public virtual void TestMulipleTerms()
{
query.SetSlop(2);
query.Add(new Term("field", "one"));
query.Add(new Term("field", "three"));
query.Add(new Term("field", "five"));
ScoreDoc[] hits = searcher.Search(query, null, 1000).scoreDocs;
Assert.AreEqual(1, hits.Length, "two total moves");
QueryUtils.Check(query, searcher);
query = new PhraseQuery();
query.SetSlop(5); // it takes six moves to match this phrase
query.Add(new Term("field", "five"));
query.Add(new Term("field", "three"));
query.Add(new Term("field", "one"));
hits = searcher.Search(query, null, 1000).scoreDocs;
Assert.AreEqual(0, hits.Length, "slop of 5 not close enough");
QueryUtils.Check(query, searcher);
query.SetSlop(6);
hits = searcher.Search(query, null, 1000).scoreDocs;
Assert.AreEqual(1, hits.Length, "slop of 6 just right");
QueryUtils.Check(query, searcher);
}
public virtual void TestNotCloseEnough()
{
query.SetSlop(2);
query.Add(new Term("Field", "one"));
query.Add(new Term("Field", "five"));
Hits hits = searcher.Search(query);
Assert.AreEqual(0, hits.Length());
}
public void ProcessRequest(HttpContext context)
{
context.Response.ContentType = "text/plain";
string searchKey = context.Request["wd"];
string indexPath = context.Server.MapPath("../IndexData");
FSDirectory directory = FSDirectory.Open(new DirectoryInfo(indexPath), new NoLockFactory());
IndexReader reader = IndexReader.Open(directory, true);
IndexSearcher searcher = new IndexSearcher(reader);
//搜索条件
PhraseQuery query = new PhraseQuery();
//把用户输入的关键字进行分词
foreach (string word in Picture.Utility.SplitContent.SplitWords(searchKey))
{
query.Add(new Term("tag", word));
}
//query.Add(new Term("content", "C#"));//多个查询条件时 为且的关系
query.SetSlop(100); //指定关键词相隔最大距离
//TopScoreDocCollector盛放查询结果的容器
TopScoreDocCollector collector = TopScoreDocCollector.create(1000, true);
searcher.Search(query, null, collector);//根据query查询条件进行查询，查询结果放入collector容器
//TopDocs 指定0到GetTotalHits() 即所有查询结果中的文档 如果TopDocs(20,10)则意味着获取第20-30之间文档内容 达到分页的效果
ScoreDoc[] docs = collector.TopDocs(0,10).scoreDocs;
//展示数据实体对象集合
var tagModels = new List<Picture.Model.TagModel>();
for (int i = 0; i < docs.Length; i++)
{
int docId = docs[i].doc;//得到查询结果文档的id（Lucene内部分配的id）
Document doc = searcher.Doc(docId);//根据文档id来获得文档对象Document
Picture.Model.TagModel tag = new Picture.Model.TagModel();
//picture.ImgSummary = doc.Get("summary");
tag.TagName= Picture.Utility.SplitContent.HightLight(searchKey, doc.Get("tag"));
//book.ContentDescription = doc.Get("content");//未使用高亮
//搜索关键字高亮显示 使用盘古提供高亮插件
//book.ContentDescription = Picture.Utility.SplitContent.HightLight(Request.QueryString["SearchKey"], doc.Get("content"));
tag.TId = Convert.ToInt32(doc.Get("id"));
tagModels.Add(tag);
}
SearchPreviewResult result = new SearchPreviewResult()
{
q=searchKey,
p=false
};
foreach (var item in tagModels)
{
result.s.Add(item.TagName);
}
System.Web.Script.Serialization.JavaScriptSerializer jss = new System.Web.Script.Serialization.JavaScriptSerializer();
context.Response.Write(jss.Serialize(result));
}
public virtual void TestPalyndrome3()
{
// search on non palyndrome, find phrase with no slop, using exact phrase scorer
query.SetSlop(0); // to use exact phrase scorer
query.Add(new Term("field", "one"));
query.Add(new Term("field", "two"));
query.Add(new Term("field", "three"));
ScoreDoc[] hits = searcher.Search(query, null, 1000).scoreDocs;
Assert.AreEqual(1, hits.Length, "phrase found with exact phrase scorer");
float score0 = hits[0].score;
//System.out.println("(exact) field: one two three: "+score0);
QueryUtils.Check(query, searcher);
// search on non palyndrome, find phrase with slop 3, though no slop required here.
query.SetSlop(4); // to use sloppy scorer
hits = searcher.Search(query, null, 1000).scoreDocs;
Assert.AreEqual(1, hits.Length, "just sloppy enough");
float score1 = hits[0].score;
//System.out.println("(sloppy) field: one two three: "+score1);
Assert.AreEqual(score0, score1, SCORE_COMP_THRESH, "exact scorer and sloppy scorer score the same when slop does not matter");
QueryUtils.Check(query, searcher);
// search ordered in palyndrome, find it twice
query = new PhraseQuery();
query.SetSlop(4); // must be at least four for both ordered and reversed to match
query.Add(new Term("palindrome", "one"));
query.Add(new Term("palindrome", "two"));
query.Add(new Term("palindrome", "three"));
hits = searcher.Search(query, null, 1000).scoreDocs;
Assert.AreEqual(1, hits.Length, "just sloppy enough");
float score2 = hits[0].score;
//System.out.println("palindrome: one two three: "+score2);
QueryUtils.Check(query, searcher);
//commented out for sloppy-phrase efficiency (issue 736) - see SloppyPhraseScorer.phraseFreq().
//Assert.IsTrue(score1+SCORE_COMP_THRESH<score2,"ordered scores higher in palindrome");
// search reveresed in palyndrome, find it twice
query = new PhraseQuery();
query.SetSlop(4); // must be at least four for both ordered and reversed to match
query.Add(new Term("palindrome", "three"));
query.Add(new Term("palindrome", "two"));
query.Add(new Term("palindrome", "one"));
hits = searcher.Search(query, null, 1000).scoreDocs;
Assert.AreEqual(1, hits.Length, "just sloppy enough");
float score3 = hits[0].score;
//System.out.println("palindrome: three two one: "+score3);
QueryUtils.Check(query, searcher);
//commented out for sloppy-phrase efficiency (issue 736) - see SloppyPhraseScorer.phraseFreq().
//Assert.IsTrue(score1+SCORE_COMP_THRESH<score3,"reversed scores higher in palindrome");
//Assert.AreEqual(score2, score3, SCORE_COMP_THRESH, "ordered or reversed does not matter");
}
/// <summary>
/// 搜索
/// </summary>
protected void SearchContent(string kw)
{
string indexPath = @"D:\lucenedir";
kw = kw.ToLower();//默认情况下盘古分词区分大小写，需转换成小写进行搜索
FSDirectory directory = FSDirectory.Open(new DirectoryInfo(indexPath), new NoLockFactory());
IndexReader reader = IndexReader.Open(directory, true);
IndexSearcher searcher = new IndexSearcher(reader);
//搜索条件
PhraseQuery queryMsg = new PhraseQuery();
foreach (string word in Common.WebCommon.PanGuSplit(kw))//先用空格，让用户去分词，空格分隔的就是词“计算机 专业”
{
queryMsg.Add(new Term("msg", word));//根据文章内容进行搜索
}
//query.Add(new Term("body","语言"));--可以添加查询条件，两者是add关系.顺序没有关系.
//query.Add(new Term("body", "大学生"));
queryMsg.SetSlop(100);//多个查询条件的词之间的最大距离.在文章中相隔太远 也就无意义.（例如 “大学生”这个查询条件和"简历"这个查询条件之间如果间隔的词太多也就没有意义了。）
PhraseQuery queryTitle = new PhraseQuery();
foreach (string word in Common.WebCommon.PanGuSplit(kw))
{
queryTitle.Add(new Term("title", word));
}
queryTitle.SetSlop(100);
BooleanQuery query = new BooleanQuery();
query.Add(queryMsg, BooleanClause.Occur.SHOULD);
query.Add(queryTitle, BooleanClause.Occur.SHOULD);
//TopScoreDocCollector是盛放查询结果的容器
TopScoreDocCollector collector = TopScoreDocCollector.create(1000, true);
searcher.Search(query, null, collector);//根据query查询条件进行查询，查询结果放入collector容器
ScoreDoc[] docs = collector.TopDocs(0, collector.GetTotalHits()).scoreDocs;//得到所有查询结果中的文档,GetTotalHits():表示总条数 TopDocs(300, 20);//表示得到 300（从300开始）到320（结束）的文档内容.可以用来实现分页功能
List<SearchResult> list = new List<SearchResult>();
for (int i = 0; i < docs.Length; i++)
{
//搜索ScoreDoc[]只能获得文档的id,这样不会把查询结果的Document一次性加载到内存中。降低了内存压力，需要获得文档的详细内容的时候通过searcher.Doc来根据文档id来获得文档的详细内容对象Document.
int docId = docs[i].doc;//得到查询结果文档的id（Lucene内部分配的id）
Document doc = searcher.Doc(docId);//找到文档id对应的文档详细信息
SearchResult result = new SearchResult();
result.ContentDescription = WebCommon.Highlight(kw,WebCommon.CutString(doc.Get("msg"),150));//分词高亮显示
result.Title = doc.Get("title");
result.Id = Convert.ToInt32(doc.Get("id"));
result.PublishDate = Convert.ToDateTime(doc.Get("PublishDate"));
result.ISBN = doc.Get("ISBN");
result.Author = doc.Get("Author");
result.UnitPrice = Convert.ToDecimal(doc.Get("UnitPrice"));
list.Add(result);
}
this.BookListRepeater.DataSource = list;
this.BookListRepeater.DataBind();
AddKeyWord(kw);
}
public virtual void TestSimilarity_()
{
RAMDirectory store = new RAMDirectory();
IndexWriter writer = new IndexWriter(store, new SimpleAnalyzer(), true);
writer.SetSimilarity(new SimpleSimilarity());
Document d1 = new Document();
d1.Add(Field.Text("Field", "a c"));
Document d2 = new Document();
d2.Add(Field.Text("Field", "a b c"));
writer.AddDocument(d1);
writer.AddDocument(d2);
writer.Optimize();
writer.Close();
float[] scores = new float[4];
Searcher searcher = new IndexSearcher(store);
searcher.SetSimilarity(new SimpleSimilarity());
Term a = new Term("Field", "a");
Term b = new Term("Field", "b");
Term c = new Term("Field", "c");
searcher.Search(new TermQuery(b), new AnonymousClassHitCollector(this));
BooleanQuery bq = new BooleanQuery();
bq.Add(new TermQuery(a), false, false);
bq.Add(new TermQuery(b), false, false);
//System.out.println(bq.toString("Field"));
searcher.Search(bq, new AnonymousClassHitCollector1(this));
PhraseQuery pq = new PhraseQuery();
pq.Add(a);
pq.Add(c);
//System.out.println(pq.toString("Field"));
searcher.Search(pq, new AnonymousClassHitCollector2(this));
pq.SetSlop(2);
//System.out.println(pq.toString("Field"));
searcher.Search(pq, new AnonymousClassHitCollector3(this));
}
public virtual void TestMulipleTerms()
{
SetUp();
query.SetSlop(2);
query.Add(new Term("Field", "one"));
query.Add(new Term("Field", "three"));
query.Add(new Term("Field", "five"));
Hits hits = searcher.Search(query);
Assert.AreEqual(1, hits.Length(), "two total moves");
query = new PhraseQuery();
query.SetSlop(5); // it takes six moves to match this phrase
query.Add(new Term("Field", "five"));
query.Add(new Term("Field", "three"));
query.Add(new Term("Field", "one"));
hits = searcher.Search(query);
Assert.AreEqual(0, hits.Length(), "slop of 5 not close enough");
query.SetSlop(6);
hits = searcher.Search(query);
Assert.AreEqual(1, hits.Length(), "slop of 6 just right");
}
public virtual void TestSimilarity_Renamed()
{
RAMDirectory store = new RAMDirectory();
IndexWriter writer = new IndexWriter(store, new SimpleAnalyzer(), true, IndexWriter.MaxFieldLength.LIMITED);
writer.SetSimilarity(new SimpleSimilarity());
Document d1 = new Document();
d1.Add(new Field("field", "a c", Field.Store.YES, Field.Index.ANALYZED));
Document d2 = new Document();
d2.Add(new Field("field", "a b c", Field.Store.YES, Field.Index.ANALYZED));
writer.AddDocument(d1);
writer.AddDocument(d2);
writer.Optimize();
writer.Close();
Searcher searcher = new IndexSearcher(store);
searcher.SetSimilarity(new SimpleSimilarity());
Term a = new Term("field", "a");
Term b = new Term("field", "b");
Term c = new Term("field", "c");
searcher.Search(new TermQuery(b), new AnonymousClassCollector(this));
BooleanQuery bq = new BooleanQuery();
bq.Add(new TermQuery(a), BooleanClause.Occur.SHOULD);
bq.Add(new TermQuery(b), BooleanClause.Occur.SHOULD);
//System.out.println(bq.toString("field"));
searcher.Search(bq, new AnonymousClassCollector1(this));
PhraseQuery pq = new PhraseQuery();
pq.Add(a);
pq.Add(c);
//System.out.println(pq.toString("field"));
searcher.Search(pq, new AnonymousClassCollector2(this));
pq.SetSlop(2);
//System.out.println(pq.toString("field"));
searcher.Search(pq, new AnonymousClassCollector3(this));
}
public virtual void TestNonExistingPhrase()
{
// phrase without repetitions that exists in 2 docs
query.Add(new Term("nonexist", "phrase"));
query.Add(new Term("nonexist", "notexist"));
query.Add(new Term("nonexist", "found"));
query.SetSlop(2); // would be found this way
ScoreDoc[] hits = searcher.Search(query, null, 1000).scoreDocs;
Assert.AreEqual(2, hits.Length, "phrase without repetitions exists in 2 docs");
QueryUtils.Check(query, searcher);
// phrase with repetitions that exists in 2 docs
query = new PhraseQuery();
query.Add(new Term("nonexist", "phrase"));
query.Add(new Term("nonexist", "exist"));
query.Add(new Term("nonexist", "exist"));
query.SetSlop(1); // would be found
hits = searcher.Search(query, null, 1000).scoreDocs;
Assert.AreEqual(2, hits.Length, "phrase with repetitions exists in two docs");
QueryUtils.Check(query, searcher);
// phrase I with repetitions that does not exist in any doc
query = new PhraseQuery();
query.Add(new Term("nonexist", "phrase"));
query.Add(new Term("nonexist", "notexist"));
query.Add(new Term("nonexist", "phrase"));
query.SetSlop(1000); // would not be found no matter how high the slop is
hits = searcher.Search(query, null, 1000).scoreDocs;
Assert.AreEqual(0, hits.Length, "nonexisting phrase with repetitions does not exist in any doc");
QueryUtils.Check(query, searcher);
// phrase II with repetitions that does not exist in any doc
query = new PhraseQuery();
query.Add(new Term("nonexist", "phrase"));
query.Add(new Term("nonexist", "exist"));
query.Add(new Term("nonexist", "exist"));
query.Add(new Term("nonexist", "exist"));
query.SetSlop(1000); // would not be found no matter how high the slop is
hits = searcher.Search(query, null, 1000).scoreDocs;
Assert.AreEqual(0, hits.Length, "nonexisting phrase with repetitions does not exist in any doc");
QueryUtils.Check(query, searcher);
}
public virtual void TestOrderDoesntMatter()
{
SetUp();
query.SetSlop(2); // must be at least two for reverse order match
query.Add(new Term("Field", "two"));
query.Add(new Term("Field", "one"));
Hits hits = searcher.Search(query);
Assert.AreEqual(1, hits.Length(), "just sloppy enough");
query = new PhraseQuery();
query.SetSlop(2);
query.Add(new Term("Field", "three"));
query.Add(new Term("Field", "one"));
hits = searcher.Search(query);
Assert.AreEqual(0, hits.Length(), "not sloppy enough");
}
public virtual void TestOrderDoesntMatter()
{
query.SetSlop(2); // must be at least two for reverse order match
query.Add(new Term("field", "two"));
query.Add(new Term("field", "one"));
ScoreDoc[] hits = searcher.Search(query, null, 1000).scoreDocs;
Assert.AreEqual(1, hits.Length, "just sloppy enough");
QueryUtils.Check(query, searcher);
query = new PhraseQuery();
query.SetSlop(2);
query.Add(new Term("field", "three"));
query.Add(new Term("field", "one"));
hits = searcher.Search(query, null, 1000).scoreDocs;
Assert.AreEqual(0, hits.Length, "not sloppy enough");
QueryUtils.Check(query, searcher);
}
public virtual void TestSlopScoring()
{
Directory directory = new RAMDirectory();
IndexWriter writer = new IndexWriter(directory, new WhitespaceAnalyzer(), true, IndexWriter.MaxFieldLength.LIMITED);
Document doc = new Document();
doc.Add(new Field("field", "foo firstname lastname foo", Field.Store.YES, Field.Index.ANALYZED));
writer.AddDocument(doc);
Document doc2 = new Document();
doc2.Add(new Field("field", "foo firstname xxx lastname foo", Field.Store.YES, Field.Index.ANALYZED));
writer.AddDocument(doc2);
Document doc3 = new Document();
doc3.Add(new Field("field", "foo firstname xxx yyy lastname foo", Field.Store.YES, Field.Index.ANALYZED));
writer.AddDocument(doc3);
writer.Optimize();
writer.Close();
Searcher searcher = new IndexSearcher(directory);
PhraseQuery query = new PhraseQuery();
query.Add(new Term("field", "firstname"));
query.Add(new Term("field", "lastname"));
query.SetSlop(System.Int32.MaxValue);
ScoreDoc[] hits = searcher.Search(query, null, 1000).scoreDocs;
Assert.AreEqual(3, hits.Length);
// Make sure that those matches where the terms appear closer to
// each other get a higher score:
Assert.AreEqual(0.71, hits[0].score, 0.01);
Assert.AreEqual(0, hits[0].doc);
Assert.AreEqual(0.44, hits[1].score, 0.01);
Assert.AreEqual(1, hits[1].doc);
Assert.AreEqual(0.31, hits[2].score, 0.01);
Assert.AreEqual(2, hits[2].doc);
QueryUtils.Check(query, searcher);
}
public virtual void TestSlop1()
{
SetUp();
// Ensures slop of 1 works with terms in order.
query.SetSlop(1);
query.Add(new Term("Field", "one"));
query.Add(new Term("Field", "two"));
Hits hits = searcher.Search(query);
Assert.AreEqual(1, hits.Length(), "in order");
// Ensures slop of 1 does not work for phrases out of order;
// must be at least 2.
query = new PhraseQuery();
query.SetSlop(1);
query.Add(new Term("Field", "two"));
query.Add(new Term("Field", "one"));
hits = searcher.Search(query);
Assert.AreEqual(0, hits.Length(), "reversed, slop not 2 or more");
}
public virtual void TestSlop1()
{
// Ensures slop of 1 works with terms in order.
query.SetSlop(1);
query.Add(new Term("field", "one"));
query.Add(new Term("field", "two"));
ScoreDoc[] hits = searcher.Search(query, null, 1000).scoreDocs;
Assert.AreEqual(1, hits.Length, "in order");
QueryUtils.Check(query, searcher);
// Ensures slop of 1 does not work for phrases out of order;
// must be at least 2.
query = new PhraseQuery();
query.SetSlop(1);
query.Add(new Term("field", "two"));
query.Add(new Term("field", "one"));
hits = searcher.Search(query, null, 1000).scoreDocs;
Assert.AreEqual(0, hits.Length, "reversed, slop not 2 or more");
QueryUtils.Check(query, searcher);
}
public virtual int DoSloppyPhrase(IndexSearcher s, int termsInIndex, int maxClauses, int iter)
{
int ret = 0;
for (int i = 0; i < iter; i++)
{
int nClauses = r.Next(maxClauses - 1) + 2; // min 2 clauses
PhraseQuery q = new PhraseQuery();
for (int j = 0; j < nClauses; j++)
{
int tnum = r.Next(termsInIndex);
q.Add(new Term("f", System.Convert.ToString((char)(tnum + 'A'))), j);
}
q.SetSlop(termsInIndex); // this could be random too
CountingHitCollector hc = new CountingHitCollector();
s.Search(q, hc);
ret += hc.GetSum();
}
return(ret);
}
/// <summary>
/// 从索引库中检索关键字
/// </summary>
private void SearchFromIndexData() {
string indexPath = Context.Server.MapPath("~/IndexData");
FSDirectory directory = FSDirectory.Open(new DirectoryInfo(indexPath), new NoLockFactory());
IndexReader reader = IndexReader.Open(directory, true);
IndexSearcher searcher = new IndexSearcher(reader);
//搜索条件
PhraseQuery query = new PhraseQuery();
//把用户输入的关键字进行分词
foreach(string word in Common.SplitContent.SplitWords(Request.QueryString["SearchKey"])) {
query.Add(new Term("content", word));
}
//query.Add(new Term("content", "C#"));//多个查询条件时 为且的关系
query.SetSlop(100); //指定关键词相隔最大距离
//TopScoreDocCollector盛放查询结果的容器
TopScoreDocCollector collector = TopScoreDocCollector.create(1000, true);
searcher.Search(query, null, collector);//根据query查询条件进行查询，查询结果放入collector容器
//TopDocs 指定0到GetTotalHits() 即所有查询结果中的文档 如果TopDocs(20,10)则意味着获取第20-30之间文档内容 达到分页的效果
ScoreDoc[] docs = collector.TopDocs(0, collector.GetTotalHits()).scoreDocs;
//展示数据实体对象集合
List<PZYM.Shop.Model.Books> bookResult = new List<PZYM.Shop.Model.Books>();
for(int i = 0; i < docs.Length; i++) {
int docId = docs[i].doc;//得到查询结果文档的id（Lucene内部分配的id）
Document doc = searcher.Doc(docId);//根据文档id来获得文档对象Document
PZYM.Shop.Model.Books book = new PZYM.Shop.Model.Books();
book.Title = doc.Get("title");
//book.ContentDescription = doc.Get("content");//未使用高亮
//搜索关键字高亮显示 使用盘古提供高亮插件
book.ContentDescription = Common.SplitContent.HightLight(Request.QueryString["SearchKey"], doc.Get("content"));
book.Id = Convert.ToInt32(doc.Get("id"));
bookResult.Add(book);
}
Repeater1.DataSource = bookResult;
Repeater1.DataBind();
}
private float CheckPhraseQuery(Document doc, PhraseQuery query, int slop, int expectedNumResults)
{
query.SetSlop(slop);
RAMDirectory ramDir = new RAMDirectory();
WhitespaceAnalyzer analyzer = new WhitespaceAnalyzer();
IndexWriter writer = new IndexWriter(ramDir, analyzer, MaxFieldLength.UNLIMITED);
writer.AddDocument(doc);
writer.Close();
IndexSearcher searcher = new IndexSearcher(ramDir);
TopDocs td = searcher.Search(query, null, 10);
//System.out.println("slop: "+slop+" query: "+query+" doc: "+doc+" Expecting number of hits: "+expectedNumResults+" maxScore="+td.getMaxScore());
Assert.AreEqual(expectedNumResults, td.TotalHits, "slop: " + slop + " query: " + query + " doc: " + doc + " Wrong number of hits");
//QueryUtils.check(query,searcher);
searcher.Close();
ramDir.Close();
return(td.GetMaxScore());
}
/// <summary>
/// 从索引库中检索关键字
/// </summary>
private static List<CommentSet> SearchFromIndexData(string IndexSavePath, string searchKey)
{
string indexPath = IndexSavePath;
FSDirectory directory = FSDirectory.Open(new DirectoryInfo(indexPath), new NoLockFactory());
IndexReader reader = IndexReader.Open(directory, true);
IndexSearcher searcher = new IndexSearcher(reader);
//搜索条件
PhraseQuery query = new PhraseQuery();
//把用户输入的关键字进行分词
foreach (string word in Picture.Utility.SplitContent.SplitWords(searchKey))
{
query.Add(new Term("content", word));
}
//query.Add(new Term("content", "C#"));//多个查询条件时 为且的关系
query.SetSlop(100); //指定关键词相隔最大距离
//TopScoreDocCollector盛放查询结果的容器
TopScoreDocCollector collector = TopScoreDocCollector.create(1000, true);
searcher.Search(query, null, collector);//根据query查询条件进行查询，查询结果放入collector容器
//TopDocs 指定0到GetTotalHits() 即所有查询结果中的文档 如果TopDocs(20,10)则意味着获取第20-30之间文档内容 达到分页的效果
ScoreDoc[] docs = collector.TopDocs(0, collector.GetTotalHits()).scoreDocs;
//展示数据实体对象集合
var commontModels = new List<CommentSet>();
for (int i = 0; i < docs.Length; i++)
{
int docId = docs[i].doc;//得到查询结果文档的id（Lucene内部分配的id）
Document doc = searcher.Doc(docId);//根据文档id来获得文档对象Document
CommentSet commont = new CommentSet();
commont.Content = doc.Get("content");
//book.ContentDescription = doc.Get("content");//未使用高亮
//搜索关键字高亮显示 使用盘古提供高亮插件
//book.ContentDescription = Picture.Utility.SplitContent.HightLight(Request.QueryString["SearchKey"], doc.Get("content"));
commont.Id = Convert.ToInt32(doc.Get("id"));
commontModels.Add(commont);
}
return commontModels;
}
protected Query Pq(float boost, int slop, String field, params String[] texts)
{
PhraseQuery query = new PhraseQuery();
foreach (String text in texts)
{
query.Add(new Term(field, text));
}
query.SetBoost(boost);
query.SetSlop(slop);
return query;
}
public void CompositeTermPhraseWildcardTests()
{
BooleanQuery originalQuery = new BooleanQuery();
Term term = new Term("_name", "value");
TermQuery termQuery = new TermQuery(term);
originalQuery.Add(termQuery, BooleanClause.Occur.MUST);
PhraseQuery phraseQuery = new PhraseQuery();
Term phraseTerm = new Term("_name", "phrase");
phraseQuery.SetSlop(2);
phraseQuery.Add(phraseTerm);
originalQuery.Add(phraseQuery, BooleanClause.Occur.MUST);
Term wildcardTerm = new Term("_name", "*wildcard*");
WildcardQuery wildcardQuery = new WildcardQuery(wildcardTerm);
originalQuery.Add(wildcardQuery, BooleanClause.Occur.SHOULD);
string queryString = originalQuery.ToString();
QueryBuilder builder = new QueryBuilder();
builder.Setup
(
x => x.Term("_name", "value"),
x => x.Phrase(2).AddTerm("_name", "phrase"),
x => x.WildCard("_name", "*wildcard*", BooleanClause.Occur.SHOULD)
);
Query replacementQuery = builder.Build();
string newQueryString = replacementQuery.ToString();
Assert.AreEqual(queryString, newQueryString);
Console.Write(queryString);
}
public void BoostedCaseInsensitivePhrase()
{
BooleanQuery originalQuery = new BooleanQuery();
Term term = new Term("_name", "value");
PhraseQuery phraseQuery = new PhraseQuery();
phraseQuery.SetSlop(2);
phraseQuery.Add(term);
phraseQuery.SetBoost(10);
originalQuery.Add(phraseQuery, BooleanClause.Occur.MUST);
string queryString = originalQuery.ToString();
QueryBuilder builder = new QueryBuilder();
builder.Setup(x => x.Phrase(2, 10).AddTerm("_name", "Value"));
Query replacementQuery = builder.Build();
string newQueryString = replacementQuery.ToString();
Assert.AreEqual(queryString, newQueryString);
Console.Write(queryString);
}
/*
* Check if src and dest have overlapped part and if it is, create PhraseQueries and add expandQueries.
*
* ex1) src="a b", dest="c d" => no overlap
* ex2) src="a b", dest="a b c" => no overlap
* ex3) src="a b", dest="b c" => overlap; expandQueries={"a b c"}
* ex4) src="a b c", dest="b c d" => overlap; expandQueries={"a b c d"}
* ex5) src="a b c", dest="b c" => no overlap
* ex6) src="a b c", dest="b" => no overlap
* ex7) src="a a a a", dest="a a a" => overlap;
* expandQueries={"a a a a a","a a a a a a"}
* ex8) src="a b c d", dest="b c" => no overlap
*/
private void CheckOverlap(Dictionary<Query,Query> expandQueries, Term[] src, Term[] dest, int slop, float boost)
{
// beginning from 1 (not 0) is safe because that the PhraseQuery has multiple terms
// is guaranteed in flatten() method (if PhraseQuery has only one term, flatten()
// converts PhraseQuery to TermQuery)
for (int i = 1; i < src.Length; i++)
{
bool overlap = true;
for (int j = i; j < src.Length; j++)
{
if ((j - i) < dest.Length && !src[j].Text().Equals(dest[j - i].Text()))
{
overlap = false;
break;
}
}
if (overlap && src.Length - i < dest.Length)
{
PhraseQuery pq = new PhraseQuery();
foreach (Term srcTerm in src)
pq.Add(srcTerm);
for (int k = src.Length - i; k < dest.Length; k++)
{
pq.Add(new Term(src[0].Field(), dest[k].Text()));
}
pq.SetSlop(slop);
pq.SetBoost(boost);
if (!expandQueries.ContainsKey(pq))
expandQueries.Add(pq,pq);
}
}
}
public virtual void TestNotCloseEnough()
{
query.SetSlop(2);
query.Add(new Term("field", "one"));
query.Add(new Term("field", "five"));
ScoreDoc[] hits = searcher.Search(query, null, 1000).scoreDocs;
Assert.AreEqual(0, hits.Length);
QueryUtils.Check(query, searcher);
}
private Query CreateStringValueQuery(QueryFieldValue value, FieldInfo currentField)
{
switch (value.Token)
{
case SnLucLexer.Token.Number:
case SnLucLexer.Token.String:
if(value.StringValue == ContentQuery.EmptyText)
return new TermQuery(new Term(currentField.Name, value.StringValue));
if (value.StringValue == ContentQuery.EmptyInnerQueryText)
return new TermQuery(new Term("Id", NumericUtils.IntToPrefixCoded(0)));
var words = GetAnalyzedText(currentField.Name, value.StringValue);
if (words.Length == 0)
words = new String[] { String.Empty }; //return null;
if (words.Length == 1)
{
var term = new Term(currentField.Name, words[0]);
if(value.FuzzyValue == null)
return new TermQuery(term);
return new FuzzyQuery(term, Convert.ToSingle(value.FuzzyValue));
}
var phraseQuery = new PhraseQuery();
foreach(var word in words)
phraseQuery.Add(new Term(currentField.Name, word));
if (value.FuzzyValue != null)
{
var slop = Convert.ToInt32(value.FuzzyValue.Value);
phraseQuery.SetSlop(slop);
}
return phraseQuery;
case SnLucLexer.Token.WildcardString:
if (!value.StringValue.EndsWith("*"))
return new WildcardQuery(new Term(currentField.Name, value.StringValue));
var s = value.StringValue.TrimEnd('*');
if (s.Contains('?') || s.Contains('*'))
return new WildcardQuery(new Term(currentField.Name, value.StringValue));
return new PrefixQuery(new Term(currentField.Name, s));
default:
throw new NotImplementedException("CreateValueQuery with Token: " + value.Token);
}
}
public virtual int DoSloppyPhrase(IndexSearcher s, int termsInIndex, int maxClauses, int iter)
{
int ret = 0;
for (int i = 0; i < iter; i++)
{
int nClauses = r.Next(maxClauses - 1) + 2; // min 2 clauses
PhraseQuery q = new PhraseQuery();
for (int j = 0; j < nClauses; j++)
{
int tnum = r.Next(termsInIndex);
q.Add(new Term("f", System.Convert.ToString((char)(tnum + 'A'))), j);
}
q.SetSlop(termsInIndex); // this could be random too
CountingHitCollector hc = new CountingHitCollector();
s.Search(q, hc);
ret += hc.GetSum();
}
return ret;
}
public void PhraseQueryText(string[] frase, string textoParaProcurar, int distanciaEntrePalavras)
{
const string texto = "texto";
using (var directory = new RAMDirectory())
{
using (var indexWriter = new IndexWriter(directory, new WhitespaceAnalyzer(), IndexWriter.MaxFieldLength.UNLIMITED))
{
var doc = new Document();
doc.Add(new Field(texto, textoParaProcurar, Field.Store.YES, Field.Index.ANALYZED));
indexWriter.AddDocument(doc);
}
using (var searcher = new IndexSearcher(directory, true))
{
var phraseQuery = new PhraseQuery();
phraseQuery.SetSlop(distanciaEntrePalavras);
foreach (var palavra in frase)
phraseQuery.Add(new Term(texto, palavra));
var matches = searcher.Search(phraseQuery, 10);
var encontrou = matches.TotalHits > 0;
var textoResultado = NaoEncontrou(textoParaProcurar, distanciaEntrePalavras, frase);
Assert.IsTrue(encontrou, textoResultado);
}
}
}
public virtual void TestPalyndrome3()
{
// search on non palyndrome, find phrase with no slop, using exact phrase scorer
query.SetSlop(0); // to use exact phrase scorer
query.Add(new Term("field", "one"));
query.Add(new Term("field", "two"));
query.Add(new Term("field", "three"));
Hits hits = searcher.Search(query);
Assert.AreEqual(1, hits.Length(), "phrase found with exact phrase scorer");
float score0 = hits.Score(0);
//System.out.println("(exact) field: one two three: "+score0);
QueryUtils.Check(query, searcher);
// search on non palyndrome, find phrase with slop 3, though no slop required here.
query.SetSlop(4); // to use sloppy scorer
hits = searcher.Search(query);
Assert.AreEqual(1, hits.Length(), "just sloppy enough");
float score1 = hits.Score(0);
//System.out.println("(sloppy) field: one two three: "+score1);
Assert.AreEqual(score0, score1, SCORE_COMP_THRESH, "exact scorer and sloppy scorer score the same when slop does not matter");
QueryUtils.Check(query, searcher);
// search ordered in palyndrome, find it twice
query = new PhraseQuery();
query.SetSlop(4); // must be at least four for both ordered and reversed to match
query.Add(new Term("palindrome", "one"));
query.Add(new Term("palindrome", "two"));
query.Add(new Term("palindrome", "three"));
hits = searcher.Search(query);
Assert.AreEqual(1, hits.Length(), "just sloppy enough");
float score2 = hits.Score(0);
//System.out.println("palindrome: one two three: "+score2);
QueryUtils.Check(query, searcher);
//commented out for sloppy-phrase efficiency (issue 736) - see SloppyPhraseScorer.phraseFreq().
//assertTrue("ordered scores higher in palindrome",score1+SCORE_COMP_THRESH<score2);
// search reveresed in palyndrome, find it twice
query = new PhraseQuery();
query.SetSlop(4); // must be at least four for both ordered and reversed to match
query.Add(new Term("palindrome", "three"));
query.Add(new Term("palindrome", "two"));
query.Add(new Term("palindrome", "one"));
hits = searcher.Search(query);
Assert.AreEqual(1, hits.Length(), "just sloppy enough");
float score3 = hits.Score(0);
//System.out.println("palindrome: three two one: "+score3);
QueryUtils.Check(query, searcher);
//commented out for sloppy-phrase efficiency (issue 736) - see SloppyPhraseScorer.phraseFreq().
//assertTrue("reversed scores higher in palindrome",score1+SCORE_COMP_THRESH<score3);
//assertEquals("ordered or reversed does not matter",score2, score3, SCORE_COMP_THRESH);
}
public virtual void TestSlopScoring()
{
Directory directory = new RAMDirectory();
IndexWriter writer = new IndexWriter(directory, new WhitespaceAnalyzer(), true);
Lucene.Net.Documents.Document doc = new Lucene.Net.Documents.Document();
doc.Add(new Field("field", "foo firstname lastname foo", Field.Store.YES, Field.Index.TOKENIZED));
writer.AddDocument(doc);
Lucene.Net.Documents.Document doc2 = new Lucene.Net.Documents.Document();
doc2.Add(new Field("field", "foo firstname xxx lastname foo", Field.Store.YES, Field.Index.TOKENIZED));
writer.AddDocument(doc2);
Lucene.Net.Documents.Document doc3 = new Lucene.Net.Documents.Document();
doc3.Add(new Field("field", "foo firstname xxx yyy lastname foo", Field.Store.YES, Field.Index.TOKENIZED));
writer.AddDocument(doc3);
writer.Optimize();
writer.Close();
Searcher searcher = new IndexSearcher(directory);
PhraseQuery query = new PhraseQuery();
query.Add(new Term("field", "firstname"));
query.Add(new Term("field", "lastname"));
query.SetSlop(System.Int32.MaxValue);
Hits hits = searcher.Search(query);
Assert.AreEqual(3, hits.Length());
// Make sure that those matches where the terms appear closer to
// each other get a higher score:
Assert.AreEqual(0.71, hits.Score(0), 0.01);
Assert.AreEqual(0, hits.Id(0));
Assert.AreEqual(0.44, hits.Score(1), 0.01);
Assert.AreEqual(1, hits.Id(1));
Assert.AreEqual(0.31, hits.Score(2), 0.01);
Assert.AreEqual(2, hits.Id(2));
QueryUtils.Check(query, searcher);
}
public virtual void TestOrderDoesntMatter()
{
query.SetSlop(2); // must be at least two for reverse order match
query.Add(new Term("field", "two"));
query.Add(new Term("field", "one"));
Hits hits = searcher.Search(query);
Assert.AreEqual(1, hits.Length(), "just sloppy enough");
QueryUtils.Check(query, searcher);
query = new PhraseQuery();
query.SetSlop(2);
query.Add(new Term("field", "three"));
query.Add(new Term("field", "one"));
hits = searcher.Search(query);
Assert.AreEqual(0, hits.Length(), "not sloppy enough");
QueryUtils.Check(query, searcher);
}
private float CheckPhraseQuery(Document doc, PhraseQuery query, int slop, int expectedNumResults)
{
query.SetSlop(slop);
RAMDirectory ramDir = new RAMDirectory();
WhitespaceAnalyzer analyzer = new WhitespaceAnalyzer();
IndexWriter writer = new IndexWriter(ramDir, analyzer, MaxFieldLength.UNLIMITED);
writer.AddDocument(doc);
writer.Close();
IndexSearcher searcher = new IndexSearcher(ramDir);
TopDocs td = searcher.Search(query, null, 10);
//System.out.println("slop: "+slop+" query: "+query+" doc: "+doc+" Expecting number of hits: "+expectedNumResults+" maxScore="+td.getMaxScore());
Assert.AreEqual(expectedNumResults, td.totalHits, "slop: " + slop + " query: " + query + " doc: " + doc + " Wrong number of hits");
//QueryUtils.check(query,searcher);
searcher.Close();
ramDir.Close();
return td.GetMaxScore();
}
/// <summary>
/// 添加PhraseQuery
/// </summary>
/// <param name="fieldName">待搜索的字段名称</param>
/// <param name="phrase">待搜索的短语</param>
/// <param name="boostLevel">权重级别</param>
/// <param name="asFilter">是否作为过滤条件</param>
/// <returns>LuceneSearchBuilder</returns>
public LuceneSearchBuilder WithPhrase(string fieldName, string phrase, BoostLevel? boostLevel = null, bool asFilter = false)
{
string filteredPhrase = ClauseScrubber.LuceneKeywordsScrub(phrase);
if (string.IsNullOrEmpty(filteredPhrase))
return this;
if (filteredPhrase.Length == 1)
return WithField(fieldName, filteredPhrase, false, boostLevel, asFilter);
string[] nameSegments = ClauseScrubber.SegmentForPhraseQuery(filteredPhrase);
PhraseQuery phraseQuery = new PhraseQuery();
foreach (var nameSegment in nameSegments)
phraseQuery.Add(new Term(fieldName, nameSegment));
phraseQuery.SetSlop(PhraseQuerySlop);
if (boostLevel.HasValue)
SetBoost(phraseQuery, boostLevel.Value);
if (asFilter)
filters.Add(new BooleanClause(phraseQuery, BooleanClause.Occur.MUST));
else
clauses.Add(new BooleanClause(phraseQuery, BooleanClause.Occur.MUST));
return this;
}
/// <summary>
/// Sets up and adds a phrase query object allowing the search for an explcit term in the field
/// To add terms, use the AddTerm() query extension
/// </summary>
/// <param name="occur">Whether it must, must not or should occur in the field</param>
/// <param name="slop">The allowed distance between the terms</param>
/// <param name="boost">A boost multiplier (1 is default / normal).</param>
/// <param name="key">The dictionary key to allow reference beyond the initial scope</param>
/// <returns>The generated phrase query object</returns>
public virtual PhraseQuery Phrase(int slop, float? boost = null, BooleanClause.Occur occur = null, string key = null)
{
PhraseQuery query = new PhraseQuery();
SetBoostValue(query, boost);
query.SetSlop(slop);
Add(query, occur, key);
return query;
}
/// <summary>
/// 批量添加PhraseQuery
/// </summary>
/// <param name="phrase">待搜索的短语</param>
/// <param name="fieldNameAndBoosts">字段名称及权重集合</param>
/// <param name="occur">搜索条件间的关系</param>
/// <param name="asFilter">是否作为过滤器条件</param>
/// <returns></returns>
public LuceneSearchBuilder WithPhrases(Dictionary<string, BoostLevel> fieldNameAndBoosts, string phrase, BooleanClause.Occur occur, bool asFilter = false)
{
string filteredPhrase = ClauseScrubber.LuceneKeywordsScrub(phrase);
if (string.IsNullOrEmpty(filteredPhrase))
return this;
string[] nameSegments = ClauseScrubber.SegmentForPhraseQuery(filteredPhrase);
if (nameSegments.Length == 1)
{
return WithFields(fieldNameAndBoosts, nameSegments[0], false, occur, asFilter);
}
else
{
BooleanQuery query = new BooleanQuery();
foreach (var fieldNameAndBoost in fieldNameAndBoosts)
{
PhraseQuery phraseQuery = new PhraseQuery();
foreach (var nameSegment in nameSegments)
phraseQuery.Add(new Term(fieldNameAndBoost.Key, nameSegment));
phraseQuery.SetSlop(PhraseQuerySlop);
SetBoost(phraseQuery, fieldNameAndBoost.Value);
query.Add(phraseQuery, occur);
}
if (asFilter)
filters.Add(new BooleanClause(query, BooleanClause.Occur.MUST));
else
clauses.Add(new BooleanClause(query, BooleanClause.Occur.MUST));
return this;
}
}
//搜索
protected void SearchContent()
{
string indexPath = @"D:\lucenedir";
string kw = Request.Form["txtContent"];
kw = kw.ToLower();//默认情况下盘古分词区分大小写，需转换成小写进行搜索
FSDirectory directory = FSDirectory.Open(new DirectoryInfo(indexPath), new NoLockFactory());
IndexReader reader = IndexReader.Open(directory, true);
IndexSearcher searcher = new IndexSearcher(reader);
//搜索条件
PhraseQuery query = new PhraseQuery();
foreach (string word in Common.WebCommon.PanGuSplit(kw))//先用空格，让用户去分词，空格分隔的就是词“计算机 专业”
{
query.Add(new Term("msg", word));//根据文章内容进行搜索
}
//query.Add(new Term("body","语言"));--可以添加查询条件，两者是add关系.顺序没有关系.
//query.Add(new Term("body", "大学生"));
query.SetSlop(100);//多个查询条件的词之间的最大距离.在文章中相隔太远 也就无意义.（例如 “大学生”这个查询条件和"简历"这个查询条件之间如果间隔的词太多也就没有意义了。）
//TopScoreDocCollector是盛放查询结果的容器
TopScoreDocCollector collector = TopScoreDocCollector.create(1000, true);
searcher.Search(query, null, collector);//根据query查询条件进行查询，查询结果放入collector容器
ScoreDoc[] docs = collector.TopDocs(0, collector.GetTotalHits()).scoreDocs;//得到所有查询结果中的文档,GetTotalHits():表示总条数 TopDocs(300, 20);//表示得到 300（从300开始）到320（结束）的文档内容.可以用来实现分页功能
List<SearchResult> list = new List<SearchResult>();
for (int i = 0; i < docs.Length; i++)
{
//搜索ScoreDoc[]只能获得文档的id,这样不会把查询结果的Document一次性加载到内存中。降低了内存压力，需要获得文档的详细内容的时候通过searcher.Doc来根据文档id来获得文档的详细内容对象Document.
int docId = docs[i].doc;//得到查询结果文档的id（Lucene内部分配的id）
Document doc = searcher.Doc(docId);//找到文档id对应的文档详细信息
SearchResult result = new SearchResult();
//result.Msg = Common.WebCommon.Highlight(kw,doc.Get("msg"));//分词高亮显示
//result.Title = doc.Get("title");
//result.Url = "/BookDeatail.apsx?id=" + doc.Get("id");
list.Add(result);
}
this.SearchRepeater.DataSource = list;
this.SearchRepeater.DataBind();
}
/// <summary>
/// 批量添加PhraseQuery
/// </summary>
/// <param name="phrases">待搜索的短语集合</param>
/// <param name="fieldNameAndBoosts">字段名称及权重集合</param>
/// <param name="occur">搜索条件间的关系</param>
/// <param name="asFilter">是否作为过滤器条件</param>
/// <returns></returns>
public LuceneSearchBuilder WithPhrases(Dictionary<string, BoostLevel> fieldNameAndBoosts, IEnumerable<string> phrases, BooleanClause.Occur occur, bool asFilter = false)
{
foreach (var fieldNameAndBoost in fieldNameAndBoosts)
{
BooleanQuery query = new BooleanQuery();
foreach (string phrase in phrases)
{
string filteredPhrase = ClauseScrubber.LuceneKeywordsScrub(phrase);
if (string.IsNullOrEmpty(filteredPhrase))
continue;
if (filteredPhrase.Length == 1)
{
Term term = new Term(fieldNameAndBoost.Key, filteredPhrase);
Query q = new PrefixQuery(term);
SetBoost(q, fieldNameAndBoost.Value);
query.Add(q, BooleanClause.Occur.SHOULD);
continue;
}
string[] nameSegments = ClauseScrubber.SegmentForPhraseQuery(filteredPhrase);
PhraseQuery phraseQuery = new PhraseQuery();
foreach (var nameSegment in nameSegments)
phraseQuery.Add(new Term(fieldNameAndBoost.Key, nameSegment));
phraseQuery.SetSlop(PhraseQuerySlop);
SetBoost(phraseQuery, fieldNameAndBoost.Value);
query.Add(phraseQuery, BooleanClause.Occur.SHOULD);
}
if (asFilter)
filters.Add(new BooleanClause(query, occur));
else
clauses.Add(new BooleanClause(query, occur));
}
return this;
}
public virtual void TestSlop1()
{
// Ensures slop of 1 works with terms in order.
query.SetSlop(1);
query.Add(new Term("field", "one"));
query.Add(new Term("field", "two"));
Hits hits = searcher.Search(query);
Assert.AreEqual(1, hits.Length(), "in order");
QueryUtils.Check(query, searcher);
// Ensures slop of 1 does not work for phrases out of order;
// must be at least 2.
query = new PhraseQuery();
query.SetSlop(1);
query.Add(new Term("field", "two"));
query.Add(new Term("field", "one"));
hits = searcher.Search(query);
Assert.AreEqual(0, hits.Length(), "reversed, slop not 2 or more");
QueryUtils.Check(query, searcher);
}
public virtual void TestSimilarity_Renamed()
{
RAMDirectory store = new RAMDirectory();
IndexWriter writer = new IndexWriter(store, new SimpleAnalyzer(), true, IndexWriter.MaxFieldLength.LIMITED);
writer.SetSimilarity(new SimpleSimilarity());
Document d1 = new Document();
d1.Add(new Field("field", "a c", Field.Store.YES, Field.Index.ANALYZED));
Document d2 = new Document();
d2.Add(new Field("field", "a b c", Field.Store.YES, Field.Index.ANALYZED));
writer.AddDocument(d1);
writer.AddDocument(d2);
writer.Optimize();
writer.Close();
Searcher searcher = new IndexSearcher(store);
searcher.SetSimilarity(new SimpleSimilarity());
Term a = new Term("field", "a");
Term b = new Term("field", "b");
Term c = new Term("field", "c");
searcher.Search(new TermQuery(b), new AnonymousClassCollector(this));
BooleanQuery bq = new BooleanQuery();
bq.Add(new TermQuery(a), BooleanClause.Occur.SHOULD);
bq.Add(new TermQuery(b), BooleanClause.Occur.SHOULD);
//System.out.println(bq.toString("field"));
searcher.Search(bq, new AnonymousClassCollector1(this));
PhraseQuery pq = new PhraseQuery();
pq.Add(a);
pq.Add(c);
//System.out.println(pq.toString("field"));
searcher.Search(pq, new AnonymousClassCollector2(this));
pq.SetSlop(2);
//System.out.println(pq.toString("field"));
searcher.Search(pq, new AnonymousClassCollector3(this));
}
public virtual void TestMulipleTerms()
{
query.SetSlop(2);
query.Add(new Term("field", "one"));
query.Add(new Term("field", "three"));
query.Add(new Term("field", "five"));
Hits hits = searcher.Search(query);
Assert.AreEqual(1, hits.Length(), "two total moves");
QueryUtils.Check(query, searcher);
query = new PhraseQuery();
query.SetSlop(5); // it takes six moves to match this phrase
query.Add(new Term("field", "five"));
query.Add(new Term("field", "three"));
query.Add(new Term("field", "one"));
hits = searcher.Search(query);
Assert.AreEqual(0, hits.Length(), "slop of 5 not close enough");
QueryUtils.Check(query, searcher);
query.SetSlop(6);
hits = searcher.Search(query);
Assert.AreEqual(1, hits.Length(), "slop of 6 just right");
QueryUtils.Check(query, searcher);
}
public virtual void TestSlopScoring()
{
Directory directory = new RAMDirectory();
IndexWriter writer = new IndexWriter(directory, new WhitespaceAnalyzer(), true, IndexWriter.MaxFieldLength.LIMITED);
Document doc = new Document();
doc.Add(new Field("field", "foo firstname lastname foo", Field.Store.YES, Field.Index.ANALYZED));
writer.AddDocument(doc);
Document doc2 = new Document();
doc2.Add(new Field("field", "foo firstname xxx lastname foo", Field.Store.YES, Field.Index.ANALYZED));
writer.AddDocument(doc2);
Document doc3 = new Document();
doc3.Add(new Field("field", "foo firstname xxx yyy lastname foo", Field.Store.YES, Field.Index.ANALYZED));
writer.AddDocument(doc3);
writer.Optimize();
writer.Close();
Searcher searcher = new IndexSearcher(directory);
PhraseQuery query = new PhraseQuery();
query.Add(new Term("field", "firstname"));
query.Add(new Term("field", "lastname"));
query.SetSlop(System.Int32.MaxValue);
ScoreDoc[] hits = searcher.Search(query, null, 1000).ScoreDocs;
Assert.AreEqual(3, hits.Length);
// Make sure that those matches where the terms appear closer to
// each other get a higher score:
Assert.AreEqual(0.71, hits[0].Score, 0.01);
Assert.AreEqual(0, hits[0].Doc);
Assert.AreEqual(0.44, hits[1].Score, 0.01);
Assert.AreEqual(1, hits[1].Doc);
Assert.AreEqual(0.31, hits[2].Score, 0.01);
Assert.AreEqual(2, hits[2].Doc);
QueryUtils.Check(query, searcher);
}
public virtual void TestNonExistingPhrase()
{
// phrase without repetitions that exists in 2 docs
query.Add(new Term("nonexist", "phrase"));
query.Add(new Term("nonexist", "notexist"));
query.Add(new Term("nonexist", "found"));
query.SetSlop(2); // would be found this way
Hits hits = searcher.Search(query);
Assert.AreEqual(2, hits.Length(), "phrase without repetitions exists in 2 docs");
QueryUtils.Check(query, searcher);
// phrase with repetitions that exists in 2 docs
query = new PhraseQuery();
query.Add(new Term("nonexist", "phrase"));
query.Add(new Term("nonexist", "exist"));
query.Add(new Term("nonexist", "exist"));
query.SetSlop(1); // would be found
hits = searcher.Search(query);
Assert.AreEqual(2, hits.Length(), "phrase with repetitions exists in two docs");
QueryUtils.Check(query, searcher);
// phrase I with repetitions that does not exist in any doc
query = new PhraseQuery();
query.Add(new Term("nonexist", "phrase"));
query.Add(new Term("nonexist", "notexist"));
query.Add(new Term("nonexist", "phrase"));
query.SetSlop(1000); // would not be found no matter how high the slop is
hits = searcher.Search(query);
Assert.AreEqual(0, hits.Length(), "nonexisting phrase with repetitions does not exist in any doc");
QueryUtils.Check(query, searcher);
// phrase II with repetitions that does not exist in any doc
query = new PhraseQuery();
query.Add(new Term("nonexist", "phrase"));
query.Add(new Term("nonexist", "exist"));
query.Add(new Term("nonexist", "exist"));
query.Add(new Term("nonexist", "exist"));
query.SetSlop(1000); // would not be found no matter how high the slop is
hits = searcher.Search(query);
Assert.AreEqual(0, hits.Length(), "nonexisting phrase with repetitions does not exist in any doc");
QueryUtils.Check(query, searcher);
}
public virtual void TestSimilarity_()
{
RAMDirectory store = new RAMDirectory();
IndexWriter writer = new IndexWriter(store, new SimpleAnalyzer(), true);
writer.SetSimilarity(new SimpleSimilarity());
Document d1 = new Document();
d1.Add(Field.Text("Field", "a c"));
Document d2 = new Document();
d2.Add(Field.Text("Field", "a b c"));
writer.AddDocument(d1);
writer.AddDocument(d2);
writer.Optimize();
writer.Close();
float[] scores = new float[4];
Searcher searcher = new IndexSearcher(store);
searcher.SetSimilarity(new SimpleSimilarity());
Term a = new Term("Field", "a");
Term b = new Term("Field", "b");
Term c = new Term("Field", "c");
searcher.Search(new TermQuery(b), new AnonymousClassHitCollector(this));
BooleanQuery bq = new BooleanQuery();
bq.Add(new TermQuery(a), false, false);
bq.Add(new TermQuery(b), false, false);
//System.out.println(bq.toString("Field"));
searcher.Search(bq, new AnonymousClassHitCollector1(this));
PhraseQuery pq = new PhraseQuery();
pq.Add(a);
pq.Add(c);
//System.out.println(pq.toString("Field"));
searcher.Search(pq, new AnonymousClassHitCollector2(this));
pq.SetSlop(2);
//System.out.println(pq.toString("Field"));
searcher.Search(pq, new AnonymousClassHitCollector3(this));
}
/// <exception cref="ParseException">throw in overridden method to disallow
/// </exception>
protected internal virtual Query GetFieldQuery(System.String field, System.String queryText)
{
// Use the analyzer to get all the tokens, and then build a TermQuery,
// PhraseQuery, or nothing based on the term count
TokenStream source = analyzer.TokenStream(field, new System.IO.StringReader(queryText));
System.Collections.ArrayList v = System.Collections.ArrayList.Synchronized(new System.Collections.ArrayList(10));
Lucene.Net.Analysis.Token t;
int positionCount = 0;
bool severalTokensAtSamePosition = false;
while (true)
{
try
{
t = source.Next();
}
catch (System.IO.IOException e)
{
t = null;
}
if (t == null)
break;
v.Add(t);
if (t.GetPositionIncrement() != 0)
positionCount += t.GetPositionIncrement();
else
severalTokensAtSamePosition = true;
}
try
{
source.Close();
}
catch (System.IO.IOException e)
{
// ignore
}
if (v.Count == 0)
return null;
else if (v.Count == 1)
{
t = (Lucene.Net.Analysis.Token) v[0];
return new TermQuery(new Term(field, t.TermText()));
}
else
{
if (severalTokensAtSamePosition)
{
if (positionCount == 1)
{
// no phrase query:
BooleanQuery q = new BooleanQuery(true);
for (int i = 0; i < v.Count; i++)
{
t = (Lucene.Net.Analysis.Token) v[i];
TermQuery currentQuery = new TermQuery(new Term(field, t.TermText()));
q.Add(currentQuery, BooleanClause.Occur.SHOULD);
}
return q;
}
else
{
// phrase query:
MultiPhraseQuery mpq = new MultiPhraseQuery();
System.Collections.ArrayList multiTerms = new System.Collections.ArrayList();
for (int i = 0; i < v.Count; i++)
{
t = (Lucene.Net.Analysis.Token) v[i];
if (t.GetPositionIncrement() == 1 && multiTerms.Count > 0)
{
mpq.Add((Term[]) multiTerms.ToArray(typeof(Term)));
multiTerms.Clear();
}
multiTerms.Add(new Term(field, t.TermText()));
}
mpq.Add((Term[]) multiTerms.ToArray(typeof(Term)));
return mpq;
}
}
else
{
PhraseQuery q = new PhraseQuery();
q.SetSlop(phraseSlop);
for (int i = 0; i < v.Count; i++)
{
q.Add(new Term(field, ((Lucene.Net.Analysis.Token) v[i]).TermText()));
}
return q;
}
}
}
/// <summary>
/// 搜索内容
/// </summary>
/// <returns></returns>
public ActionResult SearchContent()
{
if (Request["btnCreate"] != null)
{
return Redirect("/Search/GenrateSearchLibrary");
}
string indexPath = ConfigurationManager.AppSettings["LuceneNetPath"];
string searchStr = Request["txtSearchContent"] ?? "";
if (searchStr == "")
{
return View("Index");
}
//将用户搜索的词加入热词库
SearchDetailsBll.AddEntity(new SearchDetails()
{
Id = Guid.NewGuid(),
KeyWords = searchStr,
SearchDateTime = DateTime.Now
});
List<string> stringList = Common.SearchHelper.ChangeStringToSegment(searchStr);
FSDirectory directory = FSDirectory.Open(new DirectoryInfo(indexPath), new NoLockFactory());
IndexReader reader = IndexReader.Open(directory, true);
IndexSearcher searcher = new IndexSearcher(reader);
//搜索条件
PhraseQuery query = new PhraseQuery();
foreach (string word in stringList)//先用空格，让用户去分词，空格分隔的就是词“计算机 专业”
{
query.Add(new Term("content", word));
}
//query.Add(new Term("body","语言"));--可以添加查询条件，两者是add关系.顺序没有关系.
// query.Add(new Term("body", "大学生"));
//query.Add(new Term("body", searchStr));//body中含有kw的文章
query.SetSlop(100);//多个查询条件的词之间的最大距离.在文章中相隔太远 也就无意义.（例如 “大学生”这个查询条件和"简历"这个查询条件之间如果间隔的词太多也就没有意义了。）
//TopScoreDocCollector是盛放查询结果的容器
TopScoreDocCollector collector = TopScoreDocCollector.create(1000, true);
searcher.Search(query, null, collector);//根据query查询条件进行查询，查询结果放入collector容器
ScoreDoc[] docs = collector.TopDocs(0, collector.GetTotalHits()).scoreDocs;//得到所有查询结果中的文档的ID,GetTotalHits():表示总条数 TopDocs(300, 20);//表示得到300（从300开始），到320（结束）的文档内容.
//可以用来实现分页功能
//this.listBox1.Items.Clear();
List<SearchResultViewModel> list = new List<SearchResultViewModel>();
for (int i = 0; i < docs.Length; i++)
{
//
//搜索ScoreDoc[]只能获得文档的id,这样不会把查询结果的Document一次性加载到内存中。降低了内存压力，需要获得文档的详细内容的时候通过searcher.Doc来根据文档id来获得文档的详细内容对象Document.
int docId = docs[i].doc;//得到查询结果文档的id（Lucene内部分配的id）
Document doc = searcher.Doc(docId);//找到文档id对应的文档详细信息
list.Add(new SearchResultViewModel()
{
Id = doc.Get("id"),
Title = doc.Get("title"),
Content = Common.SearchHelper.ChangeStringToHighLight(searchStr, doc.Get("content"))
});
}
ViewData["list"] = list;
ViewData["searchContent"] = searchStr;
return View("Index");
}
