public void testSeekCeilNotFound() throws Exception {
Directory dir = newDirectory();
RandomIndexWriter w = new RandomIndexWriter(random(), dir);
Document doc = new Document();
// Get empty string in there!
doc.add(newStringField("field", "", Field.Store.NO));
w.addDocument(doc);
for (int i = 0; i < 36; i++) {
doc = new Document();
String term = "" + (char) (97 + i);
String term2 = "a" + (char) (97 + i);
doc.add(newTextField("field", term + " " + term2, Field.Store.NO));
w.addDocument(doc);
}
w.forceMerge(1);
IndexReader r = w.getReader();
TermsEnum te = MultiFields.getTerms(r, "field").iterator(null);
assertEquals(TermsEnum.SeekStatus.NOT_FOUND, te.seekCeil(new BytesRef(new byte[] {0x22})));
assertEquals("a", te.term().utf8ToString());
assertEquals(1L, te.ord());
r.close();
w.close();
dir.close();
}
/*
* listTermDictionary displays the term dictionary for a field.
*/
static void listTermDictionary(IndexReader reader, String fieldName) throws IOException {
System.out.println("\nTerm Dictionary: field " + fieldName);
/*
Grant says:
MultiFields.getTerms(IndexReader, fieldName)
*/
Terms terms = MultiFields.getTerms(reader, fieldName);
if ((terms == null) || (terms.size() == -1))
System.out.println(" The term dictionary is empty.");
else {
System.out.println(" Vocabulary size: " + terms.size() + " terms");
TermsEnum ithTerm = terms.iterator(null);
/*
* Iterate over the terms in this document.
* Information about a term's occurrences (tf and
* positions) is accessed via the indexing API, which
* returns inverted lists that describe (only) the
* current document.
*/
while (ithTerm.next() != null) {
System.out.format(
" %-30s %d %d\n",
ithTerm.term().utf8ToString(), ithTerm.docFreq(), ithTerm.totalTermFreq());
}
;
}
;
}
public void test10kPulsed() throws Exception {
// we always run this test with pulsing codec.
Codec cp = _TestUtil.alwaysPostingsFormat(new Pulsing41PostingsFormat(1));
File f = _TestUtil.getTempDir("10kpulsed");
BaseDirectoryWrapper dir = newFSDirectory(f);
dir.setCheckIndexOnClose(false); // we do this ourselves explicitly
RandomIndexWriter iw =
new RandomIndexWriter(
random(),
dir,
newIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(random())).setCodec(cp));
Document document = new Document();
FieldType ft = new FieldType(TextField.TYPE_STORED);
switch (_TestUtil.nextInt(random(), 0, 2)) {
case 0:
ft.setIndexOptions(IndexOptions.DOCS_ONLY);
break;
case 1:
ft.setIndexOptions(IndexOptions.DOCS_AND_FREQS);
break;
default:
ft.setIndexOptions(IndexOptions.DOCS_AND_FREQS_AND_POSITIONS);
break;
}
Field field = newField("field", "", ft);
document.add(field);
NumberFormat df = new DecimalFormat("00000", new DecimalFormatSymbols(Locale.ROOT));
for (int i = 0; i < 10050; i++) {
field.setStringValue(df.format(i));
iw.addDocument(document);
}
IndexReader ir = iw.getReader();
iw.close();
TermsEnum te = MultiFields.getTerms(ir, "field").iterator(null);
DocsEnum de = null;
for (int i = 0; i < 10050; i++) {
String expected = df.format(i);
assertEquals(expected, te.next().utf8ToString());
de = _TestUtil.docs(random(), te, null, de, 0);
assertTrue(de.nextDoc() != DocIdSetIterator.NO_MORE_DOCS);
assertEquals(DocIdSetIterator.NO_MORE_DOCS, de.nextDoc());
}
ir.close();
_TestUtil.checkIndex(dir);
dir.close();
}
public void testThreeBlocks() throws Exception {
Directory dir = newDirectory();
RandomIndexWriter w = new RandomIndexWriter(random(), dir);
List<String> terms = new ArrayList<>();
for (int i = 0; i < 36; i++) {
Document doc = new Document();
String term = "" + (char) (97 + i);
terms.add(term);
if (VERBOSE) {
System.out.println("i=" + i + " term=" + term);
}
doc.add(newTextField("field", term, Field.Store.NO));
w.addDocument(doc);
}
for (int i = 0; i < 36; i++) {
Document doc = new Document();
String term = "m" + (char) (97 + i);
terms.add(term);
if (VERBOSE) {
System.out.println("i=" + i + " term=" + term);
}
doc.add(newTextField("field", term, Field.Store.NO));
w.addDocument(doc);
}
for (int i = 0; i < 36; i++) {
Document doc = new Document();
String term = "mo" + (char) (97 + i);
terms.add(term);
if (VERBOSE) {
System.out.println("i=" + i + " term=" + term);
}
doc.add(newTextField("field", term, Field.Store.NO));
w.addDocument(doc);
}
w.forceMerge(1);
IndexReader r = w.getReader();
TermsEnum te = MultiFields.getTerms(r, "field").iterator(null);
if (VERBOSE) {
while (te.next() != null) {
System.out.println("TERM: " + te.ord() + " " + te.term().utf8ToString());
}
}
assertTrue(te.seekExact(new BytesRef("mo")));
assertEquals(27, te.ord());
te.seekExact(90);
assertEquals(new BytesRef("s"), te.term());
testEnum(te, terms);
r.close();
w.close();
dir.close();
}
/* Copied from lucene 4.2.x core */
private static long totalTermFreq(IndexReader r, String field, BytesRef text) throws IOException {
final Terms terms = MultiFields.getTerms(r, field);
if (terms != null) {
final TermsEnum termsEnum = terms.iterator(null);
if (termsEnum.seekExact(text, true)) {
return termsEnum.totalTermFreq();
}
}
return 0;
}
private long getSumTermFrequency(IndexReader reader, String fieldName) {
Terms collectionTermVector = null;
try {
collectionTermVector = MultiFields.getTerms(reader, fieldName);
long totalTermFreq = collectionTermVector.getSumTotalTermFreq();
return totalTermFreq;
} catch (IOException e) {
LOG.warn("Unable to get total term frequency, it might not be indexed");
}
return 0;
}
public void testNonRootFloorBlocks() throws Exception {
Directory dir = newDirectory();
IndexWriterConfig iwc = new IndexWriterConfig(new MockAnalyzer(random()));
IndexWriter w = new IndexWriter(dir, iwc);
List<String> terms = new ArrayList<>();
for (int i = 0; i < 36; i++) {
Document doc = new Document();
String term = "" + (char) (97 + i);
terms.add(term);
if (VERBOSE) {
System.out.println("i=" + i + " term=" + term);
}
doc.add(newTextField("field", term, Field.Store.NO));
w.addDocument(doc);
}
for (int i = 0; i < 128; i++) {
Document doc = new Document();
String term = "m" + (char) i;
terms.add(term);
if (VERBOSE) {
System.out.println("i=" + i + " term=" + term + " bytes=" + new BytesRef(term));
}
doc.add(newStringField("field", term, Field.Store.NO));
w.addDocument(doc);
}
w.forceMerge(1);
IndexReader r = DirectoryReader.open(w, true);
TermsEnum te = MultiFields.getTerms(r, "field").iterator(null);
BytesRef term;
int ord = 0;
while ((term = te.next()) != null) {
if (VERBOSE) {
System.out.println("TEST: " + te.ord() + ": " + term.utf8ToString());
}
assertEquals(ord, te.ord());
ord++;
}
testEnum(te, terms);
r.close();
w.close();
dir.close();
}
private int countTerms(MultiTermQuery q) throws Exception {
final Terms terms = MultiFields.getTerms(reader, q.getField());
if (terms == null) return 0;
final TermsEnum termEnum = q.getTermsEnum(terms);
assertNotNull(termEnum);
int count = 0;
BytesRef cur, last = null;
while ((cur = termEnum.next()) != null) {
count++;
if (last != null) {
assertTrue(last.compareTo(cur) < 0);
}
last = BytesRef.deepCopyOf(cur);
}
// LUCENE-3314: the results after next() already returned null are undefined,
// assertNull(termEnum.next());
return count;
}
@Override
public FieldStats.Long stats(IndexReader reader) throws IOException {
int maxDoc = reader.maxDoc();
Terms terms = org.apache.lucene.index.MultiFields.getTerms(reader, name());
if (terms == null) {
return null;
}
long minValue = LegacyNumericUtils.getMinInt(terms);
long maxValue = LegacyNumericUtils.getMaxInt(terms);
return new FieldStats.Long(
maxDoc,
terms.getDocCount(),
terms.getSumDocFreq(),
terms.getSumTotalTermFreq(),
isSearchable(),
isAggregatable(),
minValue,
maxValue);
}
/** Test the WordScorer emitted by the smoothing model */
public void testBuildWordScorer() throws IOException {
SmoothingModel testModel = createTestModel();
Map<String, Analyzer> mapping = new HashMap<>();
mapping.put("field", new WhitespaceAnalyzer());
PerFieldAnalyzerWrapper wrapper =
new PerFieldAnalyzerWrapper(new WhitespaceAnalyzer(), mapping);
IndexWriter writer = new IndexWriter(new RAMDirectory(), new IndexWriterConfig(wrapper));
Document doc = new Document();
doc.add(new Field("field", "someText", TextField.TYPE_NOT_STORED));
writer.addDocument(doc);
DirectoryReader ir = DirectoryReader.open(writer);
WordScorer wordScorer =
testModel
.buildWordScorerFactory()
.newScorer(
ir, MultiFields.getTerms(ir, "field"), "field", 0.9d, BytesRefs.toBytesRef(" "));
assertWordScorer(wordScorer, testModel);
}
public void testSeveralNonRootBlocks() throws Exception {
Directory dir = newDirectory();
IndexWriterConfig iwc = new IndexWriterConfig(new MockAnalyzer(random()));
IndexWriter w = new IndexWriter(dir, iwc);
List<String> terms = new ArrayList<>();
for (int i = 0; i < 30; i++) {
for (int j = 0; j < 30; j++) {
Document doc = new Document();
String term = "" + (char) (97 + i) + (char) (97 + j);
terms.add(term);
if (VERBOSE) {
System.out.println("term=" + term);
}
doc.add(newTextField("body", term, Field.Store.NO));
w.addDocument(doc);
}
}
w.forceMerge(1);
IndexReader r = DirectoryReader.open(w, true);
TermsEnum te = MultiFields.getTerms(r, "body").iterator(null);
for (int i = 0; i < 30; i++) {
for (int j = 0; j < 30; j++) {
String term = "" + (char) (97 + i) + (char) (97 + j);
if (VERBOSE) {
System.out.println("TEST: check term=" + term);
}
assertEquals(term, te.next().utf8ToString());
assertEquals(30 * i + j, te.ord());
}
}
testEnum(te, terms);
te.seekExact(0);
assertEquals("aa", te.term().utf8ToString());
r.close();
w.close();
dir.close();
}
public void testBasic() throws Exception {
Directory dir = newDirectory();
RandomIndexWriter w = new RandomIndexWriter(random(), dir);
Document doc = new Document();
doc.add(newTextField("field", "a b c", Field.Store.NO));
w.addDocument(doc);
IndexReader r = w.getReader();
TermsEnum te = MultiFields.getTerms(r, "field").iterator(null);
// Test next()
assertEquals(new BytesRef("a"), te.next());
assertEquals(0L, te.ord());
assertEquals(new BytesRef("b"), te.next());
assertEquals(1L, te.ord());
assertEquals(new BytesRef("c"), te.next());
assertEquals(2L, te.ord());
assertNull(te.next());
// Test seekExact by term
assertTrue(te.seekExact(new BytesRef("b")));
assertEquals(1, te.ord());
assertTrue(te.seekExact(new BytesRef("a")));
assertEquals(0, te.ord());
assertTrue(te.seekExact(new BytesRef("c")));
assertEquals(2, te.ord());
// Test seekExact by ord
te.seekExact(1);
assertEquals(new BytesRef("b"), te.term());
te.seekExact(0);
assertEquals(new BytesRef("a"), te.term());
te.seekExact(2);
assertEquals(new BytesRef("c"), te.term());
r.close();
w.close();
dir.close();
}
@Override
public void visitMatchingTerms(IndexReader reader, String fieldName, MatchingTermVisitor mtv)
throws IOException {
int prefixLength = prefix.length();
Terms terms = MultiFields.getTerms(reader, fieldName);
if (terms != null) {
Matcher matcher = pattern.matcher("");
try {
TermsEnum termsEnum = terms.iterator(null);
TermsEnum.SeekStatus status = termsEnum.seekCeil(prefixRef);
BytesRef text;
if (status == TermsEnum.SeekStatus.FOUND) {
text = prefixRef;
} else if (status == TermsEnum.SeekStatus.NOT_FOUND) {
text = termsEnum.term();
} else {
text = null;
}
while (text != null) {
if (text != null && StringHelper.startsWith(text, prefixRef)) {
String textString = text.utf8ToString();
matcher.reset(textString.substring(prefixLength));
if (matcher.matches()) {
mtv.visitMatchingTerm(new Term(fieldName, textString));
}
} else {
break;
}
text = termsEnum.next();
}
} finally {
matcher.reset();
}
}
}
public void testFloorBlocks() throws Exception {
Directory dir = newDirectory();
IndexWriterConfig iwc = new IndexWriterConfig(new MockAnalyzer(random()));
IndexWriter w = new IndexWriter(dir, iwc);
for (int i = 0; i < 128; i++) {
Document doc = new Document();
String term = "" + (char) i;
if (VERBOSE) {
System.out.println("i=" + i + " term=" + term + " bytes=" + new BytesRef(term));
}
doc.add(newStringField("field", term, Field.Store.NO));
w.addDocument(doc);
}
w.forceMerge(1);
IndexReader r = DirectoryReader.open(w, true);
TermsEnum te = MultiFields.getTerms(r, "field").iterator(null);
if (VERBOSE) {
BytesRef term;
while ((term = te.next()) != null) {
System.out.println(" " + te.ord() + ": " + term.utf8ToString());
}
}
assertTrue(te.seekExact(new BytesRef("a")));
assertEquals(97, te.ord());
te.seekExact(98);
assertEquals(new BytesRef("b"), te.term());
assertTrue(te.seekExact(new BytesRef("z")));
assertEquals(122, te.ord());
r.close();
w.close();
dir.close();
}
/**
* 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;
}
private void addTerms(IndexReader reader, FieldVals f, ScoreTermQueue q) throws IOException {
if (f.queryString == null) return;
final Terms terms = MultiFields.getTerms(reader, f.fieldName);
if (terms == null) {
return;
}
try (TokenStream ts = analyzer.tokenStream(f.fieldName, f.queryString)) {
CharTermAttribute termAtt = ts.addAttribute(CharTermAttribute.class);
int corpusNumDocs = reader.numDocs();
HashSet<String> processedTerms = new HashSet<>();
ts.reset();
while (ts.incrementToken()) {
String term = termAtt.toString();
if (!processedTerms.contains(term)) {
processedTerms.add(term);
ScoreTermQueue variantsQ =
new ScoreTermQueue(
MAX_VARIANTS_PER_TERM); // maxNum variants considered for any one term
float minScore = 0;
Term startTerm = new Term(f.fieldName, term);
AttributeSource atts = new AttributeSource();
MaxNonCompetitiveBoostAttribute maxBoostAtt =
atts.addAttribute(MaxNonCompetitiveBoostAttribute.class);
SlowFuzzyTermsEnum fe =
new SlowFuzzyTermsEnum(terms, atts, startTerm, f.minSimilarity, f.prefixLength);
// store the df so all variants use same idf
int df = reader.docFreq(startTerm);
int numVariants = 0;
int totalVariantDocFreqs = 0;
BytesRef possibleMatch;
BoostAttribute boostAtt = fe.attributes().addAttribute(BoostAttribute.class);
while ((possibleMatch = fe.next()) != null) {
numVariants++;
totalVariantDocFreqs += fe.docFreq();
float score = boostAtt.getBoost();
if (variantsQ.size() < MAX_VARIANTS_PER_TERM || score > minScore) {
ScoreTerm st =
new ScoreTerm(
new Term(startTerm.field(), BytesRef.deepCopyOf(possibleMatch)),
score,
startTerm);
variantsQ.insertWithOverflow(st);
minScore = variantsQ.top().score; // maintain minScore
}
maxBoostAtt.setMaxNonCompetitiveBoost(
variantsQ.size() >= MAX_VARIANTS_PER_TERM ? minScore : Float.NEGATIVE_INFINITY);
}
if (numVariants > 0) {
int avgDf = totalVariantDocFreqs / numVariants;
if (df == 0) // no direct match we can use as df for all variants
{
df = avgDf; // use avg df of all variants
}
// take the top variants (scored by edit distance) and reset the score
// to include an IDF factor then add to the global queue for ranking
// overall top query terms
int size = variantsQ.size();
for (int i = 0; i < size; i++) {
ScoreTerm st = variantsQ.pop();
st.score = (st.score * st.score) * sim.idf(df, corpusNumDocs);
q.insertWithOverflow(st);
}
}
}
}
ts.end();
}
}
protected void fillFloatValues(FloatValues vals, IndexReader reader, String field)
throws IOException {
if (parser == null) {
try {
parser = FieldCache.DEFAULT_FLOAT_PARSER;
fillFloatValues(vals, reader, field);
return;
} catch (NumberFormatException ne) {
vals.parserHashCode = null; // wipe the previous one
parser = FieldCache.NUMERIC_UTILS_FLOAT_PARSER;
fillFloatValues(vals, reader, field);
return;
}
}
setParserAndResetCounts(vals, parser);
Terms terms = MultiFields.getTerms(reader, field);
int maxDoc = reader.maxDoc();
vals.values = null;
if (terms != null) {
final TermsEnum termsEnum = terms.iterator();
OpenBitSet validBits = (hasOption(OPTION_CACHE_BITS)) ? new OpenBitSet(maxDoc) : null;
DocsEnum docs = null;
try {
while (true) {
final BytesRef term = termsEnum.next();
if (term == null) {
break;
}
final float termval = parser.parseFloat(term);
docs = termsEnum.docs(null, docs);
while (true) {
final int docID = docs.nextDoc();
if (docID == DocIdSetIterator.NO_MORE_DOCS) {
break;
}
if (vals.values == null) {
vals.values = new float[maxDoc];
}
vals.values[docID] = termval;
vals.numDocs++;
if (validBits != null) {
validBits.set(docID);
}
}
vals.numTerms++;
}
} catch (FieldCache.StopFillCacheException stop) {
}
if (vals.valid == null) {
vals.valid = checkMatchAllBits(validBits, vals.numDocs, maxDoc);
}
}
if (vals.values == null) {
vals.values = new float[maxDoc];
}
if (vals.valid == null && vals.numDocs < 1) {
vals.valid = new Bits.MatchNoBits(maxDoc);
}
}
/*
* More Ideas:
* - add ability to find whitespace problems -> we can build a poor mans decompounder with our index based on a automaton?
* - add ability to build different error models maybe based on a confusion matrix?
* - try to combine a token with its subsequent token to find / detect word splits (optional)
* - for this to work we need some way to defined the position length of a candidate
* - phonetic filters could be interesting here too for candidate selection
*/
@Override
public Suggestion<? extends Entry<? extends Option>> innerExecute(
String name,
PhraseSuggestionContext suggestion,
IndexSearcher searcher,
CharsRefBuilder spare)
throws IOException {
double realWordErrorLikelihood = suggestion.realworldErrorLikelyhood();
final PhraseSuggestion response = new PhraseSuggestion(name, suggestion.getSize());
final IndexReader indexReader = searcher.getIndexReader();
List<PhraseSuggestionContext.DirectCandidateGenerator> generators = suggestion.generators();
final int numGenerators = generators.size();
final List<CandidateGenerator> gens = new ArrayList<>(generators.size());
for (int i = 0; i < numGenerators; i++) {
PhraseSuggestionContext.DirectCandidateGenerator generator = generators.get(i);
DirectSpellChecker directSpellChecker = SuggestUtils.getDirectSpellChecker(generator);
Terms terms = MultiFields.getTerms(indexReader, generator.field());
if (terms != null) {
gens.add(
new DirectCandidateGenerator(
directSpellChecker,
generator.field(),
generator.suggestMode(),
indexReader,
realWordErrorLikelihood,
generator.size(),
generator.preFilter(),
generator.postFilter(),
terms));
}
}
final String suggestField = suggestion.getField();
final Terms suggestTerms = MultiFields.getTerms(indexReader, suggestField);
if (gens.size() > 0 && suggestTerms != null) {
final NoisyChannelSpellChecker checker =
new NoisyChannelSpellChecker(
realWordErrorLikelihood, suggestion.getRequireUnigram(), suggestion.getTokenLimit());
final BytesRef separator = suggestion.separator();
TokenStream stream =
checker.tokenStream(
suggestion.getAnalyzer(), suggestion.getText(), spare, suggestion.getField());
WordScorer wordScorer =
suggestion
.model()
.newScorer(
indexReader, suggestTerms, suggestField, realWordErrorLikelihood, separator);
Result checkerResult =
checker.getCorrections(
stream,
new MultiCandidateGeneratorWrapper(
suggestion.getShardSize(), gens.toArray(new CandidateGenerator[gens.size()])),
suggestion.maxErrors(),
suggestion.getShardSize(),
wordScorer,
suggestion.confidence(),
suggestion.gramSize());
PhraseSuggestion.Entry resultEntry =
buildResultEntry(suggestion, spare, checkerResult.cutoffScore);
response.addTerm(resultEntry);
final BytesRefBuilder byteSpare = new BytesRefBuilder();
final EarlyTerminatingCollector collector = Lucene.createExistsCollector();
final CompiledScript collateScript;
if (suggestion.getCollateQueryScript() != null) {
collateScript = suggestion.getCollateQueryScript();
} else if (suggestion.getCollateFilterScript() != null) {
collateScript = suggestion.getCollateFilterScript();
} else {
collateScript = null;
}
final boolean collatePrune = (collateScript != null) && suggestion.collatePrune();
for (int i = 0; i < checkerResult.corrections.length; i++) {
Correction correction = checkerResult.corrections[i];
spare.copyUTF8Bytes(correction.join(SEPARATOR, byteSpare, null, null));
boolean collateMatch = true;
if (collateScript != null) {
// Checks if the template query collateScript yields any documents
// from the index for a correction, collateMatch is updated
final Map<String, Object> vars = suggestion.getCollateScriptParams();
vars.put(SUGGESTION_TEMPLATE_VAR_NAME, spare.toString());
final ExecutableScript executable = scriptService.executable(collateScript, vars);
final BytesReference querySource = (BytesReference) executable.run();
final ParsedQuery parsedQuery;
if (suggestion.getCollateFilterScript() != null) {
parsedQuery =
suggestion
.getQueryParserService()
.parse(
QueryBuilders.constantScoreQuery(QueryBuilders.wrapperQuery(querySource)));
} else {
parsedQuery = suggestion.getQueryParserService().parse(querySource);
}
collateMatch = Lucene.exists(searcher, parsedQuery.query(), collector);
}
if (!collateMatch && !collatePrune) {
continue;
}
Text phrase = new StringText(spare.toString());
Text highlighted = null;
if (suggestion.getPreTag() != null) {
spare.copyUTF8Bytes(
correction.join(
SEPARATOR, byteSpare, suggestion.getPreTag(), suggestion.getPostTag()));
highlighted = new StringText(spare.toString());
}
if (collatePrune) {
resultEntry.addOption(
new Suggestion.Entry.Option(
phrase, highlighted, (float) (correction.score), collateMatch));
} else {
resultEntry.addOption(
new Suggestion.Entry.Option(phrase, highlighted, (float) (correction.score)));
}
}
} else {
response.addTerm(buildResultEntry(suggestion, spare, Double.MIN_VALUE));
}
return response;
}
public void testTwoBlocks() throws Exception {
Directory dir = newDirectory();
RandomIndexWriter w = new RandomIndexWriter(random(), dir);
List<String> terms = new ArrayList<>();
for (int i = 0; i < 36; i++) {
Document doc = new Document();
String term = "" + (char) (97 + i);
terms.add(term);
if (VERBOSE) {
System.out.println("i=" + i + " term=" + term);
}
doc.add(newTextField("field", term, Field.Store.NO));
w.addDocument(doc);
}
for (int i = 0; i < 36; i++) {
Document doc = new Document();
String term = "m" + (char) (97 + i);
terms.add(term);
if (VERBOSE) {
System.out.println("i=" + i + " term=" + term);
}
doc.add(newTextField("field", term, Field.Store.NO));
w.addDocument(doc);
}
if (VERBOSE) {
System.out.println("TEST: now forceMerge");
}
w.forceMerge(1);
IndexReader r = w.getReader();
TermsEnum te = MultiFields.getTerms(r, "field").iterator(null);
assertTrue(te.seekExact(new BytesRef("mo")));
assertEquals(27, te.ord());
te.seekExact(54);
assertEquals(new BytesRef("s"), te.term());
Collections.sort(terms);
for (int i = terms.size() - 1; i >= 0; i--) {
te.seekExact(i);
assertEquals(i, te.ord());
assertEquals(terms.get(i), te.term().utf8ToString());
}
int iters = atLeast(1000);
for (int iter = 0; iter < iters; iter++) {
int ord = random().nextInt(terms.size());
BytesRef term = new BytesRef(terms.get(ord));
if (random().nextBoolean()) {
if (VERBOSE) {
System.out.println("TEST: iter=" + iter + " seek to ord=" + ord + " of " + terms.size());
}
te.seekExact(ord);
} else {
if (VERBOSE) {
System.out.println(
"TEST: iter="
+ iter
+ " seek to term="
+ terms.get(ord)
+ " ord="
+ ord
+ " of "
+ terms.size());
}
te.seekExact(term);
}
assertEquals(ord, te.ord());
assertEquals(term, te.term());
}
r.close();
w.close();
dir.close();
}