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

SOLR-6878: support adding symmetric synonym lists using the managed synonym API 

git-svn-id: https://svn.apache.org/repos/asf/lucene/dev/trunk@1677923 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Timothy Potter 2015-05-06 05:18:47 +00:00
parent 09f3cf8680
commit 74ebcfd134
3 changed files with 112 additions and 60 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
solr/CHANGES.txt
View File

@ -320,6 +320,11 @@ Other Changes
i.e. Construct, Init, and Call so the context of the request would be available after Init
and before the actual call operation. (Anshum Gupta, Noble Paul)
* SOLR-6878: Allow symmetric lists of synonyms to be added using the managed synonym REST
API to support legacy expand=true type mappings; previously the API only allowed adding
explicit mappings, with this feature you can now add a list and have the mappings
expanded when the update is applied (Timothy Potter, Vitaliy Zhovtyuk, hossman)
================== 5.1.0 ==================
Consult the LUCENE_CHANGES.txt file for additional, low level, changes in this release

136
solr/core/src/java/org/apache/solr/rest/schema/analysis/ManagedSynonymFilterFactory.java
View File

@ -56,7 +56,7 @@ public class ManagedSynonymFilterFactory extends BaseManagedTokenFilterFactory {
public static final String SYNONYM_MAPPINGS = "synonymMappings";
public static final String IGNORE_CASE_INIT_ARG = "ignoreCase";
/**
* Used internally to preserve the case of synonym mappings regardless
* of the ignoreCase setting.
@ -102,7 +102,7 @@ public class ManagedSynonymFilterFactory extends BaseManagedTokenFilterFactory {
implements ManagedResource.ChildResourceSupport
{
protected Map<String,CasePreservedSynonymMappings> synonymMappings;
public SynonymManager(String resourceId, SolrResourceLoader loader, StorageIO storageIO)
throws SolrException {
super(resourceId, loader, storageIO);
@ -126,13 +126,13 @@ public class ManagedSynonymFilterFactory extends BaseManagedTokenFilterFactory {
if (initArgs.get(IGNORE_CASE_INIT_ARG) == null) {
initArgs.add(IGNORE_CASE_INIT_ARG, Boolean.FALSE);
}
boolean ignoreCase = getIgnoreCase(managedInitArgs);
synonymMappings = new TreeMap<>();
if (managedData != null) {
Map<String,Object> storedSyns = (Map<String,Object>)managedData;
for (String key : storedSyns.keySet()) {
String caseKey = applyCaseSetting(ignoreCase, key);
CasePreservedSynonymMappings cpsm = synonymMappings.get(caseKey);
if (cpsm == null) {
@ -155,68 +155,100 @@ public class ManagedSynonymFilterFactory extends BaseManagedTokenFilterFactory {
}
}
log.info("Loaded {} synonym mappings for {}", synonymMappings.size(), getResourceId());
}
}
@SuppressWarnings("unchecked")
@Override
protected Object applyUpdatesToManagedData(Object updates) {
if (!(updates instanceof Map)) {
throw new ResourceException(Status.CLIENT_ERROR_BAD_REQUEST,
"Unsupported data format (" + updates.getClass().getName() + "); expected a JSON object (Map)!");
}
boolean ignoreCase = getIgnoreCase();
boolean ignoreCase = getIgnoreCase();
boolean madeChanges = false;
Map<String,Object> jsonMap = (Map<String,Object>)updates;
for (String term : jsonMap.keySet()) {
if (updates instanceof List) {
madeChanges = applyListUpdates((List<String>)updates, ignoreCase);
} else if (updates instanceof Map) {
madeChanges = applyMapUpdates((Map<String,Object>)updates, ignoreCase);
} else {
throw new ResourceException(Status.CLIENT_ERROR_BAD_REQUEST,
"Unsupported data format (" + updates.getClass().getName() + "); expected a JSON object (Map or List)!");
}
return madeChanges ? getStoredView() : null;
}
protected boolean applyListUpdates(List<String> jsonList, boolean ignoreCase) {
boolean madeChanges = false;
for (String term : jsonList) {
// find the mappings using the case aware key
String origTerm = term;
term = applyCaseSetting(ignoreCase, term);
// find the mappings using the case aware key
CasePreservedSynonymMappings cpsm = synonymMappings.get(term);
if (cpsm == null) {
if (cpsm == null)
cpsm = new CasePreservedSynonymMappings();
}
Set<String> output = cpsm.mappings.get(origTerm);
Object val = jsonMap.get(origTerm); // IMPORTANT: use the original
if (val instanceof String) {
String strVal = (String)val;
if (output == null) {
output = new TreeSet<>();
cpsm.mappings.put(origTerm, output);
}
if (output.add(strVal)) {
madeChanges = true;
}
} else if (val instanceof List) {
List<String> vals = (List<String>)val;
if (output == null) {
output = new TreeSet<>();
cpsm.mappings.put(origTerm, output);
}
for (String nextVal : vals) {
if (output.add(nextVal)) {
madeChanges = true;
}
}
} else {
throw new ResourceException(Status.CLIENT_ERROR_BAD_REQUEST, "Unsupported value "+val+
" for "+term+"; expected single value or a JSON array!");
}
Set<String> treeTerms = new TreeSet<>();
treeTerms.addAll(jsonList);
treeTerms.remove(origTerm);
cpsm.mappings.put(origTerm, treeTerms);
madeChanges = true;
// only add the cpsm to the synonymMappings if it has valid data
if (!synonymMappings.containsKey(term) && cpsm.mappings.get(origTerm) != null) {
synonymMappings.put(term, cpsm);
}
}
return madeChanges ? getStoredView() : null;
return madeChanges;
}
protected boolean applyMapUpdates(Map<String,Object> jsonMap, boolean ignoreCase) {
boolean madeChanges = false;
for (String term : jsonMap.keySet()) {
String origTerm = term;
term = applyCaseSetting(ignoreCase, term);
// find the mappings using the case aware key
CasePreservedSynonymMappings cpsm = synonymMappings.get(term);
if (cpsm == null)
cpsm = new CasePreservedSynonymMappings();
Set<String> output = cpsm.mappings.get(origTerm);
Object val = jsonMap.get(origTerm); // IMPORTANT: use the original
if (val instanceof String) {
String strVal = (String)val;
if (output == null) {
output = new TreeSet<>();
cpsm.mappings.put(origTerm, output);
}
if (output.add(strVal)) {
madeChanges = true;
}
} else if (val instanceof List) {
List<String> vals = (List<String>)val;
if (output == null) {
output = new TreeSet<>();
cpsm.mappings.put(origTerm, output);
}
for (String nextVal : vals) {
if (output.add(nextVal)) {
madeChanges = true;
}
}
} else {
throw new ResourceException(Status.CLIENT_ERROR_BAD_REQUEST, "Unsupported value "+val+
" for "+term+"; expected single value or a JSON array!");
}
// only add the cpsm to the synonymMappings if it has valid data
if (!synonymMappings.containsKey(term) && cpsm.mappings.get(origTerm) != null) {
synonymMappings.put(term, cpsm);
}
}
return madeChanges;
}
/**

31
solr/core/src/test/org/apache/solr/rest/schema/analysis/TestManagedSynonymFilterFactory.java
View File

@ -17,6 +17,7 @@ package org.apache.solr.rest.schema.analysis;
*/
import java.io.File;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.HashMap;
import java.util.List;
@ -189,17 +190,31 @@ public class TestManagedSynonymFilterFactory extends RestTestBase {
// test for SOLR-6015
syns = new HashMap<>();
syns.put("mb", Arrays.asList("megabyte"));
assertJPut(endpoint,
JSONUtil.toJSON(syns),
"/responseHeader/status==0");
assertJPut(endpoint,
JSONUtil.toJSON(syns),
"/responseHeader/status==0");
syns.put("MB", Arrays.asList("MiB", "Megabyte"));
assertJPut(endpoint,
JSONUtil.toJSON(syns),
"/responseHeader/status==0");
assertJPut(endpoint,
JSONUtil.toJSON(syns),
"/responseHeader/status==0");
assertJQ(endpoint+"/MB",
"/MB==['Megabyte','MiB','megabyte']");
assertJQ(endpoint + "/MB",
"/MB==['Megabyte','MiB','megabyte']");
// test for SOLR-6878 - by default, expand is true, but only applies when sending in a list
List<String> m2mSyns = new ArrayList<>();
m2mSyns.addAll(Arrays.asList("funny", "entertaining", "whimiscal", "jocular"));
assertJPut(endpoint, JSONUtil.toJSON(m2mSyns), "/responseHeader/status==0");
assertJQ(endpoint + "/funny",
"/funny==['entertaining','jocular','whimiscal']");
assertJQ(endpoint + "/entertaining",
"/entertaining==['funny','jocular','whimiscal']");
assertJQ(endpoint + "/jocular",
"/jocular==['entertaining','funny','whimiscal']");
assertJQ(endpoint + "/whimiscal",
"/whimiscal==['entertaining','funny','jocular']");
}
/**