C# IndexSearcher.SetSimilarity方法代码示例

		public override void SetUp()
		{
			
			base.SetUp();
			
			index = new RAMDirectory();
			IndexWriter writer = new IndexWriter(index, new WhitespaceAnalyzer(), true);
			writer.SetSimilarity(sim);
			
			// hed is the most important field, dek is secondary
			
			// d1 is an "ok" match for:  albino elephant
			{
				Lucene.Net.Documents.Document d1 = new Lucene.Net.Documents.Document();
				d1.Add(new Field("id", "d1", Field.Store.YES, Field.Index.UN_TOKENIZED)); //Field.Keyword("id", "d1"));
				d1.Add(new Field("hed", "elephant", Field.Store.YES, Field.Index.TOKENIZED)); //Field.Text("hed", "elephant"));
				d1.Add(new Field("dek", "elephant", Field.Store.YES, Field.Index.TOKENIZED)); //Field.Text("dek", "elephant"));
				writer.AddDocument(d1);
			}
			
			// d2 is a "good" match for:  albino elephant
			{
				Lucene.Net.Documents.Document d2 = new Lucene.Net.Documents.Document();
				d2.Add(new Field("id", "d2", Field.Store.YES, Field.Index.UN_TOKENIZED)); //Field.Keyword("id", "d2"));
				d2.Add(new Field("hed", "elephant", Field.Store.YES, Field.Index.TOKENIZED)); //Field.Text("hed", "elephant"));
				d2.Add(new Field("dek", "albino", Field.Store.YES, Field.Index.TOKENIZED)); //Field.Text("dek", "albino"));
				d2.Add(new Field("dek", "elephant", Field.Store.YES, Field.Index.TOKENIZED)); //Field.Text("dek", "elephant"));
				writer.AddDocument(d2);
			}
			
			// d3 is a "better" match for:  albino elephant
			{
				Lucene.Net.Documents.Document d3 = new Lucene.Net.Documents.Document();
				d3.Add(new Field("id", "d3", Field.Store.YES, Field.Index.UN_TOKENIZED)); //Field.Keyword("id", "d3"));
				d3.Add(new Field("hed", "albino", Field.Store.YES, Field.Index.TOKENIZED)); //Field.Text("hed", "albino"));
				d3.Add(new Field("hed", "elephant", Field.Store.YES, Field.Index.TOKENIZED)); //Field.Text("hed", "elephant"));
				writer.AddDocument(d3);
			}
			
			// d4 is the "best" match for:  albino elephant
			{
				Lucene.Net.Documents.Document d4 = new Lucene.Net.Documents.Document();
				d4.Add(new Field("id", "d4", Field.Store.YES, Field.Index.UN_TOKENIZED)); //Field.Keyword("id", "d4"));
				d4.Add(new Field("hed", "albino", Field.Store.YES, Field.Index.TOKENIZED)); //Field.Text("hed", "albino"));
				d4.Add(new Field("hed", "elephant", Field.Store.YES, Field.Index.TOKENIZED)); //Field.Text("hed", "elephant"));
				d4.Add(new Field("dek", "albino", Field.Store.YES, Field.Index.TOKENIZED)); //Field.Text("dek", "albino"));
				writer.AddDocument(d4);
			}
			
			writer.Close();
			
			r = IndexReader.Open(index);
			s = new IndexSearcher(r);
			s.SetSimilarity(sim);
		}

		public virtual void  TestCustomSimilarity()
		{
			RAMDirectory dir = new RAMDirectory();
			InitIndex(dir, 10, true, "x"); // documents with two tokens "doc0" and "x", "doc1" and x, etc...
			IndexSearcher srchr = new IndexSearcher(dir);
			MultiSearcher msrchr = GetMultiSearcherInstance(new Searcher[]{srchr});
			
			Similarity customSimilarity = new AnonymousClassDefaultSimilarity(this);
			
			srchr.SetSimilarity(customSimilarity);
			msrchr.SetSimilarity(customSimilarity);
			
			Query query = new TermQuery(new Term("contents", "doc0"));
			
			// Get a score from IndexSearcher
			TopDocs topDocs = srchr.Search(query, null, 1);
			float score1 = topDocs.GetMaxScore();
			
			// Get the score from MultiSearcher
			topDocs = msrchr.Search(query, null, 1);
			float scoreN = topDocs.GetMaxScore();
			
			// The scores from the IndexSearcher and Multisearcher should be the same
			// if the same similarity is used.
			Assert.AreEqual(score1, scoreN, 1e-6, "MultiSearcher score must be equal to single esrcher score!");
		}

		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));
		}

		/// <summary> Given an IndexSearcher, returns a new IndexSearcher whose IndexReader 
		/// is a MultiReader containing the Reader of the original IndexSearcher, 
		/// as well as several "empty" IndexReaders -- some of which will have 
		/// deleted documents in them.  This new IndexSearcher should 
		/// behave exactly the same as the original IndexSearcher.
		/// </summary>
		/// <param name="s">the searcher to wrap
		/// </param>
		/// <param name="edge">if negative, s will be the first sub; if 0, s will be in the middle, if positive s will be the last sub
		/// </param>
		public static IndexSearcher WrapUnderlyingReader(IndexSearcher s, int edge)
		{
			
			IndexReader r = s.GetIndexReader();
			
			// we can't put deleted docs before the nested reader, because
			// it will throw off the docIds
            IndexReader[] readers = new IndexReader[] {
                edge < 0 ? r : IndexReader.Open(MakeEmptyIndex(0), true),
                IndexReader.Open(MakeEmptyIndex(0), true),
                new MultiReader(new IndexReader[] {
                    IndexReader.Open(MakeEmptyIndex(edge < 0 ? 4 : 0), true),
                    IndexReader.Open(MakeEmptyIndex(0), true),
                    0 == edge ? r : IndexReader.Open(MakeEmptyIndex(0), true)
                }),
                IndexReader.Open(MakeEmptyIndex(0 < edge ? 0 : 7), true),
                IndexReader.Open(MakeEmptyIndex(0), true),
                new MultiReader(new IndexReader[] {
                    IndexReader.Open(MakeEmptyIndex(0 < edge ? 0 : 5), true),
                    IndexReader.Open(MakeEmptyIndex(0), true),
                    0 < edge ? r : IndexReader.Open(MakeEmptyIndex(0), true)
                })
            };

			IndexSearcher out_Renamed = new IndexSearcher(new MultiReader(readers));
			out_Renamed.SetSimilarity(s.GetSimilarity());
			return out_Renamed;
		}

		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 override void  SetUp()
		{
			base.SetUp();
			RAMDirectory directory = new RAMDirectory();
			PayloadAnalyzer analyzer = new PayloadAnalyzer(this);
			IndexWriter writer = new IndexWriter(directory, analyzer, true);
			writer.SetSimilarity(similarity);
			//writer.infoStream = System.out;
			for (int i = 0; i < 1000; i++)
			{
				Document doc = new Document();
				Field noPayloadField = new Field("noPayLoad", English.IntToEnglish(i), Field.Store.YES, Field.Index.TOKENIZED);
				noPayloadField.SetBoost(0);
				doc.Add(noPayloadField);
				doc.Add(new Field("field", English.IntToEnglish(i), Field.Store.YES, Field.Index.TOKENIZED));
				doc.Add(new Field("multiField", English.IntToEnglish(i) + "  " + English.IntToEnglish(i), Field.Store.YES, Field.Index.TOKENIZED));
				writer.AddDocument(doc);
			}
			writer.Optimize();
			writer.Close();
			
			searcher = new IndexSearcher(directory);
			searcher.SetSimilarity(similarity);
		}

		public IndexSearcher GetSearcher(string indexName)
		{
			IndexSearcher ret;
			if (indexSearchers.TryGetValue(indexName, out ret) && ret != null)
				return ret;

			var indexPath = Path.Combine(IndexesStoragePath, indexName);
			if (!Directory.Exists(indexPath))
				return null;

			ret = new IndexSearcher(FSDirectory.Open(new DirectoryInfo(indexPath)), true);
			ret.SetSimilarity(new BinaryCoordSimilarity()); // to boost docs containing all terms, to overcome issues
																// concerning many terms in index produced by HebMorph

			if (!indexSearchers.TryAdd(indexName, ret))
			{
				ret.Close();
				return null;
			}

			return ret;
		}