C# PhraseQuery.Add方法代码示例

        public override Query VisitPhraseQuery(PhraseQuery phraseq)
        {
            _dump.Append("PhraseQ(");

            var terms = phraseq.GetTerms();
            PhraseQuery newQuery = null;

            int index = 0;
            int count = terms.Length;
            while (index < count)
            {
                var visitedTerm = VisitTerm(terms[index]);
                if (newQuery != null)
                {
                    newQuery.Add(visitedTerm);
                }
                else if (visitedTerm != terms[index])
                {
                    newQuery = new PhraseQuery();
                    for (int i = 0; i < index; i++)
                        newQuery.Add(terms[i]);
                    newQuery.Add(visitedTerm);
                }
                index++;
                if (index < count)
                    _dump.Append(", ");
            }
            _dump.Append(", Slop:").Append(phraseq.GetSlop()).Append(BoostToString(phraseq)).Append(")");
            if (newQuery != null)
                return newQuery;
            return phraseq;
        }

		public virtual void  TestPhrase()
		{
			PhraseQuery query = new PhraseQuery();
			query.Add(new Term("Field", "seventy"));
			query.Add(new Term("Field", "seven"));
			CheckHits(query, new int[]{77, 177, 277, 377, 477, 577, 677, 777, 877, 977});
		}

        public virtual void Test1()
        {
            BooleanQuery q = new BooleanQuery();

            PhraseQuery phraseQuery = new PhraseQuery();
            phraseQuery.Slop = 1;
            phraseQuery.Add(new Term(FIELD, "w1"));
            phraseQuery.Add(new Term(FIELD, "w2"));
            q.Add(phraseQuery, Occur.MUST);
            q.Add(Snear(St("w2"), Sor("w5", "zz"), 4, true), Occur.SHOULD);
            q.Add(Snear(Sf("w3", 2), St("w2"), St("w3"), 5, true), Occur.SHOULD);

            Query t = new FilteredQuery(new TermQuery(new Term(FIELD, "xx")), new ItemizedFilter(new int[] { 1, 3 }));
            t.Boost = 1000;
            q.Add(t, Occur.SHOULD);

            t = new ConstantScoreQuery(new ItemizedFilter(new int[] { 0, 2 }));
            t.Boost = 30;
            q.Add(t, Occur.SHOULD);

            DisjunctionMaxQuery dm = new DisjunctionMaxQuery(0.2f);
            dm.Add(Snear(St("w2"), Sor("w5", "zz"), 4, true));
            dm.Add(new TermQuery(new Term(FIELD, "QQ")));

            BooleanQuery xxYYZZ = new BooleanQuery();
            xxYYZZ.Add(new TermQuery(new Term(FIELD, "xx")), Occur.SHOULD);
            xxYYZZ.Add(new TermQuery(new Term(FIELD, "yy")), Occur.SHOULD);
            xxYYZZ.Add(new TermQuery(new Term(FIELD, "zz")), Occur.MUST_NOT);

            dm.Add(xxYYZZ);

            BooleanQuery xxW1 = new BooleanQuery();
            xxW1.Add(new TermQuery(new Term(FIELD, "xx")), Occur.MUST_NOT);
            xxW1.Add(new TermQuery(new Term(FIELD, "w1")), Occur.MUST_NOT);

            dm.Add(xxW1);

            DisjunctionMaxQuery dm2 = new DisjunctionMaxQuery(0.5f);
            dm2.Add(new TermQuery(new Term(FIELD, "w1")));
            dm2.Add(new TermQuery(new Term(FIELD, "w2")));
            dm2.Add(new TermQuery(new Term(FIELD, "w3")));
            dm.Add(dm2);

            q.Add(dm, Occur.SHOULD);

            BooleanQuery b = new BooleanQuery();
            b.MinimumNumberShouldMatch = 2;
            b.Add(Snear("w1", "w2", 1, true), Occur.SHOULD);
            b.Add(Snear("w2", "w3", 1, true), Occur.SHOULD);
            b.Add(Snear("w1", "w3", 3, true), Occur.SHOULD);

            q.Add(b, Occur.SHOULD);

            Qtest(q, new int[] { 0, 1, 2 });
        }

        // This is a simplified query builder which works for single Terms and single Phrases
        // Returns null, TermQuery, or PhraseQuery
        public static Lucene.Net.Search.Query GetFieldQuery(Analyzer analyzer, string field, string queryText)
        {
            TokenStream stream = analyzer.TokenStream(field, new StringReader(queryText));
            TokenFilter filter = new CachingTokenFilter(stream);
            filter.Reset();

            // This attribute way of getting token properties isn't very good, but it's the non-obsolete one.
            var attr1 = filter.GetAttribute<ITermAttribute>();
            Func<string> getText = () => attr1 != null ? attr1.Term : null;

            Func<int> getPositionIncrement;
            if (filter.HasAttribute<IPositionIncrementAttribute>())
            {
                var attr = filter.GetAttribute<IPositionIncrementAttribute>();
                getPositionIncrement = () => attr.PositionIncrement;
            }
            else
            {
                getPositionIncrement = () => 1;
            }

            // 0 tokens
            if (!filter.IncrementToken())
            {
                return new BooleanQuery();
            }

            // 1 token?
            string token1 = getText();
            int position = 0;
            if (!filter.IncrementToken())
            {
                return new TermQuery(new Term(field, token1));
            }

            // many tokens - handle first token
            PhraseQuery ret = new PhraseQuery();
            ret.Add(new Term(field, token1));

            do
            {
                // handle rest of tokens
                string tokenNext = getText();
                position += getPositionIncrement();
                ret.Add(new Term(field, tokenNext), position);
            }
            while (filter.IncrementToken());

            return ret;
        }

        public virtual void TestANDPhrase()
        {
            PhraseQuery phrase1 = new PhraseQuery();
            phrase1.Add(new Term("field", "foo"));
            phrase1.Add(new Term("field", "bar"));
            PhraseQuery phrase2 = new PhraseQuery();
            phrase2.Add(new Term("field", "star"));
            phrase2.Add(new Term("field", "wars"));
            BooleanQuery expected = new BooleanQuery();
            expected.Add(phrase1, BooleanClause.Occur.MUST);
            expected.Add(phrase2, BooleanClause.Occur.MUST);

            assertEquals(expected, Parse("\"foo bar\"+\"star wars\""));
        }

 public virtual void TestIncreasingSloppiness()
 {
     Term t1 = RandomTerm();
     Term t2 = RandomTerm();
     PhraseQuery q1 = new PhraseQuery();
     q1.Add(t1);
     q1.Add(t2);
     PhraseQuery q2 = new PhraseQuery();
     q2.Add(t1);
     q2.Add(t2);
     for (int i = 0; i < 10; i++)
     {
         q1.Slop = i;
         q2.Slop = i + 1;
         AssertSubsetOf(q1, q2);
     }
 }

        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));
        }

        protected void AddExactFieldValueClause(Index index, BooleanQuery query, string fieldName, string fieldValue)
        {
            //if (String.IsNullOrEmpty(fieldValue)) return;

            fieldValue = IdHelper.ProcessGUIDs(fieldValue);

            var phraseQuery = new PhraseQuery();
            phraseQuery.Add(new Term(fieldName.ToLowerInvariant(), fieldValue));

            query.Add(phraseQuery, BooleanClause.Occur.MUST);
        }

        /// <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 string GetSearchQuery(params ISearchableMetadataValue[] searchableMetadataValues)
        {
            var query = new PhraseQuery();

            foreach (var searchableMetadataValue in searchableMetadataValues)
            {
                query.Add(new Term(searchableMetadataValue.Metadata.SearchName, searchableMetadataValue.Value));
            }

            var filter = query.ToString();

            return GetSearchQuery(filter, new SearchableMetadata[] { });
        }

        public static Query BuildExactFieldValueClause(Index index, string fieldName, string fieldValue)
        {
            Assert.ArgumentNotNull(index, "Index");

            if (string.IsNullOrEmpty(fieldName) || string.IsNullOrEmpty(fieldValue))
            {
                return null;
            }

            fieldValue = IdHelper.ProcessGUIDs(fieldValue);

            var phraseQuery = new PhraseQuery();
            phraseQuery.Add(new Term(fieldName.ToLowerInvariant(), fieldValue.ToLowerInvariant()));

            return phraseQuery;
        }

 protected bool PhaseQueryHasHits(string[] phrases, int i)
 {
     using (var dir = FSDirectory.Open(TestEnvironment.TestIndexDirectory))
     using (var indexSearcher = new IndexSearcher(dir))
     {
         var phraseQuery = new PhraseQuery
         {
             Slop = 0
         };
         foreach (var phrase in phrases)
         {
             phraseQuery.Add(new Term("field", phrase));
         }
         // Search, without subcategories
         var topDocs = indexSearcher.Search(phraseQuery, 10);
         return topDocs.TotalHits > 0;
     }
 }

        /// <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();
        }

        public void AutomaticallyClosesDanglingQuotes()
        {
            // arrange
            var queryText = "title:\"dot NET version:1.2.3";
            var phraseQuery = new PhraseQuery();
            phraseQuery.Add(new Term("Title", "dot"));
            phraseQuery.Add(new Term("Title", "net"));
            phraseQuery.Add(new Term("Title", "version"));
            phraseQuery.Add(new Term("Title", "1"));
            phraseQuery.Add(new Term("Title", "2"));
            phraseQuery.Add(new Term("Title", "3"));

            var expected = new BooleanQuery
            {
                new BooleanClause(new BooleanQuery { new BooleanClause(new BooleanQuery { new BooleanClause(phraseQuery, Occur.SHOULD) }, Occur.SHOULD) }, Occur.MUST)
            };

            // act
            var actual = NuGetQuery.MakeQuery(queryText);

            // assert
            Assert.Equal(expected, actual);
        }

		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.Slop = 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.Slop = 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.Slop = 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.Slop = 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);
		}