@Test(groups = "ch12")
public void vectorTest() throws Exception {
FullTextSession session = Search.getFullTextSession(openSession());
Transaction tx = session.beginTransaction();
buildIndex(session, tx);
try {
tx = session.beginTransaction();
Query query = new TermQuery(new Term("content", "properties"));
System.out.println(query.toString());
FullTextQuery hibQuery = session.createFullTextQuery(query, ElectricalProperties.class);
hibQuery.setProjection(
FullTextQuery.DOCUMENT, FullTextQuery.DOCUMENT_ID, FullTextQuery.SCORE);
reader = getReader(session);
List<Object[]> results = hibQuery.list();
assert results.size() > 0 : "no results returned";
for (int x = 0; x < results.size(); x++) {
Integer docId = (Integer) results.get(x)[1];
TermPositionVector vector = (TermPositionVector) reader.getTermFreqVector(docId, "content");
String[] terms = vector.getTerms();
int[] f = vector.getTermFrequencies();
System.out.println(results.get(x)[2]);
for (int y = 0; y < vector.size(); y++) {
System.out.print("docID# =>" + docId);
System.out.print(" term => " + terms[y]);
System.out.print(" freq => " + f[y]);
int[] positions = vector.getTermPositions(y);
TermVectorOffsetInfo[] offsets = vector.getOffsets(y);
for (int z = 0; z < positions.length; z++) {
System.out.print(" position => " + positions[z]);
System.out.print(" starting offset => " + offsets[z].getStartOffset());
System.out.println(" ending offset => " + offsets[z].getEndOffset());
}
System.out.println("---------------");
}
}
for (Object element :
session.createQuery("from " + ElectricalProperties.class.getName()).list())
session.delete(element);
tx.commit();
} finally {
session.close();
if (provider != null) {
provider.closeReader(reader);
}
}
}
/**
* Reconstruct document fields.
*
* @param docNum document number. If this document is deleted, but the index is not optimized yet,
* the reconstruction process may still yield the reconstructed field content even from
* deleted documents.
* @return reconstructed document
* @throws Exception
*/
public Reconstructed reconstruct(int docNum) throws Exception {
if (docNum < 0 || docNum > reader.maxDoc()) {
throw new Exception("Document number outside of valid range.");
}
Reconstructed res = new Reconstructed();
if (deleted != null && deleted.get(docNum)) {
throw new Exception("Document is deleted.");
} else {
Document doc = reader.document(docNum);
for (int i = 0; i < fieldNames.length; i++) {
Field[] fs = doc.getFields(fieldNames[i]);
if (fs != null && fs.length > 0) {
res.getStoredFields().put(fieldNames[i], fs);
}
}
}
// collect values from unstored fields
HashSet<String> fields = new HashSet<String>(Arrays.asList(fieldNames));
// try to use term vectors if available
progress.maxValue = fieldNames.length;
progress.curValue = 0;
progress.minValue = 0;
for (int i = 0; i < fieldNames.length; i++) {
TermFreqVector tvf = reader.getTermFreqVector(docNum, fieldNames[i]);
if (tvf != null && tvf.size() > 0 && (tvf instanceof TermPositionVector)) {
TermPositionVector tpv = (TermPositionVector) tvf;
progress.message = "Reading term vectors ...";
progress.curValue = i;
setChanged();
notifyObservers(progress);
BytesRef[] tv = tpv.getTerms();
for (int k = 0; k < tv.length; k++) {
// do we have positions?
int[] posArr = tpv.getTermPositions(k);
if (posArr == null) {
// only offsets
TermVectorOffsetInfo[] offsets = tpv.getOffsets(k);
if (offsets.length == 0) {
continue;
}
// convert offsets into positions
posArr = convertOffsets(offsets);
}
GrowableStringArray gsa = res.getReconstructedFields().get(fieldNames[i]);
if (gsa == null) {
gsa = new GrowableStringArray();
res.getReconstructedFields().put(fieldNames[i], gsa);
}
for (int m = 0; m < posArr.length; m++) {
gsa.append(posArr[m], "|", tv[k].utf8ToString());
}
}
fields.remove(fieldNames[i]); // got what we wanted
}
}
// this loop collects data only from left-over fields
// not yet collected through term vectors
progress.maxValue = fields.size();
progress.curValue = 0;
progress.minValue = 0;
for (String fld : fields) {
progress.message = "Collecting terms in " + fld + " ...";
progress.curValue++;
setChanged();
notifyObservers(progress);
Terms terms = MultiFields.getTerms(reader, fld);
if (terms == null) { // no terms in this field
continue;
}
TermsEnum te = terms.iterator();
while (te.next() != null) {
DocsAndPositionsEnum dpe = te.docsAndPositions(deleted, null);
if (dpe == null) { // no position info for this field
break;
}
int num = dpe.advance(docNum);
if (num != docNum) { // either greater than or NO_MORE_DOCS
continue; // no data for this term in this doc
}
String term = te.term().utf8ToString();
GrowableStringArray gsa = (GrowableStringArray) res.getReconstructedFields().get(fld);
if (gsa == null) {
gsa = new GrowableStringArray();
res.getReconstructedFields().put(fld, gsa);
}
for (int k = 0; k < dpe.freq(); k++) {
int pos = dpe.nextPosition();
gsa.append(pos, "|", term);
}
}
}
progress.message = "Done.";
progress.curValue = 100;
setChanged();
notifyObservers(progress);
return res;
}
