// builds an index with payloads in the given Directory and performs
// different tests to verify the payload encoding
private void PerformTest(Directory dir)
{
PayloadAnalyzer analyzer = new PayloadAnalyzer();
IndexWriter writer = new IndexWriter(dir, NewIndexWriterConfig(TEST_VERSION_CURRENT, analyzer).SetOpenMode(OpenMode.CREATE).SetMergePolicy(NewLogMergePolicy()));
// should be in sync with value in TermInfosWriter
const int skipInterval = 16;
const int numTerms = 5;
const string fieldName = "f1";
int numDocs = skipInterval + 1;
// create content for the test documents with just a few terms
Term[] terms = GenerateTerms(fieldName, numTerms);
StringBuilder sb = new StringBuilder();
for (int i = 0; i < terms.Length; i++)
{
sb.Append(terms[i].Text());
sb.Append(" ");
}
string content = sb.ToString();
int payloadDataLength = numTerms * numDocs * 2 + numTerms * numDocs * (numDocs - 1) / 2;
var payloadData = GenerateRandomData(payloadDataLength);
Document d = new Document();
d.Add(NewTextField(fieldName, content, Field.Store.NO));
// add the same document multiple times to have the same payload lengths for all
// occurrences within two consecutive skip intervals
int offset = 0;
for (int i = 0; i < 2 * numDocs; i++)
{
analyzer = new PayloadAnalyzer(fieldName, payloadData, offset, 1);
offset += numTerms;
writer.AddDocument(d, analyzer);
}
// make sure we create more than one segment to test merging
writer.Commit();
// now we make sure to have different payload lengths next at the next skip point
for (int i = 0; i < numDocs; i++)
{
analyzer = new PayloadAnalyzer(fieldName, payloadData, offset, i);
offset += i * numTerms;
writer.AddDocument(d, analyzer);
}
writer.ForceMerge(1);
// flush
writer.Dispose();
/*
* Verify the index
* first we test if all payloads are stored correctly
*/
IndexReader reader = DirectoryReader.Open(dir);
var verifyPayloadData = new byte[payloadDataLength];
offset = 0;
var tps = new DocsAndPositionsEnum[numTerms];
for (int i = 0; i < numTerms; i++)
{
tps[i] = MultiFields.GetTermPositionsEnum(reader, MultiFields.GetLiveDocs(reader), terms[i].Field, new BytesRef(terms[i].Text()));
}
while (tps[0].NextDoc() != DocIdSetIterator.NO_MORE_DOCS)
{
for (int i = 1; i < numTerms; i++)
{
tps[i].NextDoc();
}
int freq = tps[0].Freq;
for (int i = 0; i < freq; i++)
{
for (int j = 0; j < numTerms; j++)
{
tps[j].NextPosition();
BytesRef br = tps[j].GetPayload();
if (br != null)
{
Array.Copy(br.Bytes, br.Offset, verifyPayloadData, offset, br.Length);
offset += br.Length;
}
}
}
}
AssertByteArrayEquals(payloadData, verifyPayloadData);
/*
* test lazy skipping
*/
DocsAndPositionsEnum tp = MultiFields.GetTermPositionsEnum(reader, MultiFields.GetLiveDocs(reader), terms[0].Field, new BytesRef(terms[0].Text()));
tp.NextDoc();
tp.NextPosition();
// NOTE: prior rev of this test was failing to first
// call next here:
tp.NextDoc();
// now we don't read this payload
tp.NextPosition();
BytesRef payload = tp.GetPayload();
Assert.AreEqual(1, payload.Length, "Wrong payload length.");
Assert.AreEqual(payload.Bytes[payload.Offset], payloadData[numTerms]);
tp.NextDoc();
tp.NextPosition();
// we don't read this payload and skip to a different document
tp.Advance(5);
tp.NextPosition();
payload = tp.GetPayload();
Assert.AreEqual(1, payload.Length, "Wrong payload length.");
Assert.AreEqual(payload.Bytes[payload.Offset], payloadData[5 * numTerms]);
/*
* Test different lengths at skip points
*/
tp = MultiFields.GetTermPositionsEnum(reader, MultiFields.GetLiveDocs(reader), terms[1].Field, new BytesRef(terms[1].Text()));
tp.NextDoc();
tp.NextPosition();
Assert.AreEqual(1, tp.GetPayload().Length, "Wrong payload length.");
tp.Advance(skipInterval - 1);
tp.NextPosition();
Assert.AreEqual(1, tp.GetPayload().Length, "Wrong payload length.");
tp.Advance(2 * skipInterval - 1);
tp.NextPosition();
Assert.AreEqual(1, tp.GetPayload().Length, "Wrong payload length.");
tp.Advance(3 * skipInterval - 1);
tp.NextPosition();
Assert.AreEqual(3 * skipInterval - 2 * numDocs - 1, tp.GetPayload().Length, "Wrong payload length.");
reader.Dispose();
// test long payload
analyzer = new PayloadAnalyzer();
writer = new IndexWriter(dir, NewIndexWriterConfig(TEST_VERSION_CURRENT, analyzer).SetOpenMode(OpenMode.CREATE));
string singleTerm = "lucene";
d = new Document();
d.Add(NewTextField(fieldName, singleTerm, Field.Store.NO));
// add a payload whose length is greater than the buffer size of BufferedIndexOutput
payloadData = GenerateRandomData(2000);
analyzer.SetPayloadData(fieldName, payloadData, 100, 1500);
writer.AddDocument(d);
writer.ForceMerge(1);
// flush
writer.Dispose();
reader = DirectoryReader.Open(dir);
tp = MultiFields.GetTermPositionsEnum(reader, MultiFields.GetLiveDocs(reader), fieldName, new BytesRef(singleTerm));
tp.NextDoc();
tp.NextPosition();
BytesRef bref = tp.GetPayload();
verifyPayloadData = new byte[bref.Length];
var portion = new byte[1500];
Array.Copy(payloadData, 100, portion, 0, 1500);
AssertByteArrayEquals(portion, bref.Bytes, bref.Offset, bref.Length);
reader.Dispose();
}
public virtual void TestRandom()
{
// token -> docID -> tokens
IDictionary <string, IDictionary <int?, IList <Token> > > actualTokens = new Dictionary <string, IDictionary <int?, IList <Token> > >();
Directory dir = NewDirectory();
RandomIndexWriter w = new RandomIndexWriter(Random(), dir, Iwc);
int numDocs = AtLeast(20);
//final int numDocs = AtLeast(5);
FieldType ft = new FieldType(TextField.TYPE_NOT_STORED);
// TODO: randomize what IndexOptions we use; also test
// changing this up in one IW buffered segment...:
ft.IndexOptions = IndexOptions.DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS;
if (Random().NextBoolean())
{
ft.StoreTermVectors = true;
ft.StoreTermVectorOffsets = Random().NextBoolean();
ft.StoreTermVectorPositions = Random().NextBoolean();
}
for (int docCount = 0; docCount < numDocs; docCount++)
{
Document doc = new Document();
doc.Add(new Int32Field("id", docCount, Field.Store.NO));
IList <Token> tokens = new List <Token>();
int numTokens = AtLeast(100);
//final int numTokens = AtLeast(20);
int pos = -1;
int offset = 0;
//System.out.println("doc id=" + docCount);
for (int tokenCount = 0; tokenCount < numTokens; tokenCount++)
{
string text;
if (Random().NextBoolean())
{
text = "a";
}
else if (Random().NextBoolean())
{
text = "b";
}
else if (Random().NextBoolean())
{
text = "c";
}
else
{
text = "d";
}
int posIncr = Random().NextBoolean() ? 1 : Random().Next(5);
if (tokenCount == 0 && posIncr == 0)
{
posIncr = 1;
}
int offIncr = Random().NextBoolean() ? 0 : Random().Next(5);
int tokenOffset = Random().Next(5);
Token token = MakeToken(text, posIncr, offset + offIncr, offset + offIncr + tokenOffset);
if (!actualTokens.ContainsKey(text))
{
actualTokens[text] = new Dictionary <int?, IList <Token> >();
}
IDictionary <int?, IList <Token> > postingsByDoc = actualTokens[text];
if (!postingsByDoc.ContainsKey(docCount))
{
postingsByDoc[docCount] = new List <Token>();
}
postingsByDoc[docCount].Add(token);
tokens.Add(token);
pos += posIncr;
// stuff abs position into type:
token.Type = "" + pos;
offset += offIncr + tokenOffset;
//System.out.println(" " + token + " posIncr=" + token.getPositionIncrement() + " pos=" + pos + " off=" + token.StartOffset + "/" + token.EndOffset + " (freq=" + postingsByDoc.Get(docCount).Size() + ")");
}
doc.Add(new Field("content", new CannedTokenStream(tokens.ToArray()), ft));
w.AddDocument(doc);
}
DirectoryReader r = w.Reader;
w.Dispose();
string[] terms = new string[] { "a", "b", "c", "d" };
foreach (AtomicReaderContext ctx in r.Leaves)
{
// TODO: improve this
AtomicReader sub = (AtomicReader)ctx.Reader;
//System.out.println("\nsub=" + sub);
TermsEnum termsEnum = sub.Fields.GetTerms("content").GetIterator(null);
DocsEnum docs = null;
DocsAndPositionsEnum docsAndPositions = null;
DocsAndPositionsEnum docsAndPositionsAndOffsets = null;
FieldCache.Int32s docIDToID = FieldCache.DEFAULT.GetInt32s(sub, "id", false);
foreach (string term in terms)
{
//System.out.println(" term=" + term);
if (termsEnum.SeekExact(new BytesRef(term)))
{
docs = termsEnum.Docs(null, docs);
Assert.IsNotNull(docs);
int doc;
//System.out.println(" doc/freq");
while ((doc = docs.NextDoc()) != DocIdSetIterator.NO_MORE_DOCS)
{
IList <Token> expected = actualTokens[term][docIDToID.Get(doc)];
//System.out.println(" doc=" + docIDToID.Get(doc) + " docID=" + doc + " " + expected.Size() + " freq");
Assert.IsNotNull(expected);
Assert.AreEqual(expected.Count, docs.Freq);
}
// explicitly exclude offsets here
docsAndPositions = termsEnum.DocsAndPositions(null, docsAndPositions, DocsAndPositionsFlags.PAYLOADS);
Assert.IsNotNull(docsAndPositions);
//System.out.println(" doc/freq/pos");
while ((doc = docsAndPositions.NextDoc()) != DocIdSetIterator.NO_MORE_DOCS)
{
IList <Token> expected = actualTokens[term][docIDToID.Get(doc)];
//System.out.println(" doc=" + docIDToID.Get(doc) + " " + expected.Size() + " freq");
Assert.IsNotNull(expected);
Assert.AreEqual(expected.Count, docsAndPositions.Freq);
foreach (Token token in expected)
{
int pos = Convert.ToInt32(token.Type);
//System.out.println(" pos=" + pos);
Assert.AreEqual(pos, docsAndPositions.NextPosition());
}
}
docsAndPositionsAndOffsets = termsEnum.DocsAndPositions(null, docsAndPositions);
Assert.IsNotNull(docsAndPositionsAndOffsets);
//System.out.println(" doc/freq/pos/offs");
while ((doc = docsAndPositionsAndOffsets.NextDoc()) != DocIdSetIterator.NO_MORE_DOCS)
{
IList <Token> expected = actualTokens[term][docIDToID.Get(doc)];
//System.out.println(" doc=" + docIDToID.Get(doc) + " " + expected.Size() + " freq");
Assert.IsNotNull(expected);
Assert.AreEqual(expected.Count, docsAndPositionsAndOffsets.Freq);
foreach (Token token in expected)
{
int pos = Convert.ToInt32(token.Type);
//System.out.println(" pos=" + pos);
Assert.AreEqual(pos, docsAndPositionsAndOffsets.NextPosition());
Assert.AreEqual(token.StartOffset, docsAndPositionsAndOffsets.StartOffset);
Assert.AreEqual(token.EndOffset, docsAndPositionsAndOffsets.EndOffset);
}
}
}
}
// TODO: test advance:
}
r.Dispose();
dir.Dispose();
}
public override int NextDoc()
{
return(m_input.NextDoc());
}
public virtual void TestLongPostings_Mem()
{
// Don't use TestUtil.getTempDir so that we own the
// randomness (ie same seed will point to same dir):
Directory dir = NewFSDirectory(CreateTempDir("longpostings" + "." + Random().NextLong()));
int NUM_DOCS = AtLeast(2000);
if (VERBOSE)
{
Console.WriteLine("TEST: NUM_DOCS=" + NUM_DOCS);
}
string s1 = GetRandomTerm(null);
string s2 = GetRandomTerm(s1);
if (VERBOSE)
{
Console.WriteLine("\nTEST: s1=" + s1 + " s2=" + s2);
/*
* for(int idx=0;idx<s1.Length();idx++) {
* System.out.println(" s1 ch=0x" + Integer.toHexString(s1.charAt(idx)));
* }
* for(int idx=0;idx<s2.Length();idx++) {
* System.out.println(" s2 ch=0x" + Integer.toHexString(s2.charAt(idx)));
* }
*/
}
FixedBitSet isS1 = new FixedBitSet(NUM_DOCS);
for (int idx = 0; idx < NUM_DOCS; idx++)
{
if (Random().NextBoolean())
{
isS1.Set(idx);
}
}
IndexReader r;
IndexWriterConfig iwc = NewIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(Random())).SetOpenMode(IndexWriterConfig.OpenMode_e.CREATE).SetMergePolicy(NewLogMergePolicy());
iwc.SetRAMBufferSizeMB(16.0 + 16.0 * Random().NextDouble());
iwc.SetMaxBufferedDocs(-1);
RandomIndexWriter riw = new RandomIndexWriter(Random(), dir, iwc);
for (int idx = 0; idx < NUM_DOCS; idx++)
{
Document doc = new Document();
string s = isS1.Get(idx) ? s1 : s2;
Field f = NewTextField("field", s, Field.Store.NO);
int count = TestUtil.NextInt(Random(), 1, 4);
for (int ct = 0; ct < count; ct++)
{
doc.Add(f);
}
riw.AddDocument(doc);
}
r = riw.Reader;
riw.Dispose();
/*
* if (VERBOSE) {
* System.out.println("TEST: terms");
* TermEnum termEnum = r.Terms();
* while(termEnum.Next()) {
* System.out.println(" term=" + termEnum.Term() + " len=" + termEnum.Term().Text().Length());
* Assert.IsTrue(termEnum.DocFreq() > 0);
* System.out.println(" s1?=" + (termEnum.Term().Text().equals(s1)) + " s1len=" + s1.Length());
* System.out.println(" s2?=" + (termEnum.Term().Text().equals(s2)) + " s2len=" + s2.Length());
* final String s = termEnum.Term().Text();
* for(int idx=0;idx<s.Length();idx++) {
* System.out.println(" ch=0x" + Integer.toHexString(s.charAt(idx)));
* }
* }
* }
*/
Assert.AreEqual(NUM_DOCS, r.NumDocs);
Assert.IsTrue(r.DocFreq(new Term("field", s1)) > 0);
Assert.IsTrue(r.DocFreq(new Term("field", s2)) > 0);
int num = AtLeast(1000);
for (int iter = 0; iter < num; iter++)
{
string term;
bool doS1;
if (Random().NextBoolean())
{
term = s1;
doS1 = true;
}
else
{
term = s2;
doS1 = false;
}
if (VERBOSE)
{
Console.WriteLine("\nTEST: iter=" + iter + " doS1=" + doS1);
}
DocsAndPositionsEnum postings = MultiFields.GetTermPositionsEnum(r, null, "field", new BytesRef(term));
int docID = -1;
while (docID < DocIdSetIterator.NO_MORE_DOCS)
{
int what = Random().Next(3);
if (what == 0)
{
if (VERBOSE)
{
Console.WriteLine("TEST: docID=" + docID + "; do next()");
}
// nextDoc
int expected = docID + 1;
while (true)
{
if (expected == NUM_DOCS)
{
expected = int.MaxValue;
break;
}
else if (isS1.Get(expected) == doS1)
{
break;
}
else
{
expected++;
}
}
docID = postings.NextDoc();
if (VERBOSE)
{
Console.WriteLine(" got docID=" + docID);
}
Assert.AreEqual(expected, docID);
if (docID == DocIdSetIterator.NO_MORE_DOCS)
{
break;
}
if (Random().Next(6) == 3)
{
int freq = postings.Freq();
Assert.IsTrue(freq >= 1 && freq <= 4);
for (int pos = 0; pos < freq; pos++)
{
Assert.AreEqual(pos, postings.NextPosition());
if (Random().NextBoolean())
{
var dummy = postings.Payload;
if (Random().NextBoolean())
{
dummy = postings.Payload; // get it again
}
}
}
}
}
else
{
// advance
int targetDocID;
if (docID == -1)
{
targetDocID = Random().Next(NUM_DOCS + 1);
}
else
{
targetDocID = docID + TestUtil.NextInt(Random(), 1, NUM_DOCS - docID);
}
if (VERBOSE)
{
Console.WriteLine("TEST: docID=" + docID + "; do advance(" + targetDocID + ")");
}
int expected = targetDocID;
while (true)
{
if (expected == NUM_DOCS)
{
expected = int.MaxValue;
break;
}
else if (isS1.Get(expected) == doS1)
{
break;
}
else
{
expected++;
}
}
docID = postings.Advance(targetDocID);
if (VERBOSE)
{
Console.WriteLine(" got docID=" + docID);
}
Assert.AreEqual(expected, docID);
if (docID == DocIdSetIterator.NO_MORE_DOCS)
{
break;
}
if (Random().Next(6) == 3)
{
int freq = postings.Freq();
Assert.IsTrue(freq >= 1 && freq <= 4);
for (int pos = 0; pos < freq; pos++)
{
Assert.AreEqual(pos, postings.NextPosition());
if (Random().NextBoolean())
{
var dummy = postings.Payload;
if (Random().NextBoolean())
{
dummy = postings.Payload; // get it again
}
}
}
}
}
}
}
r.Dispose();
dir.Dispose();
}
public virtual void DoTestNumbers(bool withPayloads)
{
Directory dir = NewDirectory();
Analyzer analyzer = withPayloads ? (Analyzer) new MockPayloadAnalyzer() : new MockAnalyzer(Random());
Iwc = NewIndexWriterConfig(TEST_VERSION_CURRENT, analyzer);
Iwc.SetMergePolicy(NewLogMergePolicy()); // will rely on docids a bit for skipping
RandomIndexWriter w = new RandomIndexWriter(Random(), dir, Iwc);
FieldType ft = new FieldType(TextField.TYPE_STORED);
ft.IndexOptions = IndexOptions.DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS;
if (Random().NextBoolean())
{
ft.StoreTermVectors = true;
ft.StoreTermVectorOffsets = Random().NextBoolean();
ft.StoreTermVectorPositions = Random().NextBoolean();
}
int numDocs = AtLeast(500);
for (int i = 0; i < numDocs; i++)
{
Document doc = new Document();
doc.Add(new Field("numbers", English.IntToEnglish(i), ft));
doc.Add(new Field("oddeven", (i % 2) == 0 ? "even" : "odd", ft));
doc.Add(new StringField("id", "" + i, Field.Store.NO));
w.AddDocument(doc);
}
IndexReader reader = w.Reader;
w.Dispose();
string[] terms = new string[] { "one", "two", "three", "four", "five", "six", "seven", "eight", "nine", "ten", "hundred" };
foreach (string term in terms)
{
DocsAndPositionsEnum dp = MultiFields.GetTermPositionsEnum(reader, null, "numbers", new BytesRef(term));
int doc;
while ((doc = dp.NextDoc()) != DocIdSetIterator.NO_MORE_DOCS)
{
string storedNumbers = reader.Document(doc).Get("numbers");
int freq = dp.Freq;
for (int i = 0; i < freq; i++)
{
dp.NextPosition();
int start = dp.StartOffset;
Debug.Assert(start >= 0);
int end = dp.EndOffset;
Debug.Assert(end >= 0 && end >= start);
// check that the offsets correspond to the term in the src text
Assert.IsTrue(storedNumbers.Substring(start, end - start).Equals(term, StringComparison.Ordinal));
if (withPayloads)
{
// check that we have a payload and it starts with "pos"
Assert.IsNotNull(dp.GetPayload());
BytesRef payload = dp.GetPayload();
Assert.IsTrue(payload.Utf8ToString().StartsWith("pos:", StringComparison.Ordinal));
} // note: withPayloads=false doesnt necessarily mean we dont have them from MockAnalyzer!
}
}
}
// check we can skip correctly
int numSkippingTests = AtLeast(50);
for (int j = 0; j < numSkippingTests; j++)
{
int num = TestUtil.NextInt(Random(), 100, Math.Min(numDocs - 1, 999));
DocsAndPositionsEnum dp = MultiFields.GetTermPositionsEnum(reader, null, "numbers", new BytesRef("hundred"));
int doc = dp.Advance(num);
Assert.AreEqual(num, doc);
int freq = dp.Freq;
for (int i = 0; i < freq; i++)
{
string storedNumbers = reader.Document(doc).Get("numbers");
dp.NextPosition();
int start = dp.StartOffset;
Debug.Assert(start >= 0);
int end = dp.EndOffset;
Debug.Assert(end >= 0 && end >= start);
// check that the offsets correspond to the term in the src text
Assert.IsTrue(storedNumbers.Substring(start, end - start).Equals("hundred", StringComparison.Ordinal));
if (withPayloads)
{
// check that we have a payload and it starts with "pos"
Assert.IsNotNull(dp.GetPayload());
BytesRef payload = dp.GetPayload();
Assert.IsTrue(payload.Utf8ToString().StartsWith("pos:", StringComparison.Ordinal));
} // note: withPayloads=false doesnt necessarily mean we dont have them from MockAnalyzer!
}
}
// check that other fields (without offsets) work correctly
for (int i = 0; i < numDocs; i++)
{
DocsEnum dp = MultiFields.GetTermDocsEnum(reader, null, "id", new BytesRef("" + i), 0);
Assert.AreEqual(i, dp.NextDoc());
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, dp.NextDoc());
}
reader.Dispose();
dir.Dispose();
}
/// <summary>
/// checks docs + freqs + positions + payloads, sequentially
/// </summary>
public void AssertDocsAndPositionsEnumEquals(string info, DocsAndPositionsEnum leftDocs, DocsAndPositionsEnum rightDocs)
{
if (leftDocs == null || rightDocs == null)
{
Assert.IsNull(leftDocs);
Assert.IsNull(rightDocs);
return;
}
Assert.AreEqual(-1, leftDocs.DocID(), info);
Assert.AreEqual(-1, rightDocs.DocID(), info);
int docid;
while ((docid = leftDocs.NextDoc()) != DocIdSetIterator.NO_MORE_DOCS)
{
Assert.AreEqual(docid, rightDocs.NextDoc(), info);
int freq = leftDocs.Freq();
Assert.AreEqual(freq, rightDocs.Freq(), info);
for (int i = 0; i < freq; i++)
{
Assert.AreEqual(leftDocs.NextPosition(), rightDocs.NextPosition(), info);
Assert.AreEqual(leftDocs.Payload, rightDocs.Payload, info);
Assert.AreEqual(leftDocs.StartOffset(), rightDocs.StartOffset(), info);
Assert.AreEqual(leftDocs.EndOffset(), rightDocs.EndOffset(), info);
}
}
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, rightDocs.NextDoc(), info);
}
public virtual void TestPositionReader()
{
TermVectorsReader reader = Codec.Default.TermVectorsFormat().VectorsReader(Dir, Seg.Info, FieldInfos, NewIOContext(Random()));
BytesRef[] terms;
Terms vector = reader.Get(0).Terms(TestFields[0]);
Assert.IsNotNull(vector);
Assert.AreEqual(TestTerms.Length, vector.Size());
TermsEnum termsEnum = vector.Iterator(null);
DocsAndPositionsEnum dpEnum = null;
for (int i = 0; i < TestTerms.Length; i++)
{
BytesRef text = termsEnum.Next();
Assert.IsNotNull(text);
string term = text.Utf8ToString();
//System.out.println("Term: " + term);
Assert.AreEqual(TestTerms[i], term);
dpEnum = termsEnum.DocsAndPositions(null, dpEnum);
Assert.IsNotNull(dpEnum);
int doc = dpEnum.DocID();
Assert.AreEqual(-1, doc);
Assert.IsTrue(dpEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
Assert.AreEqual(dpEnum.Freq(), Positions[i].Length);
for (int j = 0; j < Positions[i].Length; j++)
{
Assert.AreEqual(Positions[i][j], dpEnum.NextPosition());
}
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, dpEnum.NextDoc());
dpEnum = termsEnum.DocsAndPositions(null, dpEnum);
doc = dpEnum.DocID();
Assert.AreEqual(-1, doc);
Assert.IsTrue(dpEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
Assert.IsNotNull(dpEnum);
Assert.AreEqual(dpEnum.Freq(), Positions[i].Length);
for (int j = 0; j < Positions[i].Length; j++)
{
Assert.AreEqual(Positions[i][j], dpEnum.NextPosition());
Assert.AreEqual(j * 10, dpEnum.StartOffset());
Assert.AreEqual(j * 10 + TestTerms[i].Length, dpEnum.EndOffset());
}
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, dpEnum.NextDoc());
}
Terms freqVector = reader.Get(0).Terms(TestFields[1]); //no pos, no offset
Assert.IsNotNull(freqVector);
Assert.AreEqual(TestTerms.Length, freqVector.Size());
termsEnum = freqVector.Iterator(null);
Assert.IsNotNull(termsEnum);
for (int i = 0; i < TestTerms.Length; i++)
{
BytesRef text = termsEnum.Next();
Assert.IsNotNull(text);
string term = text.Utf8ToString();
//System.out.println("Term: " + term);
Assert.AreEqual(TestTerms[i], term);
Assert.IsNotNull(termsEnum.Docs(null, null));
Assert.IsNull(termsEnum.DocsAndPositions(null, null)); // no pos
}
reader.Dispose();
}
public override int NextDoc()
{
return(@in.NextDoc());
}
public virtual void VerifyEquals(DirectoryReader r1, DirectoryReader r2, string idField)
{
if (Verbose)
{
Console.WriteLine("\nr1 docs:");
PrintDocs(r1);
Console.WriteLine("\nr2 docs:");
PrintDocs(r2);
}
if (r1.NumDocs != r2.NumDocs)
{
if (Debugging.AssertsEnabled)
{
Debugging.Assert(false, "r1.NumDocs={0} vs r2.NumDocs={1}", 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
// create mapping from id2 space to id2 based on idField
Fields f1 = MultiFields.GetFields(r1);
if (f1 == null)
{
// make sure r2 is empty
Assert.IsNull(MultiFields.GetFields(r2));
return;
}
Terms terms1 = f1.GetTerms(idField);
if (terms1 == null)
{
Assert.IsTrue(MultiFields.GetFields(r2) == null || MultiFields.GetFields(r2).GetTerms(idField) == null);
return;
}
TermsEnum termsEnum = terms1.GetEnumerator();
IBits liveDocs1 = MultiFields.GetLiveDocs(r1);
IBits liveDocs2 = MultiFields.GetLiveDocs(r2);
Fields fields = MultiFields.GetFields(r2);
if (fields == null)
{
// make sure r1 is in fact empty (eg has only all
// deleted docs):
IBits liveDocs = MultiFields.GetLiveDocs(r1);
DocsEnum docs = null;
while (termsEnum.MoveNext())
{
docs = TestUtil.Docs(Random, termsEnum, liveDocs, docs, DocsFlags.NONE);
while (docs.NextDoc() != DocIdSetIterator.NO_MORE_DOCS)
{
Assert.Fail("r1 is not empty but r2 is");
}
}
return;
}
Terms terms2 = fields.GetTerms(idField);
TermsEnum termsEnum2 = terms2.GetEnumerator();
DocsEnum termDocs1 = null;
DocsEnum termDocs2 = null;
while (termsEnum.MoveNext())
{
BytesRef term = termsEnum.Term;
//System.out.println("TEST: match id term=" + term);
termDocs1 = TestUtil.Docs(Random, termsEnum, liveDocs1, termDocs1, DocsFlags.NONE);
if (termsEnum2.SeekExact(term))
{
termDocs2 = TestUtil.Docs(Random, termsEnum2, liveDocs2, termDocs2, DocsFlags.NONE);
}
else
{
termDocs2 = null;
}
if (termDocs1.NextDoc() == DocIdSetIterator.NO_MORE_DOCS)
{
// this doc is deleted and wasn't replaced
Assert.IsTrue(termDocs2 == null || termDocs2.NextDoc() == DocIdSetIterator.NO_MORE_DOCS);
continue;
}
int id1 = termDocs1.DocID;
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, termDocs1.NextDoc());
Assert.IsTrue(termDocs2.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
int id2 = termDocs2.DocID;
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, termDocs2.NextDoc());
r2r1[id2] = id1;
// verify stored fields are equivalent
try
{
VerifyEquals(r1.Document(id1), r2.Document(id2));
}
catch (Exception /*t*/)
{
Console.WriteLine("FAILED id=" + term + " id1=" + id1 + " id2=" + id2 + " term=" + term);
Console.WriteLine(" d1=" + r1.Document(id1));
Console.WriteLine(" d2=" + r2.Document(id2));
throw; // LUCENENET: CA2200: Rethrow to preserve stack details (https://docs.microsoft.com/en-us/visualstudio/code-quality/ca2200-rethrow-to-preserve-stack-details)
}
try
{
// verify term vectors are equivalent
VerifyEquals(r1.GetTermVectors(id1), r2.GetTermVectors(id2));
}
catch (Exception /*e*/)
{
Console.WriteLine("FAILED id=" + term + " id1=" + id1 + " id2=" + id2);
Fields tv1 = r1.GetTermVectors(id1);
Console.WriteLine(" d1=" + tv1);
if (tv1 != null)
{
DocsAndPositionsEnum dpEnum = null;
DocsEnum dEnum = null;
foreach (string field in tv1)
{
Console.WriteLine(" " + field + ":");
Terms terms3 = tv1.GetTerms(field);
Assert.IsNotNull(terms3);
TermsEnum termsEnum3 = terms3.GetEnumerator();
while (termsEnum3.MoveNext())
{
Console.WriteLine(" " + termsEnum3.Term.Utf8ToString() + ": freq=" + termsEnum3.TotalTermFreq);
dpEnum = termsEnum3.DocsAndPositions(null, dpEnum);
if (dpEnum != null)
{
Assert.IsTrue(dpEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
int freq = dpEnum.Freq;
Console.WriteLine(" doc=" + dpEnum.DocID + " freq=" + freq);
for (int posUpto = 0; posUpto < freq; posUpto++)
{
Console.WriteLine(" pos=" + dpEnum.NextPosition());
}
}
else
{
dEnum = TestUtil.Docs(Random, termsEnum3, null, dEnum, DocsFlags.FREQS);
Assert.IsNotNull(dEnum);
Assert.IsTrue(dEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
int freq = dEnum.Freq;
Console.WriteLine(" doc=" + dEnum.DocID + " freq=" + freq);
}
}
}
}
Fields tv2 = r2.GetTermVectors(id2);
Console.WriteLine(" d2=" + tv2);
if (tv2 != null)
{
DocsAndPositionsEnum dpEnum = null;
DocsEnum dEnum = null;
foreach (string field in tv2)
{
Console.WriteLine(" " + field + ":");
Terms terms3 = tv2.GetTerms(field);
Assert.IsNotNull(terms3);
TermsEnum termsEnum3 = terms3.GetEnumerator();
while (termsEnum3.MoveNext())
{
Console.WriteLine(" " + termsEnum3.Term.Utf8ToString() + ": freq=" + termsEnum3.TotalTermFreq);
dpEnum = termsEnum3.DocsAndPositions(null, dpEnum);
if (dpEnum != null)
{
Assert.IsTrue(dpEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
int freq = dpEnum.Freq;
Console.WriteLine(" doc=" + dpEnum.DocID + " freq=" + freq);
for (int posUpto = 0; posUpto < freq; posUpto++)
{
Console.WriteLine(" pos=" + dpEnum.NextPosition());
}
}
else
{
dEnum = TestUtil.Docs(Random, termsEnum3, null, dEnum, DocsFlags.FREQS);
Assert.IsNotNull(dEnum);
Assert.IsTrue(dEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
int freq = dEnum.Freq;
Console.WriteLine(" doc=" + dEnum.DocID + " freq=" + freq);
}
}
}
}
throw; // LUCENENET: CA2200: Rethrow to preserve stack details (https://docs.microsoft.com/en-us/visualstudio/code-quality/ca2200-rethrow-to-preserve-stack-details)
}
}
//System.out.println("TEST: done match id");
// Verify postings
//System.out.println("TEST: create te1");
Fields fields1 = MultiFields.GetFields(r1);
IEnumerator <string> fields1Enum = fields1.GetEnumerator();
Fields fields2 = MultiFields.GetFields(r2);
IEnumerator <string> fields2Enum = fields2.GetEnumerator();
string field1 = null, field2 = null;
TermsEnum termsEnum1 = null;
termsEnum2 = null;
DocsEnum docs1 = null, docs2 = null;
// pack both doc and freq into single element for easy sorting
long[] info1 = new long[r1.NumDocs];
long[] info2 = new long[r2.NumDocs];
for (; ;)
{
BytesRef term1 = null, term2 = null;
// iterate until we get some docs
int len1;
for (; ;)
{
len1 = 0;
if (termsEnum1 == null)
{
if (!fields1Enum.MoveNext())
{
break;
}
field1 = fields1Enum.Current;
Terms terms = fields1.GetTerms(field1);
if (terms == null)
{
continue;
}
termsEnum1 = terms.GetEnumerator();
}
if (!termsEnum1.MoveNext())
{
term1 = null;
// no more terms in this field
termsEnum1 = null;
continue;
}
term1 = termsEnum1.Term;
//System.out.println("TEST: term1=" + term1);
docs1 = TestUtil.Docs(Random, termsEnum1, liveDocs1, docs1, DocsFlags.FREQS);
while (docs1.NextDoc() != DocIdSetIterator.NO_MORE_DOCS)
{
int d = docs1.DocID;
int f = docs1.Freq;
info1[len1] = (((long)d) << 32) | (uint)f;
len1++;
}
if (len1 > 0)
{
break;
}
}
// iterate until we get some docs
int len2;
for (; ;)
{
len2 = 0;
if (termsEnum2 == null)
{
if (!fields2Enum.MoveNext())
{
break;
}
field2 = fields2Enum.Current;
Terms terms = fields2.GetTerms(field2);
if (terms == null)
{
continue;
}
termsEnum2 = terms.GetEnumerator();
}
if (!termsEnum2.MoveNext())
{
term2 = null;
// no more terms in this field
termsEnum2 = null;
continue;
}
term2 = termsEnum2.Term;
//System.out.println("TEST: term1=" + term1);
docs2 = TestUtil.Docs(Random, termsEnum2, liveDocs2, docs2, DocsFlags.FREQS);
while (docs2.NextDoc() != DocIdSetIterator.NO_MORE_DOCS)
{
int d = r2r1[docs2.DocID];
int f = docs2.Freq;
info2[len2] = (((long)d) << 32) | (uint)f;
len2++;
}
if (len2 > 0)
{
break;
}
}
Assert.AreEqual(len1, len2);
if (len1 == 0) // no more terms
{
break;
}
Assert.AreEqual(field1, field2);
Assert.IsTrue(term1.BytesEquals(term2));
if (!hasDeletes)
{
Assert.AreEqual(termsEnum1.DocFreq, termsEnum2.DocFreq);
}
Assert.AreEqual(term1, term2, "len1=" + len1 + " len2=" + len2 + " deletes?=" + hasDeletes);
// sort info2 to get it into ascending docid
Array.Sort(info2, 0, len2);
// now compare
for (int i = 0; i < len1; i++)
{
Assert.AreEqual(info1[i], info2[i], "i=" + i + " len=" + len1 + " d1=" + (info1[i].TripleShift(32)) + " f1=" + (info1[i] & int.MaxValue) + " d2=" + (info2[i].TripleShift(32)) + " f2=" + (info2[i] & int.MaxValue) + " field=" + field1 + " term=" + term1.Utf8ToString());
}
}
}
public virtual void TestDoubleOffsetCounting()
{
Directory dir = NewDirectory();
IndexWriter w = new IndexWriter(dir, NewIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(Random())));
Document doc = new Document();
FieldType customType = new FieldType(StringField.TYPE_NOT_STORED);
customType.StoreTermVectors = true;
customType.StoreTermVectorPositions = true;
customType.StoreTermVectorOffsets = true;
Field f = NewField("field", "abcd", customType);
doc.Add(f);
doc.Add(f);
Field f2 = NewField("field", "", customType);
doc.Add(f2);
doc.Add(f);
w.AddDocument(doc);
w.Dispose();
IndexReader r = DirectoryReader.Open(dir);
Terms vector = r.GetTermVectors(0).Terms("field");
Assert.IsNotNull(vector);
TermsEnum termsEnum = vector.Iterator(null);
Assert.IsNotNull(termsEnum.Next());
Assert.AreEqual("", termsEnum.Term().Utf8ToString());
// Token "" occurred once
Assert.AreEqual(1, termsEnum.TotalTermFreq());
DocsAndPositionsEnum dpEnum = termsEnum.DocsAndPositions(null, null);
Assert.IsTrue(dpEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
dpEnum.NextPosition();
Assert.AreEqual(8, dpEnum.StartOffset());
Assert.AreEqual(8, dpEnum.EndOffset());
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, dpEnum.NextDoc());
// Token "abcd" occurred three times
Assert.AreEqual(new BytesRef("abcd"), termsEnum.Next());
dpEnum = termsEnum.DocsAndPositions(null, dpEnum);
Assert.AreEqual(3, termsEnum.TotalTermFreq());
Assert.IsTrue(dpEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
dpEnum.NextPosition();
Assert.AreEqual(0, dpEnum.StartOffset());
Assert.AreEqual(4, dpEnum.EndOffset());
dpEnum.NextPosition();
Assert.AreEqual(4, dpEnum.StartOffset());
Assert.AreEqual(8, dpEnum.EndOffset());
dpEnum.NextPosition();
Assert.AreEqual(8, dpEnum.StartOffset());
Assert.AreEqual(12, dpEnum.EndOffset());
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, dpEnum.NextDoc());
Assert.IsNull(termsEnum.Next());
r.Dispose();
dir.Dispose();
}
protected internal virtual void AssertEquals(RandomTokenStream tk, FieldType ft, Terms terms)
{
Assert.AreEqual(1, terms.DocCount);
int termCount = (new HashSet <string>(Arrays.AsList(tk.Terms))).Count;
Assert.AreEqual(termCount, terms.Size());
Assert.AreEqual(termCount, terms.SumDocFreq);
Assert.AreEqual(ft.StoreTermVectorPositions, terms.HasPositions());
Assert.AreEqual(ft.StoreTermVectorOffsets, terms.HasOffsets());
Assert.AreEqual(ft.StoreTermVectorPayloads && tk.HasPayloads(), terms.HasPayloads());
HashSet <BytesRef> uniqueTerms = new HashSet <BytesRef>();
foreach (string term in tk.Freqs.Keys)
{
uniqueTerms.Add(new BytesRef(term));
}
BytesRef[] sortedTerms = uniqueTerms.ToArray(/*new BytesRef[0]*/);
Array.Sort(sortedTerms, terms.Comparator);
TermsEnum termsEnum = terms.Iterator(Random().NextBoolean() ? null : this.termsEnum.Value);
this.termsEnum.Value = termsEnum;
for (int i = 0; i < sortedTerms.Length; ++i)
{
BytesRef nextTerm = termsEnum.Next();
Assert.AreEqual(sortedTerms[i], nextTerm);
Assert.AreEqual(sortedTerms[i], termsEnum.Term());
Assert.AreEqual(1, termsEnum.DocFreq());
FixedBitSet bits = new FixedBitSet(1);
DocsEnum docsEnum = termsEnum.Docs(bits, Random().NextBoolean() ? null : this.docsEnum.Value);
Assert.AreEqual(DocsEnum.NO_MORE_DOCS, docsEnum.NextDoc());
bits.Set(0);
docsEnum = termsEnum.Docs(Random().NextBoolean() ? bits : null, Random().NextBoolean() ? null : docsEnum);
Assert.IsNotNull(docsEnum);
Assert.AreEqual(0, docsEnum.NextDoc());
Assert.AreEqual(0, docsEnum.DocID());
Assert.AreEqual(tk.Freqs[termsEnum.Term().Utf8ToString()], (int?)docsEnum.Freq());
Assert.AreEqual(DocsEnum.NO_MORE_DOCS, docsEnum.NextDoc());
this.docsEnum.Value = docsEnum;
bits.Clear(0);
DocsAndPositionsEnum docsAndPositionsEnum = termsEnum.DocsAndPositions(bits, Random().NextBoolean() ? null : this.docsAndPositionsEnum.Value);
Assert.AreEqual(ft.StoreTermVectorOffsets || ft.StoreTermVectorPositions, docsAndPositionsEnum != null);
if (docsAndPositionsEnum != null)
{
Assert.AreEqual(DocsEnum.NO_MORE_DOCS, docsAndPositionsEnum.NextDoc());
}
bits.Set(0);
docsAndPositionsEnum = termsEnum.DocsAndPositions(Random().NextBoolean() ? bits : null, Random().NextBoolean() ? null : docsAndPositionsEnum);
Assert.AreEqual(ft.StoreTermVectorOffsets || ft.StoreTermVectorPositions, docsAndPositionsEnum != null);
if (terms.HasPositions() || terms.HasOffsets())
{
Assert.AreEqual(0, docsAndPositionsEnum.NextDoc());
int freq = docsAndPositionsEnum.Freq();
Assert.AreEqual(tk.Freqs[termsEnum.Term().Utf8ToString()], (int?)freq);
if (docsAndPositionsEnum != null)
{
for (int k = 0; k < freq; ++k)
{
int position = docsAndPositionsEnum.NextPosition();
ISet <int?> indexes;
if (terms.HasPositions())
{
indexes = tk.PositionToTerms[position];
Assert.IsNotNull(indexes);
}
else
{
indexes = tk.StartOffsetToTerms[docsAndPositionsEnum.StartOffset()];
Assert.IsNotNull(indexes);
}
if (terms.HasPositions())
{
bool foundPosition = false;
foreach (int index in indexes)
{
if (tk.TermBytes[index].Equals(termsEnum.Term()) && tk.Positions[index] == position)
{
foundPosition = true;
break;
}
}
Assert.IsTrue(foundPosition);
}
if (terms.HasOffsets())
{
bool foundOffset = false;
foreach (int index in indexes)
{
if (tk.TermBytes[index].Equals(termsEnum.Term()) && tk.StartOffsets[index] == docsAndPositionsEnum.StartOffset() && tk.EndOffsets[index] == docsAndPositionsEnum.EndOffset())
{
foundOffset = true;
break;
}
}
Assert.IsTrue(foundOffset);
}
if (terms.HasPayloads())
{
bool foundPayload = false;
foreach (int index in indexes)
{
if (tk.TermBytes[index].Equals(termsEnum.Term()) && Equals(tk.Payloads[index], docsAndPositionsEnum.Payload))
{
foundPayload = true;
break;
}
}
Assert.IsTrue(foundPayload);
}
}
try
{
docsAndPositionsEnum.NextPosition();
Assert.Fail();
}
catch (Exception e)
{
// ok
}
}
Assert.AreEqual(DocsEnum.NO_MORE_DOCS, docsAndPositionsEnum.NextDoc());
}
this.docsAndPositionsEnum.Value = docsAndPositionsEnum;
}
Assert.IsNull(termsEnum.Next());
for (int i = 0; i < 5; ++i)
{
if (Random().NextBoolean())
{
Assert.IsTrue(termsEnum.SeekExact(RandomInts.RandomFrom(Random(), tk.TermBytes)));
}
else
{
Assert.AreEqual(SeekStatus.FOUND, termsEnum.SeekCeil(RandomInts.RandomFrom(Random(), tk.TermBytes)));
}
}
}
public virtual void TestLargeNumberOfPositions()
{
Directory dir = NewDirectory();
RandomIndexWriter writer = new RandomIndexWriter(Random, dir, NewIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(Random)));
int howMany = 1000;
FieldType customType = new FieldType(TextField.TYPE_NOT_STORED);
customType.OmitNorms = true;
for (int i = 0; i < 39; i++)
{
Document doc = new Document();
StringBuilder builder = new StringBuilder();
for (int j = 0; j < howMany; j++)
{
if (j % 2 == 0)
{
builder.Append("even ");
}
else
{
builder.Append("odd ");
}
}
doc.Add(NewField(fieldName, builder.ToString(), customType));
writer.AddDocument(doc);
}
// now do searches
IndexReader reader = writer.GetReader();
writer.Dispose();
int num = AtLeast(13);
for (int i = 0; i < num; i++)
{
BytesRef bytes = new BytesRef("even");
IndexReaderContext topReaderContext = reader.Context;
foreach (AtomicReaderContext atomicReaderContext in topReaderContext.Leaves)
{
DocsAndPositionsEnum docsAndPosEnum = GetDocsAndPositions((AtomicReader)atomicReaderContext.Reader, bytes, null);
Assert.IsNotNull(docsAndPosEnum);
int initDoc = 0;
int maxDoc = atomicReaderContext.Reader.MaxDoc;
// initially advance or do next doc
if (Random.NextBoolean())
{
initDoc = docsAndPosEnum.NextDoc();
}
else
{
initDoc = docsAndPosEnum.Advance(Random.Next(maxDoc));
}
string msg = "Iteration: " + i + " initDoc: " + initDoc; // TODO: + " payloads: " + usePayload;
Assert.AreEqual(howMany / 2, docsAndPosEnum.Freq);
for (int j = 0; j < howMany; j += 2)
{
Assert.AreEqual(j, docsAndPosEnum.NextPosition(), "position missmatch index: " + j + " with freq: " + docsAndPosEnum.Freq + " -- " + msg);
}
}
}
reader.Dispose();
dir.Dispose();
}
public virtual void TestRandomPositions()
{
Directory dir = NewDirectory();
RandomIndexWriter writer = new RandomIndexWriter(Random, dir, NewIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(Random)).SetMergePolicy(NewLogMergePolicy()));
int numDocs = AtLeast(47);
int max = 1051;
int term = Random.Next(max);
int?[][] positionsInDoc = new int?[numDocs][];
FieldType customType = new FieldType(TextField.TYPE_NOT_STORED);
customType.OmitNorms = true;
for (int i = 0; i < numDocs; i++)
{
Document doc = new Document();
List <int?> positions = new List <int?>();
StringBuilder builder = new StringBuilder();
int num = AtLeast(131);
for (int j = 0; j < num; j++)
{
int nextInt = Random.Next(max);
builder.Append(nextInt).Append(" ");
if (nextInt == term)
{
positions.Add(Convert.ToInt32(j));
}
}
if (positions.Count == 0)
{
builder.Append(term);
positions.Add(num);
}
doc.Add(NewField(fieldName, builder.ToString(), customType));
positionsInDoc[i] = positions.ToArray();
writer.AddDocument(doc);
}
IndexReader reader = writer.GetReader();
writer.Dispose();
int num_ = AtLeast(13);
for (int i = 0; i < num_; i++)
{
BytesRef bytes = new BytesRef("" + term);
IndexReaderContext topReaderContext = reader.Context;
foreach (AtomicReaderContext atomicReaderContext in topReaderContext.Leaves)
{
DocsAndPositionsEnum docsAndPosEnum = GetDocsAndPositions((AtomicReader)atomicReaderContext.Reader, bytes, null);
Assert.IsNotNull(docsAndPosEnum);
int initDoc = 0;
int maxDoc = atomicReaderContext.Reader.MaxDoc;
// initially advance or do next doc
if (Random.NextBoolean())
{
initDoc = docsAndPosEnum.NextDoc();
}
else
{
initDoc = docsAndPosEnum.Advance(Random.Next(maxDoc));
}
// now run through the scorer and check if all positions are there...
do
{
int docID = docsAndPosEnum.DocID;
if (docID == DocIdSetIterator.NO_MORE_DOCS)
{
break;
}
int?[] pos = positionsInDoc[atomicReaderContext.DocBase + docID];
Assert.AreEqual(pos.Length, docsAndPosEnum.Freq);
// number of positions read should be random - don't read all of them
// allways
int howMany = Random.Next(20) == 0 ? pos.Length - Random.Next(pos.Length) : pos.Length;
for (int j = 0; j < howMany; j++)
{
Assert.AreEqual(pos[j], docsAndPosEnum.NextPosition(), "iteration: " + i + " initDoc: " + initDoc + " doc: " + docID + " base: " + atomicReaderContext.DocBase + " positions: " + pos); /* TODO: + " usePayloads: "
+ usePayload*/
}
if (Random.Next(10) == 0) // once is a while advance
{
if (docsAndPosEnum.Advance(docID + 1 + Random.Next((maxDoc - docID))) == DocIdSetIterator.NO_MORE_DOCS)
{
break;
}
}
} while (docsAndPosEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
}
}
reader.Dispose();
dir.Dispose();
}
public virtual void TestArbitraryFields()
{
Directory dir = NewDirectory();
RandomIndexWriter w = new RandomIndexWriter(
#if FEATURE_INSTANCE_TESTDATA_INITIALIZATION
this,
#endif
Random, dir);
int NUM_DOCS = AtLeast(27);
if (VERBOSE)
{
Console.WriteLine("TEST: " + NUM_DOCS + " docs");
}
int[] fieldsPerDoc = new int[NUM_DOCS];
int baseCount = 0;
for (int docCount = 0; docCount < NUM_DOCS; docCount++)
{
int fieldCount = TestUtil.NextInt32(Random, 1, 17);
fieldsPerDoc[docCount] = fieldCount - 1;
int finalDocCount = docCount;
if (VERBOSE)
{
Console.WriteLine("TEST: " + fieldCount + " fields in doc " + docCount);
}
int finalBaseCount = baseCount;
baseCount += fieldCount - 1;
w.AddDocument(new IterableAnonymousInnerClassHelper(this, fieldCount, finalDocCount, finalBaseCount));
}
IndexReader r = w.GetReader();
w.Dispose();
IndexSearcher s = NewSearcher(r);
int counter = 0;
for (int id = 0; id < NUM_DOCS; id++)
{
if (VERBOSE)
{
Console.WriteLine("TEST: verify doc id=" + id + " (" + fieldsPerDoc[id] + " fields) counter=" + counter);
}
TopDocs hits = s.Search(new TermQuery(new Term("id", "" + id)), 1);
Assert.AreEqual(1, hits.TotalHits);
int docID = hits.ScoreDocs[0].Doc;
Document doc = s.Doc(docID);
int endCounter = counter + fieldsPerDoc[id];
while (counter < endCounter)
{
string name = "f" + counter;
int fieldID = counter % 10;
bool stored = (counter & 1) == 0 || fieldID == 3;
bool binary = fieldID == 3;
bool indexed = fieldID != 3;
string stringValue;
if (fieldID != 3 && fieldID != 9)
{
stringValue = "text " + counter;
}
else
{
stringValue = null;
}
// stored:
if (stored)
{
IIndexableField f = doc.GetField(name);
Assert.IsNotNull(f, "doc " + id + " doesn't have field f" + counter);
if (binary)
{
Assert.IsNotNull(f, "doc " + id + " doesn't have field f" + counter);
BytesRef b = f.GetBinaryValue();
Assert.IsNotNull(b);
Assert.AreEqual(10, b.Length);
for (int idx = 0; idx < 10; idx++)
{
Assert.AreEqual((byte)(idx + counter), b.Bytes[b.Offset + idx]);
}
}
else
{
Debug.Assert(stringValue != null);
Assert.AreEqual(stringValue, f.GetStringValue());
}
}
if (indexed)
{
bool tv = counter % 2 == 1 && fieldID != 9;
if (tv)
{
Terms tfv = r.GetTermVectors(docID).GetTerms(name);
Assert.IsNotNull(tfv);
TermsEnum termsEnum = tfv.GetIterator(null);
Assert.AreEqual(new BytesRef("" + counter), termsEnum.Next());
Assert.AreEqual(1, termsEnum.TotalTermFreq);
DocsAndPositionsEnum dpEnum = termsEnum.DocsAndPositions(null, null);
Assert.IsTrue(dpEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
Assert.AreEqual(1, dpEnum.Freq);
Assert.AreEqual(1, dpEnum.NextPosition());
Assert.AreEqual(new BytesRef("text"), termsEnum.Next());
Assert.AreEqual(1, termsEnum.TotalTermFreq);
dpEnum = termsEnum.DocsAndPositions(null, dpEnum);
Assert.IsTrue(dpEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
Assert.AreEqual(1, dpEnum.Freq);
Assert.AreEqual(0, dpEnum.NextPosition());
Assert.IsNull(termsEnum.Next());
// TODO: offsets
}
else
{
Fields vectors = r.GetTermVectors(docID);
Assert.IsTrue(vectors == null || vectors.GetTerms(name) == null);
}
BooleanQuery bq = new BooleanQuery();
bq.Add(new TermQuery(new Term("id", "" + id)), Occur.MUST);
bq.Add(new TermQuery(new Term(name, "text")), Occur.MUST);
TopDocs hits2 = s.Search(bq, 1);
Assert.AreEqual(1, hits2.TotalHits);
Assert.AreEqual(docID, hits2.ScoreDocs[0].Doc);
bq = new BooleanQuery();
bq.Add(new TermQuery(new Term("id", "" + id)), Occur.MUST);
bq.Add(new TermQuery(new Term(name, "" + counter)), Occur.MUST);
TopDocs hits3 = s.Search(bq, 1);
Assert.AreEqual(1, hits3.TotalHits);
Assert.AreEqual(docID, hits3.ScoreDocs[0].Doc);
}
counter++;
}
}
r.Dispose();
dir.Dispose();
}
/// <summary>
/// checks docs + freqs + positions + payloads, sequentially
/// </summary>
public virtual void AssertDocsAndPositionsEnum(DocsAndPositionsEnum leftDocs, DocsAndPositionsEnum rightDocs)
{
if (leftDocs == null || rightDocs == null)
{
Assert.IsNull(leftDocs);
Assert.IsNull(rightDocs);
return;
}
Assert.AreEqual(-1, leftDocs.DocID());
Assert.AreEqual(-1, rightDocs.DocID());
int docid;
while ((docid = leftDocs.NextDoc()) != DocIdSetIterator.NO_MORE_DOCS)
{
Assert.AreEqual(docid, rightDocs.NextDoc());
int freq = leftDocs.Freq();
Assert.AreEqual(freq, rightDocs.Freq());
for (int i = 0; i < freq; i++)
{
Assert.AreEqual(leftDocs.NextPosition(), rightDocs.NextPosition());
// we don't assert offsets/payloads, they are allowed to be different
}
}
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, rightDocs.NextDoc());
}
public static void VerifyEquals(Fields d1, Fields d2)
{
if (d1 == null)
{
Assert.IsTrue(d2 == null || d2.Count == 0);
return;
}
Assert.IsTrue(d2 != null);
IEnumerator <string> fieldsEnum2 = d2.GetEnumerator();
foreach (string field1 in d1)
{
fieldsEnum2.MoveNext();
string field2 = fieldsEnum2.Current;
Assert.AreEqual(field1, field2);
Terms terms1 = d1.GetTerms(field1);
Assert.IsNotNull(terms1);
TermsEnum termsEnum1 = terms1.GetEnumerator();
Terms terms2 = d2.GetTerms(field2);
Assert.IsNotNull(terms2);
TermsEnum termsEnum2 = terms2.GetEnumerator();
DocsAndPositionsEnum dpEnum1 = null;
DocsAndPositionsEnum dpEnum2 = null;
DocsEnum dEnum1 = null;
DocsEnum dEnum2 = null;
BytesRef term1;
while (termsEnum1.MoveNext())
{
term1 = termsEnum1.Term;
termsEnum2.MoveNext();
BytesRef term2 = termsEnum2.Term;
Assert.AreEqual(term1, term2);
Assert.AreEqual(termsEnum1.TotalTermFreq, termsEnum2.TotalTermFreq);
dpEnum1 = termsEnum1.DocsAndPositions(null, dpEnum1);
dpEnum2 = termsEnum2.DocsAndPositions(null, dpEnum2);
if (dpEnum1 != null)
{
Assert.IsNotNull(dpEnum2);
int docID1 = dpEnum1.NextDoc();
dpEnum2.NextDoc();
// docIDs are not supposed to be equal
//int docID2 = dpEnum2.NextDoc();
//Assert.AreEqual(docID1, docID2);
Assert.IsTrue(docID1 != DocIdSetIterator.NO_MORE_DOCS);
int freq1 = dpEnum1.Freq;
int freq2 = dpEnum2.Freq;
Assert.AreEqual(freq1, freq2);
IOffsetAttribute offsetAtt1 = dpEnum1.Attributes.HasAttribute <IOffsetAttribute>() ? dpEnum1.Attributes.GetAttribute <IOffsetAttribute>() : null;
IOffsetAttribute offsetAtt2 = dpEnum2.Attributes.HasAttribute <IOffsetAttribute>() ? dpEnum2.Attributes.GetAttribute <IOffsetAttribute>() : null;
if (offsetAtt1 != null)
{
Assert.IsNotNull(offsetAtt2);
}
else
{
Assert.IsNull(offsetAtt2);
}
for (int posUpto = 0; posUpto < freq1; posUpto++)
{
int pos1 = dpEnum1.NextPosition();
int pos2 = dpEnum2.NextPosition();
Assert.AreEqual(pos1, pos2);
if (offsetAtt1 != null)
{
Assert.AreEqual(offsetAtt1.StartOffset, offsetAtt2.StartOffset);
Assert.AreEqual(offsetAtt1.EndOffset, offsetAtt2.EndOffset);
}
}
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, dpEnum1.NextDoc());
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, dpEnum2.NextDoc());
}
else
{
dEnum1 = TestUtil.Docs(Random, termsEnum1, null, dEnum1, DocsFlags.FREQS);
dEnum2 = TestUtil.Docs(Random, termsEnum2, null, dEnum2, DocsFlags.FREQS);
Assert.IsNotNull(dEnum1);
Assert.IsNotNull(dEnum2);
int docID1 = dEnum1.NextDoc();
dEnum2.NextDoc();
// docIDs are not supposed to be equal
//int docID2 = dEnum2.NextDoc();
//Assert.AreEqual(docID1, docID2);
Assert.IsTrue(docID1 != DocIdSetIterator.NO_MORE_DOCS);
int freq1 = dEnum1.Freq;
int freq2 = dEnum2.Freq;
Assert.AreEqual(freq1, freq2);
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, dEnum1.NextDoc());
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, dEnum2.NextDoc());
}
}
Assert.IsFalse(termsEnum2.MoveNext());
}
Assert.IsFalse(fieldsEnum2.MoveNext());
}
/// <summary>
/// checks advancing docs + positions
/// </summary>
public virtual void AssertPositionsSkipping(int docFreq, DocsAndPositionsEnum leftDocs, DocsAndPositionsEnum rightDocs)
{
if (leftDocs == null || rightDocs == null)
{
Assert.IsNull(leftDocs);
Assert.IsNull(rightDocs);
return;
}
int docid = -1;
int averageGap = MAXDOC / (1 + docFreq);
int skipInterval = 16;
while (true)
{
if (Random().NextBoolean())
{
// nextDoc()
docid = leftDocs.NextDoc();
Assert.AreEqual(docid, rightDocs.NextDoc());
}
else
{
// advance()
int skip = docid + (int)Math.Ceiling(Math.Abs(skipInterval + Random().NextDouble() * averageGap));
docid = leftDocs.Advance(skip);
Assert.AreEqual(docid, rightDocs.Advance(skip));
}
if (docid == DocIdSetIterator.NO_MORE_DOCS)
{
return;
}
int freq = leftDocs.Freq();
Assert.AreEqual(freq, rightDocs.Freq());
for (int i = 0; i < freq; i++)
{
Assert.AreEqual(leftDocs.NextPosition(), rightDocs.NextPosition());
// we don't compare the payloads, its allowed that one is empty etc
}
}
}
public virtual void TestPositionReader()
{
TermVectorsReader reader = Codec.Default.TermVectorsFormat.VectorsReader(dir, seg.Info, fieldInfos, NewIOContext(Random));
//BytesRef[] terms; // LUCENENET NOTE: Not used in Lucene
Terms vector = reader.Get(0).GetTerms(testFields[0]);
Assert.IsNotNull(vector);
Assert.AreEqual(testTerms.Length, vector.Count);
TermsEnum termsEnum = vector.GetEnumerator();
DocsAndPositionsEnum dpEnum = null;
for (int i = 0; i < testTerms.Length; i++)
{
Assert.IsTrue(termsEnum.MoveNext());
BytesRef text = termsEnum.Term;
string term = text.Utf8ToString();
//System.out.println("Term: " + term);
Assert.AreEqual(testTerms[i], term);
dpEnum = termsEnum.DocsAndPositions(null, dpEnum);
Assert.IsNotNull(dpEnum);
int doc = dpEnum.DocID;
Assert.AreEqual(-1, doc);
Assert.IsTrue(dpEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
Assert.AreEqual(dpEnum.Freq, positions[i].Length);
for (int j = 0; j < positions[i].Length; j++)
{
Assert.AreEqual(positions[i][j], dpEnum.NextPosition());
}
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, dpEnum.NextDoc());
dpEnum = termsEnum.DocsAndPositions(null, dpEnum);
doc = dpEnum.DocID;
Assert.AreEqual(-1, doc);
Assert.IsTrue(dpEnum.NextDoc() != DocIdSetIterator.NO_MORE_DOCS);
Assert.IsNotNull(dpEnum);
Assert.AreEqual(dpEnum.Freq, positions[i].Length);
for (int j = 0; j < positions[i].Length; j++)
{
Assert.AreEqual(positions[i][j], dpEnum.NextPosition());
Assert.AreEqual(j * 10, dpEnum.StartOffset);
Assert.AreEqual(j * 10 + testTerms[i].Length, dpEnum.EndOffset);
}
Assert.AreEqual(DocIdSetIterator.NO_MORE_DOCS, dpEnum.NextDoc());
}
Terms freqVector = reader.Get(0).GetTerms(testFields[1]); //no pos, no offset
Assert.IsNotNull(freqVector);
Assert.AreEqual(testTerms.Length, freqVector.Count);
termsEnum = freqVector.GetEnumerator();
Assert.IsNotNull(termsEnum);
for (int i = 0; i < testTerms.Length; i++)
{
Assert.IsTrue(termsEnum.MoveNext());
BytesRef text = termsEnum.Term;
string term = text.Utf8ToString();
//System.out.println("Term: " + term);
Assert.AreEqual(testTerms[i], term);
Assert.IsNotNull(termsEnum.Docs(null, null));
Assert.IsNull(termsEnum.DocsAndPositions(null, null)); // no pos
}
reader.Dispose();
}
/// <summary>
/// checks advancing docs + positions
/// </summary>
public void AssertPositionsSkippingEquals(string info, IndexReader leftReader, int docFreq, DocsAndPositionsEnum leftDocs, DocsAndPositionsEnum rightDocs)
{
if (leftDocs == null || rightDocs == null)
{
Assert.IsNull(leftDocs);
Assert.IsNull(rightDocs);
return;
}
int docid = -1;
int averageGap = leftReader.MaxDoc / (1 + docFreq);
int skipInterval = 16;
while (true)
{
if (Random().NextBoolean())
{
// nextDoc()
docid = leftDocs.NextDoc();
Assert.AreEqual(docid, rightDocs.NextDoc(), info);
}
else
{
// advance()
int skip = docid + (int)Math.Ceiling(Math.Abs(skipInterval + Random().NextDouble() * averageGap));
docid = leftDocs.Advance(skip);
Assert.AreEqual(docid, rightDocs.Advance(skip), info);
}
if (docid == DocIdSetIterator.NO_MORE_DOCS)
{
return;
}
int freq = leftDocs.Freq();
Assert.AreEqual(freq, rightDocs.Freq(), info);
for (int i = 0; i < freq; i++)
{
Assert.AreEqual(leftDocs.NextPosition(), rightDocs.NextPosition(), info);
Assert.AreEqual(leftDocs.Payload, rightDocs.Payload, info);
}
}
}
protected virtual void AssertEquals(RandomTokenStream tk, FieldType ft, Terms terms)
{
Assert.AreEqual(1, terms.DocCount);
int termCount = new JCG.HashSet <string>(tk.terms).Count;
Assert.AreEqual((long)termCount, terms.Count); // LUCENENET specific - cast required because types don't match (xUnit checks this)
Assert.AreEqual((long)termCount, terms.SumDocFreq); // LUCENENET specific - cast required because types don't match (xUnit checks this)
Assert.AreEqual(ft.StoreTermVectorPositions, terms.HasPositions);
Assert.AreEqual(ft.StoreTermVectorOffsets, terms.HasOffsets);
Assert.AreEqual(ft.StoreTermVectorPayloads && tk.HasPayloads(), terms.HasPayloads);
ISet <BytesRef> uniqueTerms = new JCG.HashSet <BytesRef>();
foreach (string term in tk.freqs.Keys)
{
uniqueTerms.Add(new BytesRef(term));
}
BytesRef[] sortedTerms = uniqueTerms.ToArray(/*new BytesRef[0]*/);
Array.Sort(sortedTerms, terms.Comparer);
TermsEnum termsEnum = terms.GetEnumerator(Random.NextBoolean() ? null : this.termsEnum.Value);
this.termsEnum.Value = termsEnum;
for (int i = 0; i < sortedTerms.Length; ++i)
{
Assert.IsTrue(termsEnum.MoveNext());
Assert.AreEqual(sortedTerms[i], termsEnum.Term);
Assert.AreEqual(1, termsEnum.DocFreq);
FixedBitSet bits = new FixedBitSet(1);
DocsEnum docsEnum = termsEnum.Docs(bits, Random.NextBoolean() ? null : this.docsEnum.Value);
Assert.AreEqual(DocsEnum.NO_MORE_DOCS, docsEnum.NextDoc());
bits.Set(0);
docsEnum = termsEnum.Docs(Random.NextBoolean() ? bits : null, Random.NextBoolean() ? null : docsEnum);
Assert.IsNotNull(docsEnum);
Assert.AreEqual(0, docsEnum.NextDoc());
Assert.AreEqual(0, docsEnum.DocID);
Assert.AreEqual(tk.freqs[termsEnum.Term.Utf8ToString()], docsEnum.Freq);
Assert.AreEqual(DocsEnum.NO_MORE_DOCS, docsEnum.NextDoc());
this.docsEnum.Value = docsEnum;
bits.Clear(0);
DocsAndPositionsEnum docsAndPositionsEnum = termsEnum.DocsAndPositions(bits, Random.NextBoolean() ? null : this.docsAndPositionsEnum.Value);
Assert.AreEqual(ft.StoreTermVectorOffsets || ft.StoreTermVectorPositions, docsAndPositionsEnum != null);
if (docsAndPositionsEnum != null)
{
Assert.AreEqual(DocsEnum.NO_MORE_DOCS, docsAndPositionsEnum.NextDoc());
}
bits.Set(0);
docsAndPositionsEnum = termsEnum.DocsAndPositions(Random.NextBoolean() ? bits : null, Random.NextBoolean() ? null : docsAndPositionsEnum);
Assert.AreEqual(ft.StoreTermVectorOffsets || ft.StoreTermVectorPositions, docsAndPositionsEnum != null);
if (terms.HasPositions || terms.HasOffsets)
{
Assert.AreEqual(0, docsAndPositionsEnum.NextDoc());
int freq = docsAndPositionsEnum.Freq;
Assert.AreEqual(tk.freqs[termsEnum.Term.Utf8ToString()], freq);
if (docsAndPositionsEnum != null)
{
for (int k = 0; k < freq; ++k)
{
int position = docsAndPositionsEnum.NextPosition();
ISet <int> indexes;
if (terms.HasPositions)
{
indexes = tk.positionToTerms[position];
Assert.IsNotNull(indexes);
}
else
{
indexes = tk.startOffsetToTerms[docsAndPositionsEnum.StartOffset];
Assert.IsNotNull(indexes);
}
if (terms.HasPositions)
{
bool foundPosition = false;
foreach (int index in indexes)
{
if (tk.termBytes[index].Equals(termsEnum.Term) && tk.positions[index] == position)
{
foundPosition = true;
break;
}
}
Assert.IsTrue(foundPosition);
}
if (terms.HasOffsets)
{
bool foundOffset = false;
foreach (int index in indexes)
{
if (tk.termBytes[index].Equals(termsEnum.Term) && tk.startOffsets[index] == docsAndPositionsEnum.StartOffset && tk.endOffsets[index] == docsAndPositionsEnum.EndOffset)
{
foundOffset = true;
break;
}
}
Assert.IsTrue(foundOffset);
}
if (terms.HasPayloads)
{
bool foundPayload = false;
foreach (int index in indexes)
{
if (tk.termBytes[index].Equals(termsEnum.Term) && Equals(tk.payloads[index], docsAndPositionsEnum.GetPayload()))
{
foundPayload = true;
break;
}
}
Assert.IsTrue(foundPayload);
}
}
try
{
docsAndPositionsEnum.NextPosition();
Assert.Fail();
}
catch (Exception e) when(e.IsException())
{
// ok
}
catch (Exception e) when(e.IsAssertionError())
{
// ok
}
}
Assert.AreEqual(DocsEnum.NO_MORE_DOCS, docsAndPositionsEnum.NextDoc());
}
this.docsAndPositionsEnum.Value = docsAndPositionsEnum;
}
Assert.IsFalse(termsEnum.MoveNext());
for (int i = 0; i < 5; ++i)
{
if (Random.NextBoolean())
{
Assert.IsTrue(termsEnum.SeekExact(RandomPicks.RandomFrom(Random, tk.termBytes)));
}
else
{
Assert.AreEqual(SeekStatus.FOUND, termsEnum.SeekCeil(RandomPicks.RandomFrom(Random, tk.termBytes)));
}
}
}