added search spelling suggestions (UI needs work)

scripts/search/search

@@ -1,2 +1,6 @@
 #!/bin/sh
-curl 'http://localhost:8080/api/search?q=*'
+if [ -z "$1" ]; then
+ curl 'http://localhost:8080/api/search?q=*'
+else
+ curl "http://localhost:8080/api/search?$1"
+fi

scripts/search/spellcheck

@@ -0,0 +1,5 @@
+#!/bin/sh
+# output:
+# "hits",1,
+# "misspellingsAndCorrections",["datvrse","dataverse"]
+curl -s 'http://localhost:8983/solr/spell?spellcheck=true&wt=json&indent=true&q=datvrse'

src/main/java/edu/harvard/iq/dataverse/SearchPage.java

@@ -2,6 +2,7 @@
 
 import java.util.ArrayList;
 import java.util.List;
+import java.util.Map;
 import java.util.logging.Logger;
 import javax.ejb.EJB;
 import javax.faces.view.ViewScoped;
@@ -18,6 +19,7 @@ public class SearchPage implements java.io.Serializable {
  private List<Dataset> datasets = new ArrayList<>();
  private List<DataverseUser> dataverseUsers = new ArrayList<>();
  private List<DataFile> dataFiles = new ArrayList<>();
+ private List<String> spelling_alternatives = new ArrayList<>();
 
  @EJB
  SearchServiceBean searchService;
@@ -39,14 +41,20 @@ public void search() {
  datasets = new ArrayList();
 
  query = query == null ? "*" : query;
- List<SolrSearchResult> searchResults = searchService.search(query);
+ SolrQueryResponse solrQueryResponse = searchService.search(query);
+ List<SolrSearchResult> searchResults = solrQueryResponse.getSolrSearchResults();
+ for (Map.Entry<String, List<String>> entry : solrQueryResponse.getSpellingSuggestionsByToken().entrySet()) {
+ spelling_alternatives.add(entry.getValue().toString());
+ }
  for (SolrSearchResult searchResult : searchResults) {
  String type = searchResult.getType();
  switch (type) {
  case "dataverses":
  Dataverse dataverse = dataverseService.find(searchResult.getEntityId());
  if (searchResult.getHighlightSnippets() != null) {
- /** @todo when does long description truncate? */
+ /**
+ * @todo when does long description truncate?
+ */
  dataverse.setDescription(searchResult.getHighlightSnippets().get(0));
  }
  dataverses.add(dataverse);
@@ -124,4 +132,12 @@ public void setDataFiles(List<DataFile> dataFiles) {
  this.dataFiles = dataFiles;
  }
 
+ public List<String> getSpelling_alternatives() {
+ return spelling_alternatives;
+ }
+
+ public void setSpelling_alternatives(List<String> spelling_alternatives) {
+ this.spelling_alternatives = spelling_alternatives;
+ }
+
 }

src/main/java/edu/harvard/iq/dataverse/SearchServiceBean.java

@@ -3,8 +3,10 @@
 import edu.harvard.iq.dataverse.api.SearchFields;
 import java.util.ArrayList;
 import java.util.Collection;
+import java.util.HashMap;
 import java.util.Iterator;
 import java.util.List;
+import java.util.Map;
 import java.util.logging.Logger;
 import javax.ejb.Stateless;
 import javax.inject.Named;
@@ -13,6 +15,7 @@
 import org.apache.solr.client.solrj.SolrServerException;
 import org.apache.solr.client.solrj.impl.HttpSolrServer;
 import org.apache.solr.client.solrj.response.QueryResponse;
+import org.apache.solr.client.solrj.response.SpellCheckResponse;
 import org.apache.solr.common.SolrDocument;
 import org.apache.solr.common.SolrDocumentList;
 
@@ -28,7 +31,7 @@ public class SearchServiceBean {
 
  private static final Logger logger = Logger.getLogger(SearchServiceBean.class.getCanonicalName());
 
- public List<SolrSearchResult> search(String queryFromUser) {
+ public SolrQueryResponse search(String queryFromUser) {
  /**
  * @todo make "localhost" and port number a config option
  */
@@ -37,6 +40,7 @@ public List<SolrSearchResult> search(String queryFromUser) {
  solrQuery.setQuery(queryFromUser);
  solrQuery.setHighlight(true).setHighlightSnippets(1);
  solrQuery.setParam("hl.fl", SearchFields.DESCRIPTION);
+ solrQuery.setParam("qt", "/spell");
 
  QueryResponse queryResponse;
  try {
@@ -74,7 +78,18 @@ public List<SolrSearchResult> search(String queryFromUser) {
  solrSearchResult.setType(type);
  solrSearchResults.add(solrSearchResult);
  }
- return solrSearchResults;
- }
+ Map<String, List<String>> spellingSuggestionsByToken = new HashMap<>();
+ SpellCheckResponse spellCheckResponse = queryResponse.getSpellCheckResponse();
+ if (spellCheckResponse != null) {
+ List<SpellCheckResponse.Suggestion> suggestions = spellCheckResponse.getSuggestions();
+ for (SpellCheckResponse.Suggestion suggestion : suggestions) {
+ spellingSuggestionsByToken.put(suggestion.getToken(), suggestion.getAlternatives());
+ }
+ }
 
+ SolrQueryResponse solrQueryResponse = new SolrQueryResponse();
+ solrQueryResponse.setSolrSearchResults(solrSearchResults);
+ solrQueryResponse.setSpellingSuggestionsByToken(spellingSuggestionsByToken);
+ return solrQueryResponse;
+ }
 }

src/main/java/edu/harvard/iq/dataverse/SolrQueryResponse.java

@@ -0,0 +1,27 @@
+package edu.harvard.iq.dataverse;
+
+import java.util.List;
+import java.util.Map;
+
+public class SolrQueryResponse {
+
+ private List<SolrSearchResult> solrSearchResults;
+ private Map<String, List<String>> spellingSuggestionsByToken;
+
+ public List<SolrSearchResult> getSolrSearchResults() {
+ return solrSearchResults;
+ }
+
+ public void setSolrSearchResults(List<SolrSearchResult> solrSearchResults) {
+ this.solrSearchResults = solrSearchResults;
+ }
+
+ public Map<String, List<String>> getSpellingSuggestionsByToken() {
+ return spellingSuggestionsByToken;
+ }
+
+ public void setSpellingSuggestionsByToken(Map<String, List<String>> spellingSuggestionsByToken) {
+ this.spellingSuggestionsByToken = spellingSuggestionsByToken;
+ }
+
+}

src/main/java/edu/harvard/iq/dataverse/api/Search.java

@@ -2,12 +2,15 @@
 
 import edu.harvard.iq.dataverse.SolrSearchResult;
 import edu.harvard.iq.dataverse.SearchServiceBean;
+import edu.harvard.iq.dataverse.SolrQueryResponse;
 import java.util.List;
+import java.util.Map;
 import java.util.logging.Logger;
 import javax.ejb.EJB;
 import javax.json.Json;
 import javax.json.JsonArrayBuilder;
 import javax.json.JsonObject;
+import javax.json.JsonObjectBuilder;
 import javax.ws.rs.GET;
 import javax.ws.rs.Path;
 import javax.ws.rs.QueryParam;
@@ -24,14 +27,24 @@ public class Search {
 // public JsonObject search(@QueryParam("q") String query) {
  public String search(@QueryParam("q") String query) {
  if (query != null) {
- List<SolrSearchResult> solrSearchResults = searchService.search(query);
+ SolrQueryResponse solrQueryResponse = searchService.search(query);
+
  JsonArrayBuilder filesArrayBuilder = Json.createArrayBuilder();
+ List<SolrSearchResult> solrSearchResults = solrQueryResponse.getSolrSearchResults();
  for (SolrSearchResult solrSearchResult : solrSearchResults) {
  filesArrayBuilder.add(solrSearchResult.toJsonObject());
+
  }
+
+ JsonObjectBuilder spelling_alternatives = Json.createObjectBuilder();
+ for (Map.Entry<String, List<String>> entry : solrQueryResponse.getSpellingSuggestionsByToken().entrySet()) {
+ spelling_alternatives.add(entry.getKey(), entry.getValue().toString());
+ }
+
  JsonObject value = Json.createObjectBuilder()
  .add("total_count", solrSearchResults.size())
  .add("items", solrSearchResults.toString())
+ .add("spelling_alternatives", spelling_alternatives)
  .add("itemsJson", filesArrayBuilder.build())
  .build();
  logger.info("value: " + value);

src/main/webapp/search.xhtml

@@ -47,6 +47,7 @@
 
  <!-- Content Panel -->
  <p:panel id="content">
+ <h:outputText style="color: red" value="Did you mean #{SearchPage.spelling_alternatives}?" rendered="#{SearchPage.spelling_alternatives.size() != 0}"/>
  <p:tabView id="tabView" widgetVar="content">
  <p:tab id="dataversesTab" title="Dataverses">
  <p:dataTable id="dataversesTable" value="#{SearchPage.dataverses}" var="dataverse">