void createIndex(const DirectoryPtr& dir, bool multiSegment) {
IndexWriter::unlock(dir);
IndexWriterPtr w = newLucene<IndexWriter>(dir, newLucene<WhitespaceAnalyzer>(), IndexWriter::MaxFieldLengthLIMITED);
w->setMergePolicy(newLucene<LogDocMergePolicy>(w));
for (int32_t i = 0; i < 100; ++i) {
w->addDocument(createDocument(i, 4));
if (multiSegment && (i % 10) == 0) {
w->commit();
}
}
if (!multiSegment) {
w->optimize();
}
w->close();
IndexReaderPtr r = IndexReader::open(dir, false);
if (multiSegment) {
EXPECT_TRUE(r->getSequentialSubReaders().size() > 1);
} else {
EXPECT_EQ(r->getSequentialSubReaders().size(), 1);
}
r->close();
}
void createIndex(int32_t numHits) {
int32_t numDocs = 500;
DirectoryPtr directory = newLucene<SeekCountingDirectory>();
IndexWriterPtr writer = newLucene<IndexWriter>(directory, newLucene<WhitespaceAnalyzer>(), true, IndexWriter::MaxFieldLengthLIMITED);
writer->setUseCompoundFile(false);
writer->setMaxBufferedDocs(10);
for (int32_t i = 0; i < numDocs; ++i) {
DocumentPtr doc = newLucene<Document>();
String content;
if (i % (numDocs / numHits) == 0) {
// add a document that matches the query "term1 term2"
content = term1 + L" " + term2;
} else if (i % 15 == 0) {
// add a document that only contains term1
content = term1 + L" " + term1;
} else {
// add a document that contains term2 but not term 1
content = term3 + L" " + term2;
}
doc->add(newLucene<Field>(field, content, Field::STORE_YES, Field::INDEX_ANALYZED));
writer->addDocument(doc);
}
// make sure the index has only a single segment
writer->optimize();
writer->close();
SegmentReaderPtr reader = SegmentReader::getOnlySegmentReader(directory);
searcher = newLucene<IndexSearcher>(reader);
}
TEST_F(SegmentTermEnumTest, testTermEnum) {
DirectoryPtr dir = newLucene<RAMDirectory>();
IndexWriterPtr writer = newLucene<IndexWriter>(dir, newLucene<WhitespaceAnalyzer>(), true, IndexWriter::MaxFieldLengthLIMITED);
// ADD 100 documents with term : aaa
// add 100 documents with terms: aaa bbb
// Therefore, term 'aaa' has document frequency of 200 and term 'bbb' 100
for (int32_t i = 0; i < 100; ++i) {
addDoc(writer, L"aaa");
addDoc(writer, L"aaa bbb");
}
writer->close();
// verify document frequency of terms in an unoptimized index
verifyDocFreq(dir);
// merge segments by optimizing the index
writer = newLucene<IndexWriter>(dir, newLucene<WhitespaceAnalyzer>(), false, IndexWriter::MaxFieldLengthLIMITED);
writer->optimize();
writer->close();
// verify document frequency of terms in an optimized index
verifyDocFreq(dir);
}
static void createIndexNoClose(bool multiSegment, const String& indexName, const IndexWriterPtr& w) {
for (int32_t i = 0; i < 100; ++i) {
w->addDocument(createDocument(i, indexName, 4));
}
if (!multiSegment) {
w->optimize();
}
}
/// Tests whether the DocumentWriter and SegmentMerger correctly enable the payload bit in the FieldInfo
TEST_F(PayloadsTest, testPayloadFieldBit) {
DirectoryPtr ram = newLucene<RAMDirectory>();
PayloadAnalyzerPtr analyzer = newLucene<PayloadAnalyzer>();
IndexWriterPtr writer = newLucene<IndexWriter>(ram, analyzer, true, IndexWriter::MaxFieldLengthLIMITED);
DocumentPtr d = newLucene<Document>();
// this field won't have any payloads
d->add(newLucene<Field>(L"f1", L"This field has no payloads", Field::STORE_NO, Field::INDEX_ANALYZED));
// this field will have payloads in all docs, however not for all term positions,
// so this field is used to check if the DocumentWriter correctly enables the payloads bit
// even if only some term positions have payloads
d->add(newLucene<Field>(L"f2", L"This field has payloads in all docs", Field::STORE_NO, Field::INDEX_ANALYZED));
d->add(newLucene<Field>(L"f2", L"This field has payloads in all docs", Field::STORE_NO, Field::INDEX_ANALYZED));
// this field is used to verify if the SegmentMerger enables payloads for a field if it has payloads
// enabled in only some documents
d->add(newLucene<Field>(L"f3", L"This field has payloads in some docs", Field::STORE_NO, Field::INDEX_ANALYZED));
// only add payload data for field f2
ByteArray someData(ByteArray::newInstance(8));
uint8_t input[8] = { 's', 'o', 'm', 'e', 'd', 'a', 't', 'a' };
std::memcpy(someData.get(), input, 8);
analyzer->setPayloadData(L"f2", 1, someData, 0, 1);
writer->addDocument(d);
// flush
writer->close();
SegmentReaderPtr reader = SegmentReader::getOnlySegmentReader(ram);
FieldInfosPtr fi = reader->fieldInfos();
EXPECT_TRUE(!fi->fieldInfo(L"f1")->storePayloads);
EXPECT_TRUE(fi->fieldInfo(L"f2")->storePayloads);
EXPECT_TRUE(!fi->fieldInfo(L"f3")->storePayloads);
reader->close();
// now we add another document which has payloads for field f3 and verify if the SegmentMerger
// enabled payloads for that field
writer = newLucene<IndexWriter>(ram, analyzer, true, IndexWriter::MaxFieldLengthLIMITED);
d = newLucene<Document>();
d->add(newLucene<Field>(L"f1", L"This field has no payloads", Field::STORE_NO, Field::INDEX_ANALYZED));
d->add(newLucene<Field>(L"f2", L"This field has payloads in all docs", Field::STORE_NO, Field::INDEX_ANALYZED));
d->add(newLucene<Field>(L"f2", L"This field has payloads in all docs", Field::STORE_NO, Field::INDEX_ANALYZED));
d->add(newLucene<Field>(L"f3", L"This field has payloads in some docs", Field::STORE_NO, Field::INDEX_ANALYZED));
// add payload data for field f2 and f3
analyzer->setPayloadData(L"f2", someData, 0, 1);
analyzer->setPayloadData(L"f3", someData, 0, 3);
writer->addDocument(d);
// force merge
writer->optimize();
// flush
writer->close();
reader = SegmentReader::getOnlySegmentReader(ram);
fi = reader->fieldInfos();
EXPECT_TRUE(!fi->fieldInfo(L"f1")->storePayloads);
EXPECT_TRUE(fi->fieldInfo(L"f2")->storePayloads);
EXPECT_TRUE(fi->fieldInfo(L"f3")->storePayloads);
reader->close();
}
virtual void run()
{
try
{
for (int32_t j = 0; j < numIter; ++j)
{
writerFinal->optimize(false);
for (int32_t k = 0; k < 17 * (1 + iFinal); ++k)
{
DocumentPtr d = newLucene<Document>();
d->add(newLucene<Field>(L"id", StringUtils::toString(iterFinal) + L"_" + StringUtils::toString(iFinal) + L"_" + StringUtils::toString(j) + L"_" + StringUtils::toString(k), Field::STORE_YES, Field::INDEX_NOT_ANALYZED));
d->add(newLucene<Field>(L"contents", intToEnglish(iFinal + k), Field::STORE_NO, Field::INDEX_ANALYZED));
writer->addDocument(d);
}
for (int32_t k = 0; k < 9 * (1 + iFinal); ++k)
writerFinal->deleteDocuments(newLucene<Term>(L"id", StringUtils::toString(iterFinal) + L"_" + StringUtils::toString(iFinal) + L"_" + StringUtils::toString(j) + L"_" + StringUtils::toString(k)));
writerFinal->optimize();
}
}
catch (LuceneException& e)
{
BOOST_FAIL("Unexpected exception: " << e.getError());
}
}
/// One-time setup to initialise static members
void setup() {
// set the theoretical maximum term count for 8bit (see docs for the number)
BooleanQuery::setMaxClauseCount(3 * 255 * 2 + 255);
directory = newLucene<RAMDirectory>();
IndexWriterPtr writer = newLucene<IndexWriter>(directory, newLucene<WhitespaceAnalyzer>(), true, IndexWriter::MaxFieldLengthUNLIMITED);
NumericFieldPtr field8 = newLucene<NumericField>(L"field8", 8, Field::STORE_YES, true);
NumericFieldPtr field4 = newLucene<NumericField>(L"field4", 4, Field::STORE_YES, true);
NumericFieldPtr field2 = newLucene<NumericField>(L"field2", 2, Field::STORE_YES, true);
NumericFieldPtr fieldNoTrie = newLucene<NumericField>(L"field" + StringUtils::toString(INT_MAX), INT_MAX, Field::STORE_YES, true);
NumericFieldPtr ascfield8 = newLucene<NumericField>(L"ascfield8", 8, Field::STORE_NO, true);
NumericFieldPtr ascfield4 = newLucene<NumericField>(L"ascfield4", 4, Field::STORE_NO, true);
NumericFieldPtr ascfield2 = newLucene<NumericField>(L"ascfield2", 2, Field::STORE_NO, true);
DocumentPtr doc = newLucene<Document>();
// add fields, that have a distance to test general functionality
doc->add(field8);
doc->add(field4);
doc->add(field2);
doc->add(fieldNoTrie);
// add ascending fields with a distance of 1, beginning at -noDocs/2 to test the correct splitting of range and inclusive/exclusive
doc->add(ascfield8);
doc->add(ascfield4);
doc->add(ascfield2);
// Add a series of noDocs docs with increasing int values
for (int32_t l = 0; l < noDocs; ++l) {
int32_t val = distance * l + startOffset;
field8->setIntValue(val);
field4->setIntValue(val);
field2->setIntValue(val);
fieldNoTrie->setIntValue(val);
val = l - (noDocs / 2);
ascfield8->setIntValue(val);
ascfield4->setIntValue(val);
ascfield2->setIntValue(val);
writer->addDocument(doc);
}
writer->optimize();
writer->close();
searcher = newLucene<IndexSearcher>(directory, true);
}
void doBody(int32_t j, Collection<DirectoryPtr> dirs) {
switch (j % 4) {
case 0:
mainWriter->addIndexesNoOptimize(dirs);
mainWriter->optimize();
break;
case 1:
mainWriter->addIndexesNoOptimize(dirs);
numAddIndexesNoOptimize->incrementAndGet();
break;
case 2:
mainWriter->addIndexes(readers);
break;
case 3:
mainWriter->commit();
break;
}
count->addAndGet(dirs.size() * NUM_INIT_DOCS);
}
DateSortTest() {
// Create an index writer.
directory = newLucene<RAMDirectory>();
IndexWriterPtr writer = newLucene<IndexWriter>(directory, newLucene<WhitespaceAnalyzer>(), true, IndexWriter::MaxFieldLengthLIMITED);
// oldest doc:
// Add the first document. text = "Document 1" dateTime = Oct 10 03:25:22 EDT 2007
writer->addDocument(createDocument(L"Document 1", 1192001122000LL));
// Add the second document. text = "Document 2" dateTime = Oct 10 03:25:26 EDT 2007
writer->addDocument(createDocument(L"Document 2", 1192001126000LL));
// Add the third document. text = "Document 3" dateTime = Oct 11 07:12:13 EDT 2007
writer->addDocument(createDocument(L"Document 3", 1192101133000LL));
// Add the fourth document. text = "Document 4" dateTime = Oct 11 08:02:09 EDT 2007
writer->addDocument(createDocument(L"Document 4", 1192104129000LL));
// latest doc:
// Add the fifth document. text = "Document 5" dateTime = Oct 12 13:25:43 EDT 2007
writer->addDocument(createDocument(L"Document 5", 1192209943000LL));
writer->optimize();
writer->close();
}
/// Builds an index with payloads in the given Directory and performs different
/// tests to verify the payload encoding
static void encodingTest(const DirectoryPtr& dir) {
PayloadAnalyzerPtr analyzer = newLucene<PayloadAnalyzer>();
IndexWriterPtr writer = newLucene<IndexWriter>(dir, analyzer, true, IndexWriter::MaxFieldLengthLIMITED);
// should be in sync with value in TermInfosWriter
int32_t skipInterval = 16;
int32_t numTerms = 5;
String fieldName = L"f1";
int32_t numDocs = skipInterval + 1;
// create content for the test documents with just a few terms
Collection<TermPtr> terms = generateTerms(fieldName, numTerms);
StringStream sb;
for (Collection<TermPtr>::iterator term = terms.begin(); term != terms.end(); ++term) {
sb << (*term)->text() << L" ";
}
String content = sb.str();
int32_t payloadDataLength = numTerms * numDocs * 2 + numTerms * numDocs * (numDocs - 1) / 2;
ByteArray payloadData = generateRandomData(payloadDataLength);
DocumentPtr d = newLucene<Document>();
d->add(newLucene<Field>(fieldName, content, Field::STORE_NO, Field::INDEX_ANALYZED));
// add the same document multiple times to have the same payload lengths for all
// occurrences within two consecutive skip intervals
int32_t offset = 0;
for (int32_t i = 0; i < 2 * numDocs; ++i) {
analyzer->setPayloadData(fieldName, payloadData, offset, 1);
offset += numTerms;
writer->addDocument(d);
}
// make sure we create more than one segment to test merging
writer->commit();
for (int32_t i = 0; i < numDocs; ++i) {
analyzer->setPayloadData(fieldName, payloadData, offset, i);
offset += i * numTerms;
writer->addDocument(d);
}
writer->optimize();
// flush
writer->close();
// Verify the index
IndexReaderPtr reader = IndexReader::open(dir, true);
ByteArray verifyPayloadData(ByteArray::newInstance(payloadDataLength));
offset = 0;
Collection<TermPositionsPtr> tps = Collection<TermPositionsPtr>::newInstance(numTerms);
for (int32_t i = 0; i < numTerms; ++i) {
tps[i] = reader->termPositions(terms[i]);
}
while (tps[0]->next()) {
for (int32_t i = 1; i < numTerms; ++i) {
tps[i]->next();
}
int32_t freq = tps[0]->freq();
for (int32_t i = 0; i < freq; ++i) {
for (int32_t j = 0; j < numTerms; ++j) {
tps[j]->nextPosition();
tps[j]->getPayload(verifyPayloadData, offset);
offset += tps[j]->getPayloadLength();
}
}
}
for (int32_t i = 0; i < numTerms; ++i) {
tps[i]->close();
}
EXPECT_TRUE(payloadData.equals(verifyPayloadData));
// test lazy skipping
TermPositionsPtr tp = reader->termPositions(terms[0]);
tp->next();
tp->nextPosition();
// now we don't read this payload
tp->nextPosition();
EXPECT_EQ(1, tp->getPayloadLength());
ByteArray payload = tp->getPayload(ByteArray(), 0);
EXPECT_EQ(payload[0], payloadData[numTerms]);
tp->nextPosition();
// we don't read this payload and skip to a different document
tp->skipTo(5);
tp->nextPosition();
EXPECT_EQ(1, tp->getPayloadLength());
payload = tp->getPayload(ByteArray(), 0);
EXPECT_EQ(payload[0], payloadData[5 * numTerms]);
// Test different lengths at skip points
tp->seek(terms[1]);
tp->next();
tp->nextPosition();
EXPECT_EQ(1, tp->getPayloadLength());
tp->skipTo(skipInterval - 1);
tp->nextPosition();
EXPECT_EQ(1, tp->getPayloadLength());
tp->skipTo(2 * skipInterval - 1);
tp->nextPosition();
EXPECT_EQ(1, tp->getPayloadLength());
tp->skipTo(3 * skipInterval - 1);
tp->nextPosition();
EXPECT_EQ(3 * skipInterval - 2 * numDocs - 1, tp->getPayloadLength());
// Test multiple call of getPayload()
tp->getPayload(ByteArray(), 0);
// it is forbidden to call getPayload() more than once without calling nextPosition()
try {
tp->getPayload(ByteArray(), 0);
} catch (IOException& e) {
EXPECT_TRUE(check_exception(LuceneException::IO)(e));
}
reader->close();
// test long payload
analyzer = newLucene<PayloadAnalyzer>();
writer = newLucene<IndexWriter>(dir, analyzer, true, IndexWriter::MaxFieldLengthLIMITED);
String singleTerm = L"lucene";
d = newLucene<Document>();
d->add(newLucene<Field>(fieldName, singleTerm, Field::STORE_NO, Field::INDEX_ANALYZED));
// 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->optimize();
// flush
writer->close();
reader = IndexReader::open(dir, true);
tp = reader->termPositions(newLucene<Term>(fieldName, singleTerm));
tp->next();
tp->nextPosition();
verifyPayloadData.resize(tp->getPayloadLength());
tp->getPayload(verifyPayloadData, 0);
ByteArray portion(ByteArray::newInstance(1500));
MiscUtils::arrayCopy(payloadData.get(), 100, portion.get(), 0, 1500);
EXPECT_TRUE(portion.equals(verifyPayloadData));
reader->close();
}
void checkSkipTo(int32_t indexDivisor) {
DirectoryPtr dir = newLucene<RAMDirectory>();
IndexWriterPtr writer = newLucene<IndexWriter>(dir, newLucene<WhitespaceAnalyzer>(), true, IndexWriter::MaxFieldLengthLIMITED);
TermPtr ta = newLucene<Term>(L"content", L"aaa");
for (int32_t i = 0; i < 10; ++i) {
addDoc(writer, L"aaa aaa aaa aaa");
}
TermPtr tb = newLucene<Term>(L"content", L"bbb");
for (int32_t i = 0; i < 16; ++i) {
addDoc(writer, L"bbb bbb bbb bbb");
}
TermPtr tc = newLucene<Term>(L"content", L"ccc");
for (int32_t i = 0; i < 50; ++i) {
addDoc(writer, L"ccc ccc ccc ccc");
}
// assure that we deal with a single segment
writer->optimize();
writer->close();
IndexReaderPtr reader = IndexReader::open(dir, IndexDeletionPolicyPtr(), true, indexDivisor);
TermDocsPtr tdocs = reader->termDocs();
// without optimization (assumption skipInterval == 16)
// with next
tdocs->seek(ta);
EXPECT_TRUE(tdocs->next());
EXPECT_EQ(0, tdocs->doc());
EXPECT_EQ(4, tdocs->freq());
EXPECT_TRUE(tdocs->next());
EXPECT_EQ(1, tdocs->doc());
EXPECT_EQ(4, tdocs->freq());
EXPECT_TRUE(tdocs->skipTo(0));
EXPECT_EQ(2, tdocs->doc());
EXPECT_TRUE(tdocs->skipTo(4));
EXPECT_EQ(4, tdocs->doc());
EXPECT_TRUE(tdocs->skipTo(9));
EXPECT_EQ(9, tdocs->doc());
EXPECT_TRUE(!tdocs->skipTo(10));
// without next
tdocs->seek(ta);
EXPECT_TRUE(tdocs->skipTo(0));
EXPECT_EQ(0, tdocs->doc());
EXPECT_TRUE(tdocs->skipTo(4));
EXPECT_EQ(4, tdocs->doc());
EXPECT_TRUE(tdocs->skipTo(9));
EXPECT_EQ(9, tdocs->doc());
EXPECT_TRUE(!tdocs->skipTo(10));
// exactly skipInterval documents and therefore with optimization
// with next
tdocs->seek(tb);
EXPECT_TRUE(tdocs->next());
EXPECT_EQ(10, tdocs->doc());
EXPECT_EQ(4, tdocs->freq());
EXPECT_TRUE(tdocs->next());
EXPECT_EQ(11, tdocs->doc());
EXPECT_EQ(4, tdocs->freq());
EXPECT_TRUE(tdocs->skipTo(5));
EXPECT_EQ(12, tdocs->doc());
EXPECT_TRUE(tdocs->skipTo(15));
EXPECT_EQ(15, tdocs->doc());
EXPECT_TRUE(tdocs->skipTo(24));
EXPECT_EQ(24, tdocs->doc());
EXPECT_TRUE(tdocs->skipTo(25));
EXPECT_EQ(25, tdocs->doc());
EXPECT_TRUE(!tdocs->skipTo(26));
// without next
tdocs->seek(tb);
EXPECT_TRUE(tdocs->skipTo(5));
EXPECT_EQ(10, tdocs->doc());
EXPECT_TRUE(tdocs->skipTo(15));
EXPECT_EQ(15, tdocs->doc());
EXPECT_TRUE(tdocs->skipTo(24));
EXPECT_EQ(24, tdocs->doc());
EXPECT_TRUE(tdocs->skipTo(25));
EXPECT_EQ(25, tdocs->doc());
EXPECT_TRUE(!tdocs->skipTo(26));
// much more than skipInterval documents and therefore with optimization
// with next
tdocs->seek(tc);
EXPECT_TRUE(tdocs->next());
EXPECT_EQ(26, tdocs->doc());
EXPECT_EQ(4, tdocs->freq());
EXPECT_TRUE(tdocs->next());
EXPECT_EQ(27, tdocs->doc());
EXPECT_EQ(4, tdocs->freq());
EXPECT_TRUE(tdocs->skipTo(5));
EXPECT_EQ(28, tdocs->doc());
EXPECT_TRUE(tdocs->skipTo(40));
EXPECT_EQ(40, tdocs->doc());
EXPECT_TRUE(tdocs->skipTo(57));
EXPECT_EQ(57, tdocs->doc());
EXPECT_TRUE(tdocs->skipTo(74));
EXPECT_EQ(74, tdocs->doc());
EXPECT_TRUE(tdocs->skipTo(75));
EXPECT_EQ(75, tdocs->doc());
EXPECT_TRUE(!tdocs->skipTo(76));
// without next
tdocs->seek(tc);
EXPECT_TRUE(tdocs->skipTo(5));
EXPECT_EQ(26, tdocs->doc());
EXPECT_TRUE(tdocs->skipTo(40));
EXPECT_EQ(40, tdocs->doc());
EXPECT_TRUE(tdocs->skipTo(57));
EXPECT_EQ(57, tdocs->doc());
EXPECT_TRUE(tdocs->skipTo(74));
EXPECT_EQ(74, tdocs->doc());
EXPECT_TRUE(tdocs->skipTo(75));
EXPECT_EQ(75, tdocs->doc());
EXPECT_TRUE(!tdocs->skipTo(76));
tdocs->close();
reader->close();
dir->close();
}
TEST_F(SimpleExplanationsOfNonMatchesTest, testTermQueryMultiSearcherExplain) {
// creating two directories for indices
DirectoryPtr indexStoreA = newLucene<MockRAMDirectory>();
DirectoryPtr indexStoreB = newLucene<MockRAMDirectory>();
DocumentPtr lDoc = newLucene<Document>();
lDoc->add(newLucene<Field>(L"handle", L"1 2", Field::STORE_YES, Field::INDEX_ANALYZED));
DocumentPtr lDoc2 = newLucene<Document>();
lDoc2->add(newLucene<Field>(L"handle", L"1 2", Field::STORE_YES, Field::INDEX_ANALYZED));
DocumentPtr lDoc3 = newLucene<Document>();
lDoc3->add(newLucene<Field>(L"handle", L"1 2", Field::STORE_YES, Field::INDEX_ANALYZED));
IndexWriterPtr writerA = newLucene<IndexWriter>(indexStoreA, newLucene<StandardAnalyzer>(LuceneVersion::LUCENE_CURRENT), true, IndexWriter::MaxFieldLengthLIMITED);
IndexWriterPtr writerB = newLucene<IndexWriter>(indexStoreB, newLucene<StandardAnalyzer>(LuceneVersion::LUCENE_CURRENT), true, IndexWriter::MaxFieldLengthLIMITED);
writerA->addDocument(lDoc);
writerA->addDocument(lDoc2);
writerA->optimize();
writerA->close();
writerB->addDocument(lDoc3);
writerB->close();
QueryParserPtr parser = newLucene<QueryParser>(LuceneVersion::LUCENE_CURRENT, L"fulltext", newLucene<StandardAnalyzer>(LuceneVersion::LUCENE_CURRENT));
QueryPtr query = parser->parse(L"handle:1");
Collection<SearchablePtr> searchers = newCollection<SearchablePtr>(
newLucene<IndexSearcher>(indexStoreB, true),
newLucene<IndexSearcher>(indexStoreA, true)
);
SearcherPtr mSearcher = newLucene<MultiSearcher>(searchers);
Collection<ScoreDocPtr> hits = mSearcher->search(query, FilterPtr(), 1000)->scoreDocs;
EXPECT_EQ(3, hits.size());
ExplanationPtr explain = mSearcher->explain(query, hits[0]->doc);
String exp = explain->toString();
EXPECT_TRUE(exp.find(L"maxDocs=3") != String::npos);
EXPECT_TRUE(exp.find(L"docFreq=3") != String::npos);
query = parser->parse(L"handle:\"1 2\"");
hits = mSearcher->search(query, FilterPtr(), 1000)->scoreDocs;
EXPECT_EQ(3, hits.size());
explain = mSearcher->explain(query, hits[0]->doc);
exp = explain->toString();
EXPECT_TRUE(exp.find(L"1=3") != String::npos);
EXPECT_TRUE(exp.find(L"2=3") != String::npos);
query = newLucene<SpanNearQuery>(newCollection<SpanQueryPtr>(newLucene<SpanTermQuery>(newLucene<Term>(L"handle", L"1")), newLucene<SpanTermQuery>(newLucene<Term>(L"handle", L"2"))), 0, true);
hits = mSearcher->search(query, FilterPtr(), 1000)->scoreDocs;
EXPECT_EQ(3, hits.size());
explain = mSearcher->explain(query, hits[0]->doc);
exp = explain->toString();
EXPECT_TRUE(exp.find(L"1=3") != String::npos);
EXPECT_TRUE(exp.find(L"2=3") != String::npos);
mSearcher->close();
}
/// Test that norms values are preserved as the index is maintained. Including separate norms.
/// Including merging indexes with separate norms. Including optimize.
TEST_F(IndexReaderCloneNormsTest, testNorms) {
// test with a single index: index1
String indexDir1(FileUtils::joinPath(getTempDir(), L"lucenetestindex1"));
DirectoryPtr dir1 = FSDirectory::open(indexDir1);
IndexWriter::unlock(dir1);
norms = Collection<double>::newInstance();
modifiedNorms = Collection<double>::newInstance();
createIndex(dir1);
doTestNorms(dir1);
// test with a single index: index2
Collection<double> norms1 = norms;
Collection<double> modifiedNorms1 = modifiedNorms;
int32_t numDocNorms1 = numDocNorms;
norms = Collection<double>::newInstance();
modifiedNorms = Collection<double>::newInstance();
numDocNorms = 0;
String indexDir2(FileUtils::joinPath(getTempDir(), L"lucenetestindex2"));
DirectoryPtr dir2 = FSDirectory::open(indexDir2);
createIndex(dir2);
doTestNorms(dir2);
// add index1 and index2 to a third index: index3
String indexDir3(FileUtils::joinPath(getTempDir(), L"lucenetestindex3"));
DirectoryPtr dir3 = FSDirectory::open(indexDir3);
createIndex(dir3);
IndexWriterPtr iw = newLucene<IndexWriter>(dir3, anlzr, false, IndexWriter::MaxFieldLengthLIMITED);
iw->setMaxBufferedDocs(5);
iw->setMergeFactor(3);
iw->addIndexesNoOptimize(newCollection<DirectoryPtr>(dir1, dir2));
iw->optimize();
iw->close();
norms1.addAll(norms.begin(), norms.end());
norms = norms1;
modifiedNorms1.addAll(modifiedNorms.begin(), modifiedNorms.end());
modifiedNorms = modifiedNorms1;
numDocNorms += numDocNorms1;
// test with index3
verifyIndex(dir3);
doTestNorms(dir3);
// now with optimize
iw = newLucene<IndexWriter>(dir3, anlzr, false, IndexWriter::MaxFieldLengthLIMITED);
iw->setMaxBufferedDocs(5);
iw->setMergeFactor(3);
iw->optimize();
iw->close();
verifyIndex(dir3);
dir1->close();
dir2->close();
dir3->close();
}
