/// <summary>
/// Subclass can override to choose a specific
/// <see cref="Directory"/> implementation.
/// </summary>
protected internal virtual Directory GetDirectory(DirectoryInfo path)
{
return(FSDirectory.Open(path));
}
private void InitializeLocalIndexAndDirectory(Lucene.Net.Store.Directory baseLuceneDirectory, Analyzer analyzer, string configuredPath)
{
lock (Locker)
{
if (!Directory.Exists(_tempPath))
{
Directory.CreateDirectory(_tempPath);
}
else
{
//if we are syncing storage to the main file system to temp files, then clear out whatever is
//currently there since we'll re-copy it over
if (_syncStorage)
{
//clear it!
Directory.Delete(_tempPath, true);
//recreate it
Directory.CreateDirectory(_tempPath);
}
}
//if we are syncing storage to the main file system to temp files, then sync from the main FS to our temp FS
if (_syncStorage)
{
//copy index
using (new IndexWriter(
//read from the underlying/default directory, not the temp codegen dir
baseLuceneDirectory,
analyzer,
Snapshotter,
IndexWriter.MaxFieldLength.UNLIMITED))
{
try
{
var basePath = IOHelper.MapPath(configuredPath);
var commit = Snapshotter.Snapshot();
var fileNames = commit.GetFileNames();
foreach (var fileName in fileNames)
{
File.Copy(
Path.Combine(basePath, "Index", fileName),
Path.Combine(_tempPath, Path.GetFileName(fileName)), true);
}
var segments = commit.GetSegmentsFileName();
if (segments.IsNullOrWhiteSpace() == false)
{
File.Copy(
Path.Combine(basePath, "Index", segments),
Path.Combine(_tempPath, Path.GetFileName(segments)), true);
}
}
finally
{
Snapshotter.Release();
}
}
//create the custom lucene directory which will keep the main and temp FS's in sync
LuceneDirectory = new TempStorageDirectory(
new DirectoryInfo(_tempPath),
baseLuceneDirectory);
}
else
{
//just return a normal lucene directory that uses the codegen folder
LuceneDirectory = FSDirectory.Open(new DirectoryInfo(_tempPath));
}
}
}
public void Integration()
{
var builder = new ContainerBuilder();
builder.RegisterModule(new RootModule(@"Files\Shorthand.xml"));
var container = builder.Build();
// CORRECT DATA AND INITIAL LOAD
using (var cn = new SqlServerConnectionFactory(InputConnection).GetConnection()) {
cn.Open();
Assert.AreEqual(2, cn.Execute(@"
UPDATE [Order Details] SET UnitPrice = 14.40, Quantity = 42 WHERE OrderId = 10253 AND ProductId = 39;
UPDATE Orders SET CustomerID = 'CHOPS', Freight = 22.98 WHERE OrderId = 10254;
"));
}
var root = ResolveRoot(container, TestFile, true);
var response = new PipelineAction(root).Execute();
Assert.AreEqual(200, response.Code);
Assert.AreEqual(string.Empty, response.Content);
using (var reader = IndexReader.Open(FSDirectory.Open(new DirectoryInfo(Path.Combine(OutputConnection.Folder, "Order Details"))), true)) {
Assert.AreEqual(2155, reader.NumDocs());
}
// FIRST DELTA, NO CHANGES
root = ResolveRoot(container, TestFile, false);
response = new PipelineAction(root).Execute();
Assert.AreEqual(200, response.Code);
Assert.AreEqual(string.Empty, response.Content);
using (var reader = IndexReader.Open(FSDirectory.Open(new DirectoryInfo(Path.Combine(OutputConnection.Folder, "Order Details"))), true)) {
Assert.AreEqual(2155, reader.NumDocs());
}
// CHANGE 2 FIELDS IN 1 RECORD IN MASTER TABLE THAT WILL CAUSE CALCULATED FIELD TO BE UPDATED TOO
using (var cn = new SqlServerConnectionFactory(InputConnection).GetConnection()) {
cn.Open();
const string sql = @"UPDATE [Order Details] SET UnitPrice = 15, Quantity = 40 WHERE OrderId = 10253 AND ProductId = 39;";
Assert.AreEqual(1, cn.Execute(sql));
}
root = ResolveRoot(container, TestFile, false);
response = new PipelineAction(root).Execute();
Assert.AreEqual(200, response.Code);
Assert.AreEqual(string.Empty, response.Content);
using (var searcher = new IndexSearcher(FSDirectory.Open(new DirectoryInfo(Path.Combine(OutputConnection.Folder, "Order Details"))), true)) {
var hits = searcher.Search(new TermQuery(new Term("TflId", "1025339")), null, 1);
Assert.AreEqual(1, hits.TotalHits);
var hit = searcher.Doc(hits.ScoreDocs[0].Doc);
Assert.AreEqual(15.0d, Convert.ToDecimal(hit.Get("OrderDetailsUnitPrice")));
Assert.AreEqual(40, Convert.ToInt32(hit.Get("OrderDetailsQuantity")));
Assert.AreEqual(40 * 15.0d, Convert.ToDecimal(hit.Get("OrderDetailsExtendedPrice")));
}
// CHANGE 1 RECORD'S CUSTOMERID AND FREIGHT ON ORDERS TABLE
using (var cn = new SqlServerConnectionFactory(InputConnection).GetConnection()) {
cn.Open();
Assert.AreEqual(1, cn.Execute("UPDATE Orders SET CustomerID = 'VICTE', Freight = 20.11 WHERE OrderId = 10254;"));
}
root = ResolveRoot(container, TestFile, false);
response = new PipelineAction(root).Execute();
Assert.AreEqual(200, response.Code);
Assert.AreEqual(string.Empty, response.Content);
using (var searcher = new IndexSearcher(FSDirectory.Open(new DirectoryInfo(Path.Combine(OutputConnection.Folder, "Orders"))), true)) {
var hits = searcher.Search(new TermQuery(new Term("OrdersOrderID", NumericUtils.IntToPrefixCoded(10254))), 1);
Assert.AreEqual(1, hits.TotalHits);
var hit = searcher.Doc(hits.ScoreDocs[0].Doc);
Assert.AreEqual("VICTE", hit.Get("OrdersCustomerID"));
Assert.AreEqual(20.11d, Convert.ToDecimal(hit.Get("OrdersFreight")));
}
}
/// <summary>
///
/// </summary>
public void IndexCreate()
{
analyzer = new PanGuAnalyzer();
IndexWriter writer = new IndexWriter(FSDirectory.Open(new DirectoryInfo(indexDirectory)), analyzer, true, IndexWriter.MaxFieldLength.LIMITED);
}
/// <summary>
/// Searches the datasource using the specified criteria. Criteria is parsed by the query builder specified by
/// <typeparamref />
/// .
/// </summary>
/// <param name="scope">Name of the application.</param>
/// <param name="criteria">The criteria.</param>
/// <returns></returns>
/// <exception cref="VirtoCommerce.SearchModule.Data.Providers.Lucene.LuceneSearchException"></exception>
public virtual ISearchResults Search(string scope, ISearchCriteria criteria)
{
ISearchResults result;
var directoryInfo = new DirectoryInfo(GetDirectoryPath(GetFolderName(scope, criteria.DocumentType)));
if (directoryInfo.Exists)
{
var dir = FSDirectory.Open(directoryInfo);
var searcher = new IndexSearcher(dir);
var q = (QueryBuilder)QueryBuilder.BuildQuery(criteria);
// filter out empty value
var filter = q.Filter.ToString().Equals("BooleanFilter()") ? null : q.Filter;
Debug.WriteLine("Search Lucene Query:{0}", q.ToString());
TopDocs docs;
try
{
var numDocs = criteria.StartingRecord + criteria.RecordsToRetrieve;
// numDocs must be > 0
if (numDocs < 1)
{
numDocs = 1;
}
if (criteria.Sort != null)
{
var fields = criteria.Sort.GetSort();
docs = searcher.Search(q.Query, filter, numDocs,
new Sort(fields.Select(field => new SortField(field.FieldName, field.DataType, field.IsDescending)).ToArray()));
}
else
{
docs = searcher.Search(q.Query, filter, numDocs);
}
}
catch (Exception ex)
{
throw new LuceneSearchException("Search exception", ex);
}
var results = new LuceneSearchResults(searcher, searcher.IndexReader, docs, criteria, q.Query);
// Cleanup here
searcher.IndexReader.Dispose();
searcher.Dispose();
result = results.Results;
}
else
{
result = new SearchResults(criteria, null);
}
return(result);
}
public static void Show()
{
FSDirectory dir = FSDirectory.Open(StaticConstant.TestIndexPath);
IndexSearcher searcher = new IndexSearcher(dir); //查找器
{
TermQuery query = new TermQuery(new Term("title", "图书馆")); //包含
TopDocs docs = searcher.Search(query, null, 10000); //找到的数据
foreach (ScoreDoc sd in docs.ScoreDocs)
{
Document doc = searcher.Doc(sd.Doc);
Console.WriteLine("***************************************");
Console.WriteLine(string.Format("id={0}", doc.Get("id")));
Console.WriteLine(string.Format("title={0}", doc.Get("title")));
Console.WriteLine(string.Format("time={0}", doc.Get("time")));
Console.WriteLine(string.Format("price={0}", doc.Get("price")));
Console.WriteLine(string.Format("content={0}", doc.Get("content")));
}
Console.WriteLine("1一共命中了{0}个", docs.TotalHits);
}
QueryParser parser = new QueryParser(Lucene.Net.Util.Version.LUCENE_30, "title", new PanGuAnalyzer());//解析器
{
//string keyword = "高中政治人教新课标选修生活中的法律常识";
string keyword = "高中政治 人 教 新课 标 选修 生活 中的 法律常识";
{
Query query = parser.Parse(keyword);
TopDocs docs = searcher.Search(query, null, 10000);//找到的数据
int i = 0;
foreach (ScoreDoc sd in docs.ScoreDocs)
{
if (i++ < 1000)
{
Document doc = searcher.Doc(sd.Doc);
Console.WriteLine("***************************************");
Console.WriteLine(string.Format("id={0}", doc.Get("id")));
Console.WriteLine(string.Format("title={0}", doc.Get("title")));
Console.WriteLine(string.Format("time={0}", doc.Get("time")));
Console.WriteLine(string.Format("price={0}", doc.Get("price")));
}
}
Console.WriteLine($"一共命中{docs.TotalHits}");
}
{
Query query = parser.Parse(keyword);
NumericRangeFilter <int> timeFilter = NumericRangeFilter.NewIntRange("time", 20190101, 20191231, true, true); //过滤
SortField sortPrice = new SortField("price", SortField.DOUBLE, false); //降序
SortField sortTime = new SortField("time", SortField.INT, true); //升序
Sort sort = new Sort(sortTime, sortPrice); //排序 哪个前哪个后
TopDocs docs = searcher.Search(query, timeFilter, 10000, sort); //找到的数据
int i = 0;
foreach (ScoreDoc sd in docs.ScoreDocs)
{
if (i++ < 1000)
{
Document doc = searcher.Doc(sd.Doc);
Console.WriteLine("***************************************");
Console.WriteLine(string.Format("id={0}", doc.Get("id")));
Console.WriteLine(string.Format("title={0}", doc.Get("title")));
Console.WriteLine(string.Format("time={0}", doc.Get("time")));
Console.WriteLine(string.Format("price={0}", doc.Get("price")));
}
}
Console.WriteLine("3一共命中了{0}个", docs.TotalHits);
}
}
}
/// <summary>
///根据关键字进行查询,根据int的范围进行搜索
/// </summary>
/// <param name="Title"></param>
/// <returns></returns>
public List <Bpo_JobEntity> SearchJobList4(string Title, string FullAddress, int MinId, int MaxId)
{
List <Bpo_JobEntity> jobList = new List <Bpo_JobEntity>();
string path = "F://LuceneIndexDir";
FSDirectory dir = FSDirectory.Open(path);
//IndexSearcher searcher = new IndexSearcher(dir);//声明一个查询器，或者以下面一种方式声明
IndexReader reader = IndexReader.Open(dir, true);//查询器
IndexSearcher searcher = new IndexSearcher(reader);
//搜索条件，BooleanQuery可以同时制定多个搜索条件
BooleanQuery booleanQuery = new BooleanQuery();
//Query query = new TermQuery(new Term("Title", $"*{Title}*"));//不支持通配符
//Query query = new WildcardQuery(new Term("Title", $"*{Title}*")); // 通配符
if (!string.IsNullOrEmpty(Title))//空格隔开，按照多个词进行搜索
{
Title = new LuceneAnalyze().AnalyzerKeyword("Title", Title);
QueryParser parser = new QueryParser(Lucene.Net.Util.Version.LUCENE_30, "Title", new PanGuAnalyzer());
Query query = parser.Parse(Title);
booleanQuery.Add(query, Occur.MUST);
}
if (MinId > 0)//空格隔开，按照多个词进行搜索
{
//string idStr = new LuceneAnalyze().AnalyzerKeyword("Id", MinId.ToString());
QueryParser parser = new QueryParser(Lucene.Net.Util.Version.LUCENE_30, "Id", new PanGuAnalyzer());
Query query = parser.Parse(MinId.ToString());
booleanQuery.Add(query, Occur.MUST);
}
if (!string.IsNullOrEmpty(FullAddress))
{
FullAddress = new LuceneAnalyze().AnalyzerKeyword("FullAddress", FullAddress);
QueryParser parser = new QueryParser(Lucene.Net.Util.Version.LUCENE_30, "FullAddress", new PanGuAnalyzer());
Query query = parser.Parse(FullAddress);
booleanQuery.Add(query, Occur.MUST);
//使用WildcardQuery，相当于sql的like
//Query query2 = new WildcardQuery(new Term("FullAddress", $"*{FullAddress}*")); // 通配符
//booleanQuery.Add(query2, Occur.MUST);
}
//根据id区间区间搜索(此时id存储索引时要是NumericField类型)
NumericRangeFilter <int> intFilter = null;
//if (MinId > 0 && MaxId == 0)
//{
// intFilter = NumericRangeFilter.NewIntRange("Id", MinId, int.MaxValue, true, true);
//}
//else if (MaxId > 0 && MinId == 0)
//{
// intFilter = NumericRangeFilter.NewIntRange("Id", 0, MaxId, true, true);
//}
//else if (MaxId > 0 && MinId > 0)
//{
// intFilter = NumericRangeFilter.NewIntRange("Id", MinId, MaxId, true, true);
//}
//定义排序
SortField sortField = new SortField("Id", SortField.STRING, false); //降序
SortField sortField2 = new SortField("CompanyId", SortField.INT, false); //降序
Sort sort = new Sort(sortField, sortField2);
//取搜索结果方法1：
TopDocs docs = searcher.Search(booleanQuery, intFilter, 10000, sort); //找到的结果取100条数据
foreach (ScoreDoc sd in docs.ScoreDocs) //从docs.ScoreDocs取数据
{
Document doc = searcher.Doc(sd.Doc);
jobList.Add(new Bpo_JobEntity
{
Id = Convert.ToInt32(doc.Get("Id") ?? "-100"),
Title = doc.Get("Title"),
UserId = Convert.ToInt32(doc.Get("UserId") ?? "-100"),
UserName = doc.Get("UserName"),
CompanyId = Convert.ToInt32(doc.Get("CompanyId") ?? "-100"),
CompanyName = doc.Get("CompanyName"),
FullAddress = doc.Get("FullAddress"),
CreateDate = Convert.ToDateTime(doc.Get("CreateDate"))
});
}
return(jobList);
}
public NivelDocumentalComProdutoresSearcher(string orderBy, string defaultField)
: base(FSDirectory.Open(Util.NivelDocumentalComProdutoresPath), orderBy, defaultField, new SynonymAnalyzer(new XmlSynonymEngine()))
{
}
private bool Initialize()
{
if (initialized)
{
return(true);
}
if (failedToInitialize)
{
return(false);
}
var fileSystemPath = cmsConfiguration.Search.GetValue(LuceneSearchConstants.ConfigurationKeys.LuceneFileSystemDirectory);
if (string.IsNullOrWhiteSpace(fileSystemPath))
{
Log.Error("Cannot Initialize Lucene indexer. Lucene file system path is not set.");
failedToInitialize = true;
return(false);
}
directory = GetLuceneDirectory(fileSystemPath);
index = FSDirectory.Open(directory);
try
{
configurationExcludedNodeTypes = GetCollectionFromConfiguration(LuceneSearchConstants.ConfigurationKeys.LuceneExcludedNodes);
configurationExcludedIds = GetCollectionFromConfiguration(LuceneSearchConstants.ConfigurationKeys.LuceneExcludedIds);
configurationExcludedClasses = GetCollectionFromConfiguration(LuceneSearchConstants.ConfigurationKeys.LuceneExcludedClasses);
bool.TryParse(cmsConfiguration.Search.GetValue(LuceneSearchConstants.ConfigurationKeys.LuceneSearchForPartOfWordsPrefix), out searchForPartOfWords);
bool disableStopWords;
if (!bool.TryParse(cmsConfiguration.Search.GetValue(LuceneSearchConstants.ConfigurationKeys.LuceneDisableStopWords), out disableStopWords))
{
disableStopWords = false;
}
if (disableStopWords)
{
analyzer = new StandardAnalyzer(Version.LUCENE_30, new HashSet <string>());
}
else
{
analyzer = new StandardAnalyzer(Version.LUCENE_30);
}
if (searchForPartOfWords)
{
parser = new PartialWordTermQueryParser(Version.LUCENE_30, "content", analyzer);
}
else
{
parser = new QueryParser(Version.LUCENE_30, "content", analyzer);
}
parser.AllowLeadingWildcard = true;
if (!IndexReader.IndexExists(index))
{
OpenWriter(true);
CloseWriter();
}
}
catch (Exception exc)
{
Log.Error("Failed to initialize Lucene search engine.", exc);
failedToInitialize = true;
return(false);
}
initialized = true;
return(true);
}
public LuceneEngine()
{
_Analyzer = new StandardAnalyzer(Lucene.Net.Util.Version.LUCENE_30);
_Directory = FSDirectory.Open(_IndexPath);
}
public static bool MakeIndex(List <ArticleIndex> articleList)
{
bool retult = false;
var indexDir = Path.Combine(System.IO.Directory.GetCurrentDirectory(), "lucene");
if (System.IO.Directory.Exists(indexDir) == false)
{
System.IO.Directory.CreateDirectory(indexDir);
}
var VERSION = Lucene.Net.Util.LuceneVersion.LUCENE_48;
var director = FSDirectory.Open(new DirectoryInfo(indexDir));
var analyzer = new JieBaAnalyzer(TokenizerMode.Search);
var indexWriterConfig = new IndexWriterConfig(VERSION, analyzer);
var indexWriter = new IndexWriter(director, indexWriterConfig);
try
{
//Console.WriteLine($"[{DateTime.Now}] UpdateMerchIndex job begin...");
if (File.Exists(Path.Combine(indexDir, "segments.gen")) == true)
{
indexWriter.DeleteAll();
}
var index = 1;
var size = 200;
var count = articleList.Count();
if (count > 0)
{
while (true)
{
var rs = articleList.OrderBy(t => t.CreateTime)
.Skip((index - 1) * size)
.Take(size).ToList();
if (rs.Count == 0)
{
break;
}
var docList = new List <Document>();
foreach (var item in rs)
{
var articleId = item.Id.ToString();
var doc = new Document();
var field1 = new StringField("articleId", articleId, Field.Store.YES);
var field2 = new TextField("title", item.Title?.ToLower(), Field.Store.YES);
var field3 = new TextField("summary", item.Summary?.ToLower(), Field.Store.YES);
doc.Add(field1);
doc.Add(field2);
doc.Add(field3);
docList.Add(doc);
}
if (docList.Count > 0)
{
indexWriter.AddDocuments(docList);
}
index = index + 1;
}
}
if (!indexWriter.IsClosed)
{
indexWriter.Dispose();
}
retult = true;
//Console.WriteLine($"[{DateTime.Now}] UpdateMerchIndex job end!");
}
catch (Exception ex)
{
if (!indexWriter.IsClosed)
{
indexWriter.Dispose();
}
LogHelper.WriteLog_LocalTxt(ex.ToString());
}
return(retult);
}
public static List <int> Search(string key, int page, int pageSize, out int totalHits)
{
totalHits = 0;
if (string.IsNullOrEmpty(key))
{
return(null);
}
key = key.Trim().ToLower();
var rs = new List <int>();
try
{
var indexDir = Path.Combine(System.IO.Directory.GetCurrentDirectory(), "lucene");
//var VERSION = Lucene.Net.Util.LuceneVersion.LUCENE_48;
if (System.IO.Directory.Exists(indexDir) == true)
{
var reader = DirectoryReader.Open(FSDirectory.Open(new DirectoryInfo(indexDir)));
var search = new IndexSearcher(reader);
var directory = FSDirectory.Open(new DirectoryInfo(indexDir), NoLockFactory.GetNoLockFactory());
var reader2 = DirectoryReader.Open(directory);
var search2 = new IndexSearcher(reader2);
//var parser = new QueryParser(VERSION, "title", new JieBaAnalyzer(TokenizerMode.Search));
var booleanQuery = new BooleanQuery();
var list = CutKeyWord(key);
foreach (var word in list)
{
var query1 = new TermQuery(new Term("title", word));
var query2 = new TermQuery(new Term("summary", word));
booleanQuery.Add(query1, Occur.SHOULD);
booleanQuery.Add(query2, Occur.SHOULD);
}
var collector = TopScoreDocCollector.Create(1000, true);
search2.Search(booleanQuery, null, collector);
var docs = collector.GetTopDocs((page - 1) * pageSize, (page) * pageSize).ScoreDocs;
totalHits = collector.TotalHits;
foreach (var d in docs)
{
var num = d.Doc;
var document = search.Doc(num);// 拿到指定的文档
var articleId = document.Get("articleId");
//var name = document.Get("title");
if (int.TryParse(articleId, out int mid) == true)
{
rs.Add(mid);
}
}
}
}
catch (Exception ex)
{
Console.WriteLine($"SearchMerchs ex={ex}");
}
return(rs);
}
/// <summary>
/// 提交内容
/// </summary>
/// <param name="id"></param>
/// <param name="title"></param>
/// <param name="content"></param>
/// <param name="date"></param>
/// <param name="param"></param>
/// <param name="modcode"></param>
/// <returns></returns>
public static string CommitContent(string id, string title, string content, DateTime?date, string param, string modcode, string searchparam1, string searchparam2, string searchparam3)
{
#region 验证输入参数是否合格
if (string.IsNullOrEmpty(id))
{
throw new Exception("参数id不能为空");
}
if (string.IsNullOrEmpty(title))
{
throw new Exception("参数title不能为空");
}
if (string.IsNullOrEmpty(content))
{
throw new Exception("参数content和filepathList不能同时为空");
}
#endregion
if (date == null)
{
date = DateTime.Now; //日期为null时使用当前时间
}
//Analyzer analyzer = new StandardAnalyzer(Lucene.Net.Util.Version.LUCENE_29);
Analyzer analyzer = new PanGuAnalyzer();
FSDirectory directory = FSDirectory.Open(new DirectoryInfo(indexPath), new NativeFSLockFactory());
bool isUpdate = IndexReader.IndexExists(directory);
if (isUpdate)
{
//如果索引目录被锁定（比如索引过程中程序异常退出），则首先解锁
if (IndexWriter.IsLocked(directory))
{
IndexWriter.Unlock(directory);
}
}
IndexWriter writer = new IndexWriter(directory, analyzer, !isUpdate, IndexWriter.MaxFieldLength.UNLIMITED);
//为避免重复索引，所以先删除id=id的记录，再重新添加
writer.DeleteDocuments(new Term("id", id));
string result = "";
try
{
Document doc = new Document();
doc.Add(new Field("id", id, Field.Store.YES, Field.Index.NOT_ANALYZED)); //存储
doc.Add(new Field("title", title, Field.Store.YES, Field.Index.ANALYZED, Field.TermVector.WITH_POSITIONS_OFFSETS)); //分词建立索引
doc.Add(new Field("content", content, Field.Store.YES, Field.Index.ANALYZED, Field.TermVector.WITH_POSITIONS_OFFSETS)); //分词建立索引
doc.Add(new Field("date", date.ToString(), Field.Store.YES, Field.Index.NOT_ANALYZED)); //不分词建立索引
doc.Add(new Field("param", param, Field.Store.YES, Field.Index.NO)); //存储
doc.Add(new Field("modcode", modcode, Field.Store.YES, Field.Index.NOT_ANALYZED)); //存储
doc.Add(new Field("searchparam1", searchparam1, Field.Store.YES, Field.Index.NOT_ANALYZED)); //存储
doc.Add(new Field("searchparam2", searchparam2, Field.Store.YES, Field.Index.NOT_ANALYZED)); //存储
doc.Add(new Field("searchparam3", searchparam3, Field.Store.YES, Field.Index.NOT_ANALYZED)); //存储
writer.AddDocument(doc);
//输出api调用记录
string EnableApiLog = Util.GetAppSetting("EnableApiLog");
if (EnableApiLog == "1")
{
string logMsg = GetParamList(id, title, content, date, param, modcode) + "\r\n" + Utility.Util.GetClientInfo();
Utility.Util.WriteApiLog(logMsg);
}
}
catch (Exception ex)
{
result = ex.Message;
string errMsg = ex.Message + "\r\n" + GetParamList(id, title, content, date, param, modcode) + "\r\n" + Utility.Util.GetClientInfo();
Utility.Util.WriteLog(errMsg);
}
//对索引文件进行优化
//writer.Optimize();
analyzer.Close();
writer.Dispose();
directory.Dispose();
return(result);
}
/// <summary>
/// 搜索内容
/// </summary>
/// <param name="word">搜索关键字</param>
/// <param name="pagesize">每页显示记录数</param>
/// <param name="pageindex">当前页码</param>
/// <returns></returns>
public static SearchResult SearchContent(string modcode, string word, int pagesize, int pageindex, string searchparam1, string searchparam2, string searchparam3)
{
SearchResult searchResult = new SearchResult();
FSDirectory directory = FSDirectory.Open(new DirectoryInfo(indexPath), new NativeFSLockFactory());
IndexSearcher searcher = new IndexSearcher(directory, true);
var analyzer = new PanGuAnalyzer();
//初始化MultiFieldQueryParser以便同时查询多列
Lucene.Net.QueryParsers.MultiFieldQueryParser parser = new Lucene.Net.QueryParsers.MultiFieldQueryParser(Lucene.Net.Util.Version.LUCENE_29, new string[] { "title", "content" }, analyzer);
Lucene.Net.Search.Query query = parser.Parse(word);//初始化Query
parser.DefaultOperator = QueryParser.AND_OPERATOR;
Lucene.Net.Search.BooleanQuery boolQuery = new Lucene.Net.Search.BooleanQuery();
boolQuery.Add(query, Occur.MUST);
if (!string.IsNullOrEmpty(modcode))
{
PhraseQuery queryModCode = new PhraseQuery();
queryModCode.Add(new Term("modcode", modcode));
boolQuery.Add(queryModCode, Occur.MUST);
}
if (!string.IsNullOrEmpty(searchparam1))
{
WildcardQuery query1 = new WildcardQuery(new Term("searchparam1", "*" + searchparam1 + "*"));
boolQuery.Add(query1, Occur.MUST);
}
if (!string.IsNullOrEmpty(searchparam2))
{
WildcardQuery query1 = new WildcardQuery(new Term("searchparam2", "*" + searchparam2 + "*"));
boolQuery.Add(query1, Occur.MUST);
}
if (!string.IsNullOrEmpty(searchparam3))
{
WildcardQuery query1 = new WildcardQuery(new Term("searchparam3", "*" + searchparam3 + "*"));
boolQuery.Add(query1, Occur.MUST);
}
Sort sort = new Sort(new SortField("date", SortField.STRING, true));
var result = searcher.Search(boolQuery, null, 1000, sort);
if (result.TotalHits == 0)
{
searchResult.count = 0;
}
else
{
searchResult.count = result.TotalHits;
int startNum = 0, endNum = result.TotalHits;
if (pagesize > 0)
{
//当pagesize>0时使用分页功能
startNum = (pageindex - 1) * pagesize;
endNum = startNum + pagesize;
}
ScoreDoc[] docs = result.ScoreDocs;
List <JObject> dataList = new List <JObject>();
for (int i = 0; i < docs.Length; i++)
{
if (i < startNum)
{
continue;
}
if (i >= endNum)
{
break;
}
Document doc = searcher.Doc(docs[i].Doc);
string id = doc.Get("id").ToString();
string title = doc.Get("title").ToString();
string content = doc.Get("content").ToString();
string date = doc.Get("date").ToString();
string param = doc.Get("param").ToString();
string mcode = doc.Get("modcode").ToString();
string param1 = doc.Get("searchparam1").ToString();
string param2 = doc.Get("searchparam2").ToString();
string param3 = doc.Get("searchparam3").ToString();
JObject obj = new JObject();
obj["id"] = id;
//创建HTMLFormatter,参数为高亮单词的前后缀
string highLightTag = Util.GetAppSetting("HighLightTag", "<font color=\"red\">|</font>");
string[] tarArr = highLightTag.Split('|');
var simpleHTMLFormatter = new SimpleHTMLFormatter(tarArr[0], tarArr[1]);
//创建 Highlighter ，输入HTMLFormatter 和 盘古分词对象Semgent
var highlighter = new Highlighter(simpleHTMLFormatter, new PanGu.Segment());
//设置每个摘要段的字符数
int highlightFragmentSize = Util.GetAppSetting("HighlightFragmentSize", "100").ToInt();
highlighter.FragmentSize = highlightFragmentSize;
//获取最匹配的摘要段
String bodyPreview = highlighter.GetBestFragment(word, content);
string newTitle = highlighter.GetBestFragment(word, title);
if (!string.IsNullOrEmpty(newTitle))
{
title = newTitle;
}
obj["title"] = title;
obj["content"] = bodyPreview;
obj["date"] = date;
obj["param"] = param;
obj["modcode"] = mcode;
obj["searchparam1"] = param1;
obj["searchparam2"] = param2;
obj["searchparam3"] = param3;
dataList.Add(obj);
}
searchResult.data = dataList;
}
analyzer.Close();
searcher.Dispose();
directory.Dispose();
return(searchResult);
}
public override void Start()
{
string refreshPeriod = properties.ContainsKey("refresh") ? properties["refresh"] : "3600";
long period;
if (!long.TryParse(refreshPeriod, out period))
{
period = 3600;
log.Warn("Error parsing refresh period, defaulting to 1 hour");
}
log.DebugFormat("Refresh period {0} seconds", period);
period *= 1000; // per second
try
{
// Initialize
this.directory1 = InitializeIndex(new DirectoryInfo(Path.Combine(indexName, "1")));
this.directory2 = InitializeIndex(new DirectoryInfo(Path.Combine(indexName, "2")));
string current1Marker = Path.Combine(indexName, "current1");
string current2Marker = Path.Combine(indexName, "current2");
if (File.Exists(current1Marker))
{
current = 1;
}
else if (File.Exists(current2Marker))
{
current = 2;
}
else
{
// no default
log.Debug("Setting directory 1 as current");
current = 1;
DirectoryInfo srcDir = new DirectoryInfo(source);
DirectoryInfo destDir = new DirectoryInfo(Path.Combine(indexName, current.ToString()));
int sourceCurrent = -1;
if (File.Exists(Path.Combine(srcDir.Name, "current1")))
{
sourceCurrent = 1;
}
else if (File.Exists(Path.Combine(srcDir.Name, "current2")))
{
sourceCurrent = 2;
}
if (sourceCurrent != -1)
{
try
{
FileHelper.Synchronize(
new DirectoryInfo(Path.Combine(source, sourceCurrent.ToString())), destDir, true);
}
catch (IOException e)
{
throw new HibernateException("Umable to synchonize directory: " + indexName, e);
}
}
try
{
File.Create(current1Marker).Dispose();
}
catch (IOException e)
{
throw new HibernateException("Unable to create the directory marker file: " + indexName, e);
}
}
log.Debug("Current directory: " + current);
}
catch (IOException e)
{
throw new HibernateException("Unable to initialize index: " + directoryProviderName, e);
}
task = new TriggerTask(this, source, indexName);
timer = new Timer(task.Run, null, period, period);
}
/// <summary>
/// 分页获取商品信息数据
/// </summary>
/// <param name="queryString"></param>
/// <param name="pageIndex">第一页为1</param>
/// <param name="pageSize"></param>
/// <param name="totalCount"></param>
/// <returns></returns>
public List <Commodity> QueryIndexPage(string queryString, int pageIndex, int pageSize, out int totalCount, string priceFilter, string priceOrderBy)
{
totalCount = 0;
IndexSearcher searcher = null;
try
{
List <Commodity> ciList = new List <Commodity>();
FSDirectory dir = FSDirectory.Open(StaticConstant.IndexPath);
searcher = new IndexSearcher(dir);
Analyzer analyzer = new PanGuAnalyzer();
//--------------------------------------这里配置搜索条件
QueryParser parser = new QueryParser(Lucene.Net.Util.Version.LUCENE_30, "title", analyzer);
Query query = parser.Parse(queryString);
pageIndex = Math.Max(1, pageIndex);//索引从1开始
int startIndex = (pageIndex - 1) * pageSize;
int endIndex = pageIndex * pageSize;
NumericRangeFilter <float> numPriceFilter = null;
if (!string.IsNullOrWhiteSpace(priceFilter))
{
bool isContainStart = priceFilter.StartsWith("[");
bool isContainEnd = priceFilter.EndsWith("]");
string[] floatArray = priceFilter.Replace("[", "").Replace("]", "").Replace("{", "").Replace("}", "").Split(',');
float start = 0;
float end = 0;
if (!float.TryParse(floatArray[0], out start) || !float.TryParse(floatArray[1], out end))
{
throw new Exception("Wrong priceFilter");
}
numPriceFilter = NumericRangeFilter.NewFloatRange("price", start, end, isContainStart, isContainEnd);
}
Sort sort = new Sort();
if (!string.IsNullOrWhiteSpace(priceOrderBy))
{
SortField sortField = new SortField("price", SortField.FLOAT, priceOrderBy.EndsWith("asc", StringComparison.CurrentCultureIgnoreCase));
sort.SetSort(sortField);
}
TopDocs docs = searcher.Search(query, numPriceFilter, 10000, sort);
//TopDocs docs = searcher.Search(query, null, 10000);
totalCount = docs.TotalHits;
//PrintScores(docs, startIndex, endIndex, searcher);
for (int i = startIndex; i < endIndex && i < totalCount; i++)
{
Document doc = searcher.Doc(docs.ScoreDocs[i].Doc);
ciList.Add(DocumentToCommodityInfo(doc));
}
return(ciList);
}
finally
{
if (searcher != null)
{
searcher.Dispose();
}
}
}
/// <summary>
/// Searches Lucene.
/// </summary>
/// <param name="luceneVersion">The lucene version.</param>
/// <param name="fsDirectory">The fs directory.</param>
/// <param name="hitLimit">The hit limit.</param>
/// <param name="input">The input.</param>
/// <returns></returns>
public virtual SearchResult <TSearchData> Search(Version luceneVersion, FSDirectory fsDirectory, int hitLimit, string input)
{
return(string.IsNullOrEmpty(input) ? new SearchResult <TSearchData>() : this.PerformSearch(luceneVersion, fsDirectory, hitLimit, input));
}
private static void CreateIndexContent()
{
using (FSDirectory directory = FSDirectory.Open(new DirectoryInfo(m_directoryPath), new NativeFSLockFactory())) //指定索引文件(打开索引目录) FS指的是就是FileSystem
{
bool isUpdate = IndexReader.IndexExists(directory); //IndexReader:对索引进行读取的类。该语句的作用：判断索引库文件夹是否存在以及索引特征文件是否存在。
if (isUpdate)
{
//同时只能有一段代码对索引库进行写操作。当使用IndexWriter打开directory时会自动对索引库文件上锁。
//如果索引目录被锁定（比如索引过程中程序异常退出），则首先解锁（提示一下：如果我现在正在写着已经加锁了，但是还没有写完，这时候又来一个请求，那么不就解锁了吗？这个问题后面会解决）
if (IndexWriter.IsLocked(directory))
{
IndexWriter.Unlock(directory);
}
}
using (IndexWriter writer = new IndexWriter(directory, new PanGuAnalyzer(), !isUpdate, IndexWriter.MaxFieldLength.UNLIMITED)) //向索引库中写索引。这时在这里加锁。
{
//如果队列中有数据，获取队列中的数据写到Lucene.Net中。
while (Queue.Count > 0)
{
KeyValuePair <T, LuceneTypeEnum> keyValuePair = Queue.Dequeue();
T model = keyValuePair.Key;
Type type = model.GetType();
PropertyInfo propertyInfo = type.GetProperty(m_luceneDataModels[0].PropertyName);
writer.DeleteDocuments(new Term(m_luceneDataModels[0].FieldName, propertyInfo.GetValue(model).ObjectToString())); //删除
if (keyValuePair.Value == LuceneTypeEnum.Delete)
{
continue;
}
//表示一篇文档。
Document document = new Document();
//Field.Store.YES:表示是否存储原值。只有当Field.Store.YES在后面才能用doc.Get("number")取出值来.Field.Index. NOT_ANALYZED:不进行分词保存
//Field.Index. ANALYZED:进行分词保存:也就是要进行全文的字段要设置分词 保存（因为要进行模糊查询
//Lucene.Net.Documents.Field.TermVector.WITH_POSITIONS_OFFSETS:不仅保存分词还保存分词的距离。
foreach (FieldDataModel item in m_luceneDataModels)
{
propertyInfo = type.GetProperty(item.PropertyName);
var propertyValue = propertyInfo.GetValue(model);
if (propertyValue != null)
{
string valueString = propertyValue.ToString();
IFieldable fieldable = null;
if (item.FieldType == TypeCode.String)
{
fieldable = new Field(item.FieldName, propertyInfo.GetValue(model).ObjectToString(), item.Store, item.Index, item.TermVector);
}
else
{
NumericField numericField = new NumericField(item.FieldName, item.Store, item.Index == Field.Index.ANALYZED_NO_NORMS);
switch (item.FieldType)
{
case TypeCode.Double:
numericField.SetDoubleValue(Convert.ToDouble(valueString));
break;
case TypeCode.Single:
numericField.SetFloatValue(Convert.ToSingle(valueString));
break;
case TypeCode.Int32:
numericField.SetIntValue(Convert.ToInt32(valueString));
break;
case TypeCode.Int64:
numericField.SetLongValue(Convert.ToInt64(valueString));
break;
default:
break;
}
fieldable = numericField;
}
document.Add(fieldable);
}
}
writer.AddDocument(document);
}
} //会自动解锁。
}
}
/// <summary>
/// Performs a simple search.
/// </summary>
/// <param name="luceneVersion">The lucene version.</param>
/// <param name="fsDirectory">The fs directory.</param>
/// <param name="hitLimit">The hit limit.</param>
/// <param name="input">The input.</param>
/// <returns></returns>
public virtual SearchResult <TSearchData> SimpleSearch(Version luceneVersion, FSDirectory fsDirectory, int hitLimit, string input)
{
if (string.IsNullOrEmpty(input))
{
return(new SearchResult <TSearchData>());
}
var terms = input.Trim().Replace("-", " ").Split(' ').Where(x => !string.IsNullOrEmpty(x)).Select(x => x.Trim() + "*");
input = string.Join(" ", terms);
return(this.PerformSearch(luceneVersion, fsDirectory, hitLimit, input));
}
private void BindSearchResult(string keyword, int startIndex, int pageSize, out int totalCount)
{
string indexPath = Context.Server.MapPath("~/Index"); // 索引文档保存位置
FSDirectory directory = FSDirectory.Open(new DirectoryInfo(indexPath), new NoLockFactory());
IndexReader reader = IndexReader.Open(directory, true);
IndexSearcher searcher = new IndexSearcher(reader);
#region v1.0 单条件查询
//// 查询条件
//PhraseQuery query = new PhraseQuery();
//// 分词后加入查询
//IEnumerable<string> keyList = SplitHelper.SplitWords(keyword);
//foreach (var key in keyList)
//{
// query.Add(new Term("msg", key));
//}
//// 两个词的距离大于100（经验值）就不放入搜索结果，因为距离太远相关度就不高了
//query.SetSlop(100);
#endregion
#region v2.0 多条件查询
IEnumerable <string> keyList = SplitHelper.SplitWords(keyword);
PhraseQuery queryTitle = new PhraseQuery();
foreach (var key in keyList)
{
queryTitle.Add(new Term("title", key));
}
queryTitle.SetSlop(100);
PhraseQuery queryMsg = new PhraseQuery();
foreach (var key in keyList)
{
queryMsg.Add(new Term("msg", key));
}
queryMsg.SetSlop(100);
BooleanQuery query = new BooleanQuery();
query.Add(queryTitle, BooleanClause.Occur.SHOULD); // SHOULD => 可以有，但不是必须的
query.Add(queryTitle, BooleanClause.Occur.SHOULD); // SHOULD => 可以有，但不是必须的
#endregion
// TopScoreDocCollector:盛放查询结果的容器
TopScoreDocCollector collector = TopScoreDocCollector.create(1000, true);
// 使用query这个查询条件进行搜索，搜索结果放入collector
searcher.Search(query, null, collector);
// 首先获取总条数
totalCount = collector.GetTotalHits();
// 从查询结果中取出第m条到第n条的数据
ScoreDoc[] docs = collector.TopDocs(startIndex, pageSize).scoreDocs;
// 遍历查询结果
IList <SearchResult> resultList = new List <SearchResult>();
for (int i = 0; i < docs.Length; i++)
{
// 拿到文档的id，因为Document可能非常占内存（DataSet和DataReader的区别）
int docId = docs[i].doc;
// 所以查询结果中只有id，具体内容需要二次查询
// 根据id查询内容：放进去的是Document，查出来的还是Document
Document doc = searcher.Doc(docId);
SearchResult result = new SearchResult();
result.Url = "ViewArticle.aspx?id=" + doc.Get("id");
result.Title = HighlightHelper.HighLight(keyword, doc.Get("title"));
result.Msg = HighlightHelper.HighLight(keyword, doc.Get("msg")) + "......";
resultList.Add(result);
}
// 绑定到Repeater
rptSearchResult.DataSource = resultList;
rptSearchResult.DataBind();
}
public static void Main(string[] args)
{
if (args.Length < 5)
{
// LUCENENET specific - our wrapper console shows the correct usage
throw new ArgumentException();
//Console.Error.WriteLine("Usage: MultiPassIndexSplitter -out <outputDir> -num <numParts> [-seq] <inputIndex1> [<inputIndex2 ...]");
//Console.Error.WriteLine("\tinputIndex\tpath to input index, multiple values are ok");
//Console.Error.WriteLine("\t-out ouputDir\tpath to output directory to contain partial indexes");
//Console.Error.WriteLine("\t-num numParts\tnumber of parts to produce");
//Console.Error.WriteLine("\t-seq\tsequential docid-range split (default is round-robin)");
//Environment.Exit(-1);
}
List <IndexReader> indexes = new List <IndexReader>();
try
{
string outDir = null;
int numParts = -1;
bool seq = false;
for (int i = 0; i < args.Length; i++)
{
if (args[i].Equals("-out"))
{
outDir = args[++i];
}
else if (args[i].Equals("-num"))
{
numParts = Convert.ToInt32(args[++i]);
}
else if (args[i].Equals("-seq"))
{
seq = true;
}
else
{
DirectoryInfo file = new DirectoryInfo(args[i]);
if (!file.Exists)
{
Console.Error.WriteLine("Invalid input path - skipping: " + file);
continue;
}
using (Store.Directory dir = FSDirectory.Open(new DirectoryInfo(args[i])))
{
try
{
if (!DirectoryReader.IndexExists(dir))
{
Console.Error.WriteLine("Invalid input index - skipping: " + file);
continue;
}
}
catch (Exception)
{
Console.Error.WriteLine("Invalid input index - skipping: " + file);
continue;
}
indexes.Add(DirectoryReader.Open(dir));
}
}
}
if (outDir == null)
{
throw new Exception("Required argument missing: -out outputDir");
}
if (numParts < 2)
{
throw new Exception("Invalid value of required argument: -num numParts");
}
if (indexes.Count == 0)
{
throw new Exception("No input indexes to process");
}
DirectoryInfo @out = new DirectoryInfo(outDir);
@out.Create();
if (!new DirectoryInfo(outDir).Exists)
{
throw new Exception("Can't create output directory: " + @out);
}
Store.Directory[] dirs = new Store.Directory[numParts];
try
{
for (int i = 0; i < numParts; i++)
{
dirs[i] = FSDirectory.Open(new DirectoryInfo(Path.Combine(@out.FullName, "part-" + i)));
}
MultiPassIndexSplitter splitter = new MultiPassIndexSplitter();
IndexReader input;
if (indexes.Count == 1)
{
input = indexes[0];
}
else
{
input = new MultiReader(indexes.ToArray());
}
#pragma warning disable 612, 618
splitter.Split(LuceneVersion.LUCENE_CURRENT, input, dirs, seq);
#pragma warning restore 612, 618
}
finally
{
// LUCENENET specific - properly dispose directories to prevent resource leaks
IOUtils.Dispose(dirs);
}
}
finally
{
// LUCENENET specific - properly dispose index readers to prevent resource leaks
IOUtils.Dispose(indexes);
}
}
// FIXME :default impl
/*public BrowseServiceImpl() : this(new FileInfo(System.getProperty("index.directory")))
* {
* }*/
private BoboIndexReader NewIndexReader()
{
Directory idxDir = FSDirectory.Open(_idxDir);
return(NewIndexReader(idxDir));
}
public IEnumerable <IHit> Query(int pageIndex, int pageSize, out int totalCount, out IEnumerable <FacetGroup> facetedResults)
{
totalCount = 0;
facetedResults = null;
if (searchPaths == null || searchPaths.Count <= 0)
{
searchPaths.AddRange(indexPaths.Values.Select(o => o.Path));
}
List <LuceneHit> results = new List <LuceneHit>();
List <IndexSearcher> subSearchs = new List <IndexSearcher>();
searchPaths.ForEach(o => subSearchs.Add(new IndexSearcher(FSDirectory.Open(o))));
if (facetFields != null && facetFields.Count > 0)
{
var facetGroups = new List <FacetGroup>();
var mainQueryFilter = new CachingWrapperFilter(new QueryWrapperFilter(query));
MultiReader readers = new MultiReader(subSearchs.Select(o => o.IndexReader).ToArray());
foreach (var facetField in facetFields)
{
FacetGroup fg = new FacetGroup();
fg.FieldName = facetFieldNameProvider.GetMapName(TypeName, facetField);
var items = new List <FacetItem>();
var allDistinctField = FieldCache_Fields.DEFAULT.GetStrings(readers, facetField).Distinct().ToArray();
int totalHits = 0;
Parallel.ForEach(allDistinctField, fieldValue =>
{
//foreach (var fieldValue in allDistinctField)
//{
var facetQuery = new TermQuery(new Term(facetField, fieldValue));
var facetQueryFilter = new CachingWrapperFilter(new QueryWrapperFilter(facetQuery));
var bs = new OpenBitSetDISI(facetQueryFilter.GetDocIdSet(readers).Iterator(), readers.MaxDoc);
bs.InPlaceAnd(mainQueryFilter.GetDocIdSet(readers).Iterator());
int count = (Int32)bs.Cardinality();
FacetItem item = new FacetItem();
item.GroupValue = fieldValue;
item.Count = count;
items.Add(item);
totalHits += count;
}
);
fg.FacetItems = items.OrderByDescending(o => o.Count);
fg.TotalHits = totalHits;
facetGroups.Add(fg);
}
facetedResults = facetGroups.OrderBy(o => o.FieldName);
}
ParallelMultiSearcher searcher = new ParallelMultiSearcher(subSearchs.ToArray());
Sort sort = null;
if (sortFields != null && sortFields.Count > 0)
{
sort = new Sort(sortFields.ToArray());
}
int maxDoc = searcher.MaxDoc;
int startIndex = 0;
if (pageIndex >= 0 && pageSize > 0)
{
startIndex = pageIndex * pageSize;
maxDoc = pageSize * (pageIndex + 1);
}
var docs = sort == null?searcher.Search(query, null, maxDoc) : searcher.Search(query, null, maxDoc, sort);
totalCount = docs.TotalHits;
int endIndex = docs.TotalHits - startIndex;
for (int i = startIndex; i < endIndex; i++)
{
LuceneHit h = new LuceneHit(TypeName, DocumentBuilder, searcher.Doc(docs.ScoreDocs[i].Doc));
results.Add(h);
}
return(results);
}
/// <summary>Index all text files under a directory.</summary>
public static void Main(string[] args)
{
// The <CONSOLE_APP_NAME> should be the assembly name of the application
// this code is compiled into. In .NET Framework, it is the name of the EXE file.
// In .NET Core, you have the option of compiling this into either an EXE or a DLL
// (see https://docs.microsoft.com/en-us/dotnet/core/deploying/index).
// In the latter case, the <CONSOLE_APP_NAME> will be "dotnet <DLL_NAME>.dll".
string usage = "Usage: <CONSOLE_APP_NAME> <INDEX_DIRECTORY> <SOURCE_DIRECTORY> "
+ "[-u|--update]\n\n"
+ "This indexes the documents in <SOURCE_DIRECTORY>, creating a Lucene index"
+ "in <INDEX_DIRECTORY> that can be searched with the search-files demo.";
// Validate required arguments are present.
// If not, show usage information.
if (args.Length < 2)
{
Console.WriteLine(usage);
Environment.Exit(1);
}
string indexPath = args[0];
string sourcePath = args[1];
bool create = true;
for (int i = 0; i < args.Length; i++)
{
if ("-u".Equals(args[i], StringComparison.Ordinal) || "--update".Equals(args[i], StringComparison.Ordinal))
{
create = false;
}
}
DirectoryInfo sourceDirectory = new DirectoryInfo(sourcePath);
if (!sourceDirectory.Exists)
{
Console.WriteLine("Source directory '" + sourcePath + "' does not exist, please check the path");
Environment.Exit(1);
}
DateTime start = DateTime.UtcNow;
try
{
Console.WriteLine("Indexing to directory '" + indexPath + "'...");
Store.Directory dir = FSDirectory.Open(indexPath);
// :Post-Release-Update-Version.LUCENE_XY:
Analyzer analyzer = new StandardAnalyzer(LuceneVersion.LUCENE_48);
IndexWriterConfig iwc = new IndexWriterConfig(LuceneVersion.LUCENE_48, analyzer);
if (create)
{
// Create a new index in the directory, removing any
// previously indexed documents:
iwc.OpenMode = OpenMode.CREATE;
}
else
{
// Add new documents to an existing index:
iwc.OpenMode = OpenMode.CREATE_OR_APPEND;
}
// Optional: for better indexing performance, if you
// are indexing many documents, increase the RAM
// buffer.
//
// iwc.RAMBufferSizeMB = 256.0;
using (IndexWriter writer = new IndexWriter(dir, iwc))
{
IndexDocs(writer, sourceDirectory);
// NOTE: if you want to maximize search performance,
// you can optionally call forceMerge here. This can be
// a terribly costly operation, so generally it's only
// worth it when your index is relatively static (ie
// you're done adding documents to it):
//
// writer.ForceMerge(1);
}
DateTime end = DateTime.UtcNow;
Console.WriteLine((end - start).TotalMilliseconds + " total milliseconds");
}
catch (IOException e)
{
Console.WriteLine(" caught a " + e.GetType() +
"\n with message: " + e.Message);
}
}
public SnapshotManager(FSDirectory fsdir)
{
this.fsdir = fsdir;
}
//string path = AppDomain.CurrentDomain.BaseDirectory.Substring(0, AppDomain.CurrentDomain.BaseDirectory.IndexOf("\\bin"));
//TREBA INJECTOVATI PUTANJU U OVAJ STATIC PROPERTY, STAVLJENO DA NE VRACA NISTA DA BI BUILD PROSAO
//private static string LuceneIndexDirectory
//{
// get
// {
// //return "";
// //kako radi
// //return Path.Combine(_env.WebRootPath, "/Lucene");
// return Path.Combine(_env.ContentRootPath, "~/Lucene");
// //kako je bilo
// //return HttpContext.Current.Server.MapPath("~/Lucene");
// }
//}
public List <KbSearchResultItemViewModel> DoSearch(string text, int page = 1, int resultCount = 20)
{
var LuceneIndexDirectory = Path.Combine(_env.ContentRootPath, "Lucene");
try
{
if (page < 1)
{
throw new ArgumentException("Page");
}
if (string.IsNullOrEmpty(text))
{
throw new ArgumentNullException("Search Text");
}
var results = new List <KbSearchResultItemViewModel>();
var analyzer = new StandardAnalyzer(Lucene.Net.Util.Version.LUCENE_30);
var searcher = new IndexSearcher(FSDirectory.Open(LuceneIndexDirectory));
var fields = new string[] { "Title", "Content" };
var parser = new MultiFieldQueryParser(Lucene.Net.Util.Version.LUCENE_30, fields, analyzer);
var q = ParseQuery(text, parser);
var hits = searcher.Search(q, page * resultCount);
if (hits.ScoreDocs.Any())
{
for (int i = (page - 1) * resultCount; i < hits.ScoreDocs.Count(); i++)
{
var doc = searcher.Doc(hits.ScoreDocs[i].Doc);
var item = new KbSearchResultItemViewModel
{
ArticleId = Convert.ToInt32(doc.Get("Id").ToString().Replace("KB-", string.Empty).Replace("AT-", string.Empty)),
IsArticle = doc.Get("Id").StartsWith("KB-"),
IsAttachment = doc.Get("Id").StartsWith("AT-"),
ArticleTitle = doc.Get("Title")
};
//2808
var article = _context.Articles.FirstOrDefault(a => a.Id == item.ArticleId);
if (article != null)
{
item.ArticleSefName = article.SefName;
}
else
{
item.ArticleSefName = null;
}
results.Add(item);
}
}
searcher.Dispose();
return(results);
}
catch (Exception ex)
{
LogManager.GetCurrentClassLogger().Error(ex);
throw;
}
}
private void btnFolder_Click(object sender, EventArgs e)
{
FolderBrowserDialog dia = new FolderBrowserDialog();
DialogResult res = dia.ShowDialog();
if (res != System.Windows.Forms.DialogResult.OK)
{
return;
}
FSDirectory dir = FSDirectory.GetDirectory(Environment.CurrentDirectory + "\\LuceneIndex");
//Lucene.Net.Store.RAMDirectory dir = new RAMDirectory();
Lucene.Net.Analysis.StopAnalyzer an = new Lucene.Net.Analysis.StopAnalyzer();
IndexWriter wr = new IndexWriter(dir, an, true);
DirectoryInfo diMain = new DirectoryInfo(dia.SelectedPath);
foreach (FileInfo fi in diMain.GetFiles())
{
Document doc = new Document();
doc.Add(new Field("title", fi.Name, Field.Store.YES, Field.Index.NO));
doc.Add(new Field("text", File.ReadAllText(fi.FullName), Field.Store.YES, Field.Index.TOKENIZED, Field.TermVector.YES));
wr.AddDocument(doc);
}
wr.Optimize();
wr.Flush();
wr.Close();
dir.Close();
IndexReader reader = IndexReader.Open(dir);
for (int i = 0; i < reader.MaxDoc(); i++)
{
if (reader.IsDeleted(i))
{
continue;
}
Document doc = reader.Document(i);
String docId = doc.Get("docId");
foreach (TermFreqVector vector in reader.GetTermFreqVectors(i))
{
foreach (string term in vector.GetTerms())
{
Console.WriteLine(term);
}
}
// do something with docId here...
}
//IndexSearcher search = new IndexSearcher(wr.GetReader());
//MoreLikeThis mlt = new MoreLikeThis(wr.GetReader());
//FileInfo fitarget = new FileInfo(@"C:\Users\peacemaker\Desktop\TestNoBitcoin\test.txt");
//Query query = mlt.Like(fitarget);
//var hits = search.Search(query, int.MaxValue);
//foreach (ScoreDoc doc in hits.ScoreDocs)
//{
// textBox1.Text += doc.Score + Environment.NewLine;
//}
}
public TipologiasUpdater()
: base(FSDirectory.Open(Util.TipologiasPath))
{
}
/// <summary>
/// 更新索引库操作
/// </summary>
private void CRUDIndex()
{
FSDirectory directory = FSDirectory.Open(new DirectoryInfo(indexPath), new NativeFSLockFactory());
bool isExist = IndexReader.IndexExists(directory);
if (isExist)
{
if (IndexWriter.IsLocked(directory))
{
IndexWriter.Unlock(directory);
}
}
IndexWriter writer = new IndexWriter(directory, new PanGuAnalyzer(), !isExist, IndexWriter.MaxFieldLength.UNLIMITED);
if (bookQueue.Count > 0)
{
for (int i = 0; i < 5; i++)
{
Document document = new Document();
if (bookQueue.Count <= 0)
{
break;
}
MovieViewModel book = bookQueue.Dequeue();
if (book.IT == IndexType.Delete)
{
writer.DeleteDocuments(new Term("id", book.MovieID.ToString()));
}
else if (book.IT == IndexType.Modify || book.IT == IndexType.Insert)
{
//先删除 再新增
writer.DeleteDocuments(new Term("id", book.MovieID.ToString()));
document.Add(new Field("id", book.MovieID.ToString(), Field.Store.YES, Field.Index.NOT_ANALYZED));
document.Add(new Field("title", book.MovieTitle, Field.Store.YES, Field.Index.ANALYZED,
Field.TermVector.WITH_POSITIONS_OFFSETS));
document.Add(new Field("content", book.MovieContent, Field.Store.YES, Field.Index.ANALYZED,
Field.TermVector.WITH_POSITIONS_OFFSETS));
document.Add(new Field("stars", book.Stars, Field.Store.YES, Field.Index.NO,
Field.TermVector.NO));
document.Add(new Field("director", book.Director ?? "", Field.Store.YES, Field.Index.ANALYZED,
Field.TermVector.WITH_POSITIONS_OFFSETS));
document.Add(new Field("picfile", book.PictureFile, Field.Store.YES, Field.Index.NO,
Field.TermVector.NO));
document.Add(new Field("grade", book.Grade.ToString(), Field.Store.YES, Field.Index.NO,
Field.TermVector.NO));
writer.AddDocument(document);
}
}
}
writer.Close();
directory.Close();
}
public LuceneService(Console console)
{
luceneIndexDirectory = FSDirectory.Open(CONFIG.LUCENE_INDEX_PATH);
this.console = console;
}
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public IndexSplitter(java.io.File dir) throws java.io.IOException
public IndexSplitter(File dir)
{
this.dir = dir;
fsDir = FSDirectory.open(dir);
infos = new SegmentInfos();
infos.read(fsDir);
}