.util.CharsRef;
import org.apache.lucene.util.CharsRefBuilder;
import org.apache.solr.common.SolrException;
import org.apache.solr.common.SolrException.ErrorCode;
import org.apache.solr.common.util.NamedList;
@ -358,9 +359,9 @@ public class ManagedSynonymFilterFactory extends BaseManagedTokenFilterFactory {
for (String term : cpsm.mappings.keySet()) {
for (String mapping : cpsm.mappings.get(term)) {
// apply the case setting to match the behavior of the SynonymMap builder
String casedTerm = synonymManager.applyCaseSetting(ignoreCase, term);
String casedMapping = synonymManager.applyCaseSetting(ignoreCase, mapping);
add(new CharsRef(casedTerm), new CharsRef(casedMapping), false);
CharsRef casedTerm = analyze(synonymManager.applyCaseSetting(ignoreCase, term), new CharsRefBuilder());
CharsRef casedMapping = analyze(synonymManager.applyCaseSetting(ignoreCase, mapping), new CharsRefBuilder());
add(casedTerm, casedMapping, false);
}
}
}

2
solr/core/src/test-files/solr/collection1/conf/schema-rest.xml
View File

@ -494,7 +494,7 @@
<fieldType name="location" class="solr.LatLonType" subFieldSuffix="_coordinate"/>
<!-- Field type where english stopwords are managed by the REST API -->
<fieldType name="managed_en" class="solr.TextField">
<fieldType name="managed_en" class="solr.TextField" autoGeneratePhraseQueries="false">
<analyzer>
<tokenizer class="solr.StandardTokenizerFactory"/>
<filter class="solr.ManagedStopFilterFactory" managed="english"/>

40
solr/core/src/test/org/apache/solr/rest/schema/analysis/TestManagedSynonymFilterFactory.java
View File

@ -16,6 +16,7 @@
*/
package org.apache.solr.rest.schema.analysis;
import java.io.File;
import java.net.URLEncoder;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.HashMap;
@ -154,13 +155,30 @@ public class TestManagedSynonymFilterFactory extends RestTestBase {
"count(/response/lst[@name='field']) = 1",
"/response/lst[@name='responseHeader']/int[@name='status'] = '0'");
// multi-term synonym logic - SOLR-10264
final String multiTermOrigin;
final String multiTermSynonym;
if (random().nextBoolean()) {
multiTermOrigin = "hansestadt hamburg";
multiTermSynonym = "hh";
} else {
multiTermOrigin = "hh";
multiTermSynonym = "hansestadt hamburg";
}
// multi-term logic similar to the angry/mad logic (angry ~ origin, mad ~ synonym)
assertU(adoc(newFieldName, "I am a happy test today but yesterday I was angry", "id", "5150"));
assertU(adoc(newFieldName, multiTermOrigin+" is in North Germany.", "id", "040"));
assertU(commit());
assertQ("/select?q=" + newFieldName + ":angry",
"/response/lst[@name='responseHeader']/int[@name='status'] = '0'",
"/response/result[@name='response'][@numFound='1']",
"/response/result[@name='response']/doc/str[@name='id'][.='5150']");
assertQ("/select?q=" + newFieldName + ":"+URLEncoder.encode(multiTermOrigin, "UTF-8"),
"/response/lst[@name='responseHeader']/int[@name='status'] = '0'",
"/response/result[@name='response'][@numFound='1']",
"/response/result[@name='response']/doc/str[@name='id'][.='040']");
// add a mapping that will expand a query for "mad" to match docs with "angry"
syns = new HashMap<>();
@ -172,12 +190,28 @@ public class TestManagedSynonymFilterFactory extends RestTestBase {
assertJQ(endpoint,
"/synonymMappings/managedMap/mad==['angry']");
// add a mapping that will expand a query for "multi-term synonym" to match docs with "acronym"
syns = new HashMap<>();
syns.put(multiTermSynonym, Arrays.asList(multiTermOrigin));
assertJPut(endpoint,
JSONUtil.toJSON(syns),
"/responseHeader/status==0");
assertJQ(endpoint+"/"+URLEncoder.encode(multiTermSynonym, "UTF-8"),
"/"+multiTermSynonym+"==['"+multiTermOrigin+"']");
// should not match as the synonym mapping between mad and angry does not
// get applied until core reload
assertQ("/select?q=" + newFieldName + ":mad",
"/response/lst[@name='responseHeader']/int[@name='status'] = '0'",
"/response/result[@name='response'][@numFound='0']");
// should not match as the synonym mapping between "origin" and "synonym"
// was not added before the document was indexed
assertQ("/select?q=" + newFieldName + ":("+URLEncoder.encode(multiTermSynonym, "UTF-8") + ")&sow=false",
"/response/lst[@name='responseHeader']/int[@name='status'] = '0'",
"/response/result[@name='response'][@numFound='0']");
restTestHarness.reload();
// now query for mad and we should see our test doc
@ -186,6 +220,12 @@ public class TestManagedSynonymFilterFactory extends RestTestBase {
"/response/result[@name='response'][@numFound='1']",
"/response/result[@name='response']/doc/str[@name='id'][.='5150']");
// now query for "synonym" and we should see our test doc with "origin"
assertQ("/select?q=" + newFieldName + ":("+URLEncoder.encode(multiTermSynonym, "UTF-8") + ")&sow=false",
"/response/lst[@name='responseHeader']/int[@name='status'] = '0'",
"/response/result[@name='response'][@numFound='1']",
"/response/result[@name='response']/doc/str[@name='id'][.='040']");
// test for SOLR-6015
syns = new HashMap<>();
syns.put("mb", Arrays.asList("megabyte"));