private SegmentCommitInfo merge(
Directory dir,
SegmentCommitInfo si1,
SegmentCommitInfo si2,
String merged,
boolean useCompoundFile)
throws Exception {
IOContext context = newIOContext(random());
SegmentReader r1 = new SegmentReader(si1, DirectoryReader.DEFAULT_TERMS_INDEX_DIVISOR, context);
SegmentReader r2 = new SegmentReader(si2, DirectoryReader.DEFAULT_TERMS_INDEX_DIVISOR, context);
final Codec codec = Codec.getDefault();
TrackingDirectoryWrapper trackingDir = new TrackingDirectoryWrapper(si1.info.dir);
final SegmentInfo si =
new SegmentInfo(si1.info.dir, Version.LATEST, merged, -1, false, codec, null);
SegmentMerger merger =
new SegmentMerger(
Arrays.<AtomicReader>asList(r1, r2),
si,
InfoStream.getDefault(),
trackingDir,
IndexWriterConfig.DEFAULT_TERM_INDEX_INTERVAL,
MergeState.CheckAbort.NONE,
new FieldInfos.FieldNumbers(),
context,
true);
MergeState mergeState = merger.merge();
r1.close();
r2.close();
final SegmentInfo info =
new SegmentInfo(
si1.info.dir,
Version.LATEST,
merged,
si1.info.getDocCount() + si2.info.getDocCount(),
false,
codec,
null);
info.setFiles(new HashSet<>(trackingDir.getCreatedFiles()));
if (useCompoundFile) {
Collection<String> filesToDelete =
IndexWriter.createCompoundFile(
InfoStream.getDefault(),
dir,
MergeState.CheckAbort.NONE,
info,
newIOContext(random()));
info.setUseCompoundFile(true);
for (final String fileToDelete : filesToDelete) {
si1.info.dir.deleteFile(fileToDelete);
}
}
return new SegmentCommitInfo(info, 0, -1L, -1L, -1L);
}
public void testMerge() throws IOException {
final Codec codec = Codec.getDefault();
final SegmentInfo si =
new SegmentInfo(
mergedDir,
Version.LATEST,
mergedSegment,
-1,
false,
codec,
Collections.emptyMap(),
StringHelper.randomId(),
new HashMap<>());
SegmentMerger merger =
new SegmentMerger(
Arrays.<CodecReader>asList(reader1, reader2),
si,
InfoStream.getDefault(),
mergedDir,
new FieldInfos.FieldNumbers(),
newIOContext(random(), new IOContext(new MergeInfo(-1, -1, false, -1))));
MergeState mergeState = merger.merge();
int docsMerged = mergeState.segmentInfo.maxDoc();
assertTrue(docsMerged == 2);
// Should be able to open a new SegmentReader against the new directory
SegmentReader mergedReader =
new SegmentReader(
new SegmentCommitInfo(mergeState.segmentInfo, 0, -1L, -1L, -1L),
newIOContext(random()));
assertTrue(mergedReader != null);
assertTrue(mergedReader.numDocs() == 2);
Document newDoc1 = mergedReader.document(0);
assertTrue(newDoc1 != null);
// There are 2 unstored fields on the document
assertTrue(
DocHelper.numFields(newDoc1) == DocHelper.numFields(doc1) - DocHelper.unstored.size());
Document newDoc2 = mergedReader.document(1);
assertTrue(newDoc2 != null);
assertTrue(
DocHelper.numFields(newDoc2) == DocHelper.numFields(doc2) - DocHelper.unstored.size());
PostingsEnum termDocs =
TestUtil.docs(
random(), mergedReader, DocHelper.TEXT_FIELD_2_KEY, new BytesRef("field"), null, 0);
assertTrue(termDocs != null);
assertTrue(termDocs.nextDoc() != DocIdSetIterator.NO_MORE_DOCS);
int tvCount = 0;
for (FieldInfo fieldInfo : mergedReader.getFieldInfos()) {
if (fieldInfo.hasVectors()) {
tvCount++;
}
}
// System.out.println("stored size: " + stored.size());
assertEquals("We do not have 3 fields that were indexed with term vector", 3, tvCount);
Terms vector = mergedReader.getTermVectors(0).terms(DocHelper.TEXT_FIELD_2_KEY);
assertNotNull(vector);
assertEquals(3, vector.size());
TermsEnum termsEnum = vector.iterator();
int i = 0;
while (termsEnum.next() != null) {
String term = termsEnum.term().utf8ToString();
int freq = (int) termsEnum.totalTermFreq();
// System.out.println("Term: " + term + " Freq: " + freq);
assertTrue(DocHelper.FIELD_2_TEXT.indexOf(term) != -1);
assertTrue(DocHelper.FIELD_2_FREQS[i] == freq);
i++;
}
TestSegmentReader.checkNorms(mergedReader);
mergedReader.close();
}
