solr4.7.2开发实践 ——拼写检查spellcheck【需要修改为自己配置的】

文章来源：http://my.oschina.net/MrMichael/blog/261242

①拼写检查不同于其他域，它在建立索引时需要分词，但是检索时不需要分词，所以要建立一个特殊的域，以供拼写检查：

在schema.xml文件里设置所需的拼写检查域都有哪些字段：

?

1 2 3 4 5 6 7 8 9 10 11 12 13 14

< field  name = "spell"  type = "text_spell"  indexed = "true"  stored = "false"  multiValued = "true"  />   < copyField  source = "name"  dest = "spell" />   < copyField  source = "content"  dest = "spell" />   < fieldType  name = "text_spell"  class = "solr.TextField"  positionIncrementGap = "100" >       < analyzer  type = "index" >       < tokenizer  class = "org.wltea.analyzer.lucene.IKTokenizerFactory"  useSmart = "false"  conf = "ik.conf" />       < filter  class = "solr.StopFilterFactory"  ignoreCase = "true"  words = "stopwords.txt"  enablePositionIncrements = "true"  />       </ analyzer >       < analyzer  type = "query" >         < tokenizer  class = "solr.WhitespaceTokenizerFactory" />       </ analyzer >    </ fieldType >

②在solrconfig.xml文件里设置：

?

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

< searchComponent  name = "spellerror"  class = "solr.SpellCheckComponent" >      < str  name = "queryAnalyzerFieldType" >text_spell</ str >      <!-- a spellchecker built from a field of the main index   -->       < lst  name = "spellchecker" >        < str  name = "name" >default</ str >        < str  name = "field" >spell</ str >        < str  name = "classname" >solr.DirectSolrSpellChecker</ str >        < str  name = "distanceMeasure" >internal</ str >        < float  name = "accuracy" >0.5</ float >        < int  name = "maxEdits" >2</ int >        < int  name = "minPrefix" >1</ int >        < int  name = "maxInspections" >5</ int >        < int  name = "minQueryLength" >2</ int >        < float  name = "maxQueryFrequency" >0.01</ float >      </ lst >   < lst  name = "spellchecker" >     < str  name = "classname" >solr.FileBasedSpellChecker</ str >     < str  name = "name" >file</ str >     < str  name = "sourceLocation" >spellings.txt</ str >     < str  name = "characterEncoding" >UTF-8</ str >     < str  name = "spellcheckIndexDir" >spellcheckerFile</ str >      </ lst >    </ searchComponent >    < requestHandler  name = "/spell"  class = "solr.SearchHandler"  startup = "lazy" >      < lst  name = "defaults" >        < str  name = "df" >spell</ str >        < str  name = "spellcheck.dictionary" >default</ str >        < str  name = "spellcheck" >on</ str >        <!-- <str name="spellcheck.extendedResults">true</str>               <str name="spellcheck.count">10</str>        <str name="spellcheck.alternativeTermCount">5</str>        <str name="spellcheck.maxResultsForSuggest">5</str> -->               < str  name = "spellcheck.collate" >true</ str >        < str  name = "spellcheck.collateExtendedResults" >true</ str >          <!-- <str name="spellcheck.maxCollationTries">10</str>        <str name="spellcheck.maxCollations">5</str> -->                </ lst >      < arr  name = "last-components" >        < str >spellerror</ str >      </ arr >    </ requestHandler >

③solrj里的代码

?

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33

/**      * @method: testSpellCheck      * @Description: 拼写检查       *       * @author: ChenYW      * @date 2014-4-15 下午06:14:56      */      public  String spellCheck(String word){          SolrQuery query =  new  SolrQuery();            query.set( "defType" , "edismax" ); //加权          query.set( "qf" , "name^20.0" );                    query.set( "spellcheck" ,  "true" );            query.set( "spellcheck.q" , word);          query.set( "qt" ,  "/spell" );            query.set( "spellcheck.build" ,  "true" ); //遇到新的检查词，会自动添加到索引里面            query.set( "spellcheck.count" ,  5 );                       try  {             QueryResponse rsp = server.query(query);             SpellCheckResponse re=rsp.getSpellCheckResponse();             if  (re !=  null ) {              if (!re.isCorrectlySpelled()){                String t = re.getFirstSuggestion(word); //获取第一个推荐词         System.out.println( "推荐词："  + t);       return  t;            }                             }           }  catch  (SolrServerException e) {                e.printStackTrace();            }            return  null ;      }