C# IndexReader.GetTermFreqVectors方法代码示例

		private void  CopyVectorsNoDeletions(TermVectorsWriter termVectorsWriter, TermVectorsReader matchingVectorsReader, IndexReader reader)
		{
			int maxDoc = reader.MaxDoc();
			if (matchingVectorsReader != null)
			{
				// We can bulk-copy because the fieldInfos are "congruent"
				int docCount = 0;
				while (docCount < maxDoc)
				{
					int len = System.Math.Min(MAX_RAW_MERGE_DOCS, maxDoc - docCount);
					matchingVectorsReader.RawDocs(rawDocLengths, rawDocLengths2, docCount, len);
					termVectorsWriter.AddRawDocuments(matchingVectorsReader, rawDocLengths, rawDocLengths2, len);
					docCount += len;
					checkAbort.Work(300 * len);
				}
			}
			else
			{
				for (int docNum = 0; docNum < maxDoc; docNum++)
				{
					// NOTE: it's very important to first assign to vectors then pass it to
					// termVectorsWriter.addAllDocVectors; see LUCENE-1282
					TermFreqVector[] vectors = reader.GetTermFreqVectors(docNum);
					termVectorsWriter.AddAllDocVectors(vectors);
					checkAbort.Work(300);
				}
			}
		}

		public static void  VerifyEquals(IndexReader r1, IndexReader r2, System.String idField)
		{
			Assert.AreEqual(r1.NumDocs(), r2.NumDocs());
			bool hasDeletes = !(r1.MaxDoc() == r2.MaxDoc() && r1.NumDocs() == r1.MaxDoc());
			
			int[] r2r1 = new int[r2.MaxDoc()]; // r2 id to r1 id mapping
			
			TermDocs termDocs1 = r1.TermDocs();
			TermDocs termDocs2 = r2.TermDocs();
			
			// create mapping from id2 space to id2 based on idField
			idField = StringHelper.Intern(idField);
			TermEnum termEnum = r1.Terms(new Term(idField, ""));
			do 
			{
				Term term = termEnum.Term();
				if (term == null || (System.Object) term.Field() != (System.Object) idField)
					break;
				
				termDocs1.Seek(termEnum);
				if (!termDocs1.Next())
				{
					// This doc is deleted and wasn't replaced
					termDocs2.Seek(termEnum);
					Assert.IsFalse(termDocs2.Next());
					continue;
				}
				
				int id1 = termDocs1.Doc();
				Assert.IsFalse(termDocs1.Next());
				
				termDocs2.Seek(termEnum);
				Assert.IsTrue(termDocs2.Next());
				int id2 = termDocs2.Doc();
				Assert.IsFalse(termDocs2.Next());
				
				r2r1[id2] = id1;
				
				// verify stored fields are equivalent
				try
				{
					VerifyEquals(r1.Document(id1), r2.Document(id2));
				}
				catch (System.Exception t)
				{
					System.Console.Out.WriteLine("FAILED id=" + term + " id1=" + id1 + " id2=" + id2 + " term=" + term);
					System.Console.Out.WriteLine("  d1=" + r1.Document(id1));
					System.Console.Out.WriteLine("  d2=" + r2.Document(id2));
					throw t;
				}
				
				try
				{
					// verify term vectors are equivalent        
					VerifyEquals(r1.GetTermFreqVectors(id1), r2.GetTermFreqVectors(id2));
				}
				catch (System.Exception e)
				{
					System.Console.Out.WriteLine("FAILED id=" + term + " id1=" + id1 + " id2=" + id2);
					TermFreqVector[] tv1 = r1.GetTermFreqVectors(id1);
					System.Console.Out.WriteLine("  d1=" + tv1);
					if (tv1 != null)
						for (int i = 0; i < tv1.Length; i++)
						{
							System.Console.Out.WriteLine("    " + i + ": " + tv1[i]);
						}
					
					TermFreqVector[] tv2 = r2.GetTermFreqVectors(id2);
					System.Console.Out.WriteLine("  d2=" + tv2);
					if (tv2 != null)
						for (int i = 0; i < tv2.Length; i++)
						{
							System.Console.Out.WriteLine("    " + i + ": " + tv2[i]);
						}
					
					throw e;
				}
			}
			while (termEnum.Next());
			
			termEnum.Close();
			
			// Verify postings
			TermEnum termEnum1 = r1.Terms(new Term("", ""));
			TermEnum termEnum2 = r2.Terms(new Term("", ""));
			
			// pack both doc and freq into single element for easy sorting
			long[] info1 = new long[r1.NumDocs()];
			long[] info2 = new long[r2.NumDocs()];
			
			for (; ; )
			{
				Term term1, term2;
				
				// iterate until we get some docs
				int len1;
				for (; ; )
				{
					len1 = 0;
					term1 = termEnum1.Term();
//.........这里部分代码省略.........

		private void  CopyVectorsWithDeletions(TermVectorsWriter termVectorsWriter, TermVectorsReader matchingVectorsReader, IndexReader reader)
		{
			int maxDoc = reader.MaxDoc();
			if (matchingVectorsReader != null)
			{
				// We can bulk-copy because the fieldInfos are "congruent"
				for (int docNum = 0; docNum < maxDoc; )
				{
					if (reader.IsDeleted(docNum))
					{
						// skip deleted docs
						++docNum;
						continue;
					}
					// We can optimize this case (doing a bulk byte copy) since the field 
					// numbers are identical
					int start = docNum, numDocs = 0;
					do 
					{
						docNum++;
						numDocs++;
						if (docNum >= maxDoc)
							break;
						if (reader.IsDeleted(docNum))
						{
							docNum++;
							break;
						}
					}
					while (numDocs < MAX_RAW_MERGE_DOCS);
					
					matchingVectorsReader.RawDocs(rawDocLengths, rawDocLengths2, start, numDocs);
					termVectorsWriter.AddRawDocuments(matchingVectorsReader, rawDocLengths, rawDocLengths2, numDocs);
					checkAbort.Work(300 * numDocs);
				}
			}
			else
			{
				for (int docNum = 0; docNum < maxDoc; docNum++)
				{
					if (reader.IsDeleted(docNum))
					{
						// skip deleted docs
						continue;
					}
					
					// NOTE: it's very important to first assign to vectors then pass it to
					// termVectorsWriter.addAllDocVectors; see LUCENE-1282
					TermFreqVector[] vectors = reader.GetTermFreqVectors(docNum);
					termVectorsWriter.AddAllDocVectors(vectors);
					checkAbort.Work(300);
				}
			}
		}

		private void  TestHits(ScoreDoc[] hits, int expectedCount, IndexReader reader)
		{
			int hitCount = hits.Length;
			Assert.AreEqual(expectedCount, hitCount, "wrong number of hits");
			for (int i = 0; i < hitCount; i++)
			{
				reader.Document(hits[i].doc);
				reader.GetTermFreqVectors(hits[i].doc);
			}
		}