[Lucene.Net] Search in Azure

2014-06-18

1475
0
C#
2017-01-10

Azure平台支援較普遍技術，如你現有系統有在使用Hadoop的話就可透過HDInsight來快速整合，

全文檢索方面Azure也不例外支援Lucene.net，

如要提高Lucene.net Search全文檢索效能，I/O 和分散查詢絕對是關鍵，

仗者Azure都是SSD加龐大硬體資源，讓開發人員可更專心在商業邏輯上，

下面我簡單使用asp.net mvc+lucene.net+Azure快速打造全文檢索。

1.安裝Lucene.net和Azure

install-package Lucene.Net

install-package Lucene.Net.Store.Azure

2.新增文件to Azure Blob

Model

public class LuceneModels { public string EdocId { get; set; } public string CreateUser { get; set; } public string CreateDate { get; set; } public string Content { get; set; } }

View

<h2>Lucene.net On Azure</h2> <img src="~/Images/ajax-loader.gif" id="processimg" style="display:none" /> <input type="button" value="新增文件" onclick="return AddDoc();" /> <label id="labmessage"></label><br/> <input type="button" value="全文檢索" onclick="return mysearch();" /> <input type="text" id="txtkw" /> <div id="resultdiv"> @Html.Partial("_FullText") </div> @section scripts { <script type="text/javascript"> function mysearch() { var keyword = $('#txtkw').val(); $.ajax({ type: 'post', url: '/Home/SearchResult?kw=' + escape(keyword), data: '', contentType: 'json', dataType: 'html', cache: false, async: true, beforeSend: function () { $('#processimg').show(); }, success: function (data) { $('#processimg').hide(); $('#resultdiv').html(data); return false; }, error: function (jqXHR, errorThrown) { $('#processimg').hide(); alert(errorThrown); return false; } }); } function AddDoc() { $.ajax({ type: 'post', url: '/Home/AddDoc', data: '', contentType: 'json', dataType: 'json', cache: false, async: true, beforeSend: function () { $('#processimg').show(); }, success: function (result) { var response = $(result.Data); $('#labmessage').text(response); $('#processimg').hide(); return false; }, error: function (jqXHR, errorThrown) { $('#processimg').hide(); alert(errorThrown); return false; } }); } </script> }

Controller

Add Document to Azure

string AzureBlob = ConfigurationManager.AppSettings["AzureBlob"]; string AzureContainName = ConfigurationManager.AppSettings["AzureContainName"]; public ActionResult AddDoc() { var storageAccount = CloudStorageAccount.Parse(AzureBlob); var azureDir = new AzureDirectory(storageAccount, AzureContainName); datas = InitDoc(); var findexExists = IndexReader.IndexExists(azureDir); IndexWriter indexWriter = null; while (indexWriter == null) { try { indexWriter = new IndexWriter(azureDir, new StandardAnalyzer(Lucene.Net.Util.Version.LUCENE_CURRENT), !IndexReader.IndexExists(azureDir), new Lucene.Net.Index.IndexWriter.MaxFieldLength(IndexWriter.DEFAULT_MAX_FIELD_LENGTH)); } catch (LockObtainFailedException) { azureDir.ClearLock("write.lock"); return new JsonResult { Data = "索引資料庫已解除鎖定，請重新加入文件" }; } }; indexWriter.SetRAMBufferSizeMB(10.0); foreach (var item in datas) { var doc = new Document(); doc.Add(new Field("EdocId", item.EdocId, Field.Store.YES, Field.Index.ANALYZED, Field.TermVector.NO)); doc.Add(new Field("CreateUser", item.CreateUser, Field.Store.YES, Field.Index.ANALYZED, Field.TermVector.NO)); doc.Add(new Field("CreateDate", item.CreateDate, Field.Store.YES, Field.Index.ANALYZED, Field.TermVector.NO)); doc.Add(new Field("Content", item.Content, Field.Store.YES, Field.Index.ANALYZED, Field.TermVector.NO)); indexWriter.AddDocument(doc); } indexWriter.Dispose(); return new JsonResult{ Data="加入文件數量:"+indexWriter.NumDocs().ToString()}; }

成功後可以看到Blob儲存體有Lucene.net Index Data

Search in Azure

public ActionResult SearchResult(string kw)
       {
           var storageAccount = CloudStorageAccount.Parse(AzureBlob);
           var azureDir = new AzureDirectory(storageAccount, AzureContainName);
           IndexSearcher searcher = new IndexSearcher(azureDir);
           Analyzer analyzer = new StandardAnalyzer(Lucene.Net.Util.Version.LUCENE_CURRENT);
           var parser = new QueryParser(Lucene.Net.Util.Version.LUCENE_CURRENT, "Content", analyzer);
           //QueryParser parser = new Lucene.Net.QueryParsers.QueryParser(Lucene.Net.Util.Version.LUCENE_CURRENT, "Body", new StandardAnalyzer(Lucene.Net.Util.Version.LUCENE_CURRENT));       
           Query query = parser.Parse(kw);
           TopDocs topDocs = searcher.Search(query, null,100);
           int totalnum = topDocs.TotalHits;
           int max = topDocs.ScoreDocs.Length;
           SimpleHTMLFormatter formatter = new SimpleHTMLFormatter("<font color=\"red\">", "</font>");
           QueryScorer scorer = new QueryScorer(query, "Content");
           var highlighter = new Highlighter(formatter, scorer);
           highlighter.TextFragmenter = new SimpleFragmenter(500);//500字元
           List<LuceneModels> _LuceneModels = new List<LuceneModels>();
           for (int i = 0; i < max; i++)
           {
               LuceneModels _LuceneModel = new LuceneModels();
               ScoreDoc scoreDoc = topDocs.ScoreDocs[i];
               float score = scoreDoc.Score;
               int docId = scoreDoc.Doc;
               Document doc = searcher.Doc(docId);
               TokenStream stream = TokenSources.GetAnyTokenStream(searcher.IndexReader, docId, "Content", new StandardAnalyzer(Lucene.Net.Util.Version.LUCENE_CURRENT)); 

               string strSnippet = highlighter.GetBestFragment(stream, doc.Get("Content"));
               _LuceneModel.Content = strSnippet;
               _LuceneModel.CreateDate = doc.Get("CreateDate");
               _LuceneModel.CreateUser = doc.Get("CreateUser");
               _LuceneModel.EdocId = doc.Get("EdocId"); 
               _LuceneModels.Add(_LuceneModel);
           } 

           return PartialView("_FullText", _LuceneModels);         
       }