use of org.apache.lucene.index.IndexReader in project lucene-solr by apache.
the class TestSimilarity2 method testEmptyTerm.
/** similar to the above, however the field exists, but we query with a term that doesnt exist too */
public void testEmptyTerm() throws Exception {
Directory dir = newDirectory();
RandomIndexWriter iw = new RandomIndexWriter(random(), dir);
Document doc = new Document();
doc.add(newTextField("foo", "bar", Field.Store.NO));
iw.addDocument(doc);
IndexReader ir = iw.getReader();
iw.close();
IndexSearcher is = newSearcher(ir);
for (Similarity sim : sims) {
is.setSimilarity(sim);
BooleanQuery.Builder query = new BooleanQuery.Builder();
query.add(new TermQuery(new Term("foo", "bar")), BooleanClause.Occur.SHOULD);
query.add(new TermQuery(new Term("foo", "baz")), BooleanClause.Occur.SHOULD);
assertEquals(1, is.search(query.build(), 10).totalHits);
}
ir.close();
dir.close();
}
use of org.apache.lucene.index.IndexReader in project lucene-solr by apache.
the class TestSimilarity2 method testNoFieldSkew.
/** make sure scores are not skewed by docs not containing the field */
public void testNoFieldSkew() throws Exception {
Directory dir = newDirectory();
// an evil merge policy could reorder our docs for no reason
IndexWriterConfig iwConfig = newIndexWriterConfig().setMergePolicy(newLogMergePolicy());
RandomIndexWriter iw = new RandomIndexWriter(random(), dir, iwConfig);
Document doc = new Document();
doc.add(newTextField("foo", "bar baz somethingelse", Field.Store.NO));
iw.addDocument(doc);
IndexReader ir = iw.getReader();
IndexSearcher is = newSearcher(ir);
BooleanQuery.Builder queryBuilder = new BooleanQuery.Builder();
queryBuilder.add(new TermQuery(new Term("foo", "bar")), BooleanClause.Occur.SHOULD);
queryBuilder.add(new TermQuery(new Term("foo", "baz")), BooleanClause.Occur.SHOULD);
Query query = queryBuilder.build();
// collect scores
List<Explanation> scores = new ArrayList<>();
for (Similarity sim : sims) {
is.setSimilarity(sim);
scores.add(is.explain(query, 0));
}
ir.close();
// add some additional docs without the field
int numExtraDocs = TestUtil.nextInt(random(), 1, 1000);
for (int i = 0; i < numExtraDocs; i++) {
iw.addDocument(new Document());
}
// check scores are the same
ir = iw.getReader();
is = newSearcher(ir);
for (int i = 0; i < sims.size(); i++) {
is.setSimilarity(sims.get(i));
Explanation expected = scores.get(i);
Explanation actual = is.explain(query, 0);
assertEquals(sims.get(i).toString() + ": actual=" + actual + ",expected=" + expected, expected.getValue(), actual.getValue(), 0F);
}
iw.close();
ir.close();
dir.close();
}
use of org.apache.lucene.index.IndexReader in project lucene-solr by apache.
the class TestSimilarity2 method testOmitTF.
/** make sure all sims work if TF is omitted */
public void testOmitTF() throws Exception {
Directory dir = newDirectory();
RandomIndexWriter iw = new RandomIndexWriter(random(), dir);
Document doc = new Document();
FieldType ft = new FieldType(TextField.TYPE_NOT_STORED);
ft.setIndexOptions(IndexOptions.DOCS);
ft.freeze();
Field f = newField("foo", "bar", ft);
doc.add(f);
iw.addDocument(doc);
IndexReader ir = iw.getReader();
iw.close();
IndexSearcher is = newSearcher(ir);
for (Similarity sim : sims) {
is.setSimilarity(sim);
BooleanQuery.Builder query = new BooleanQuery.Builder();
query.add(new TermQuery(new Term("foo", "bar")), BooleanClause.Occur.SHOULD);
assertEquals(1, is.search(query.build(), 10).totalHits);
}
ir.close();
dir.close();
}
use of org.apache.lucene.index.IndexReader in project lucene-solr by apache.
the class TestSimilarity2 method testNoNorms.
/** make sure we can retrieve when norms are disabled */
public void testNoNorms() throws Exception {
Directory dir = newDirectory();
RandomIndexWriter iw = new RandomIndexWriter(random(), dir);
Document doc = new Document();
FieldType ft = new FieldType(TextField.TYPE_NOT_STORED);
ft.setOmitNorms(true);
ft.freeze();
doc.add(newField("foo", "bar", ft));
iw.addDocument(doc);
IndexReader ir = iw.getReader();
iw.close();
IndexSearcher is = newSearcher(ir);
for (Similarity sim : sims) {
is.setSimilarity(sim);
BooleanQuery.Builder query = new BooleanQuery.Builder();
query.add(new TermQuery(new Term("foo", "bar")), BooleanClause.Occur.SHOULD);
assertEquals(1, is.search(query.build(), 10).totalHits);
}
ir.close();
dir.close();
}
use of org.apache.lucene.index.IndexReader in project lucene-solr by apache.
the class TestSimilarityBase method testLengthEncodingBackwardCompatibility.
public void testLengthEncodingBackwardCompatibility() throws IOException {
Similarity similarity = RandomPicks.randomFrom(random(), sims);
for (int indexCreatedVersionMajor : new int[] { Version.LUCENE_6_0_0.major, Version.LATEST.major }) {
for (int length : new int[] { 1, 2, 4 }) {
// these length values are encoded accurately on both cases
Directory dir = newDirectory();
// set the version on the directory
new SegmentInfos(indexCreatedVersionMajor).commit(dir);
IndexWriter w = new IndexWriter(dir, newIndexWriterConfig().setSimilarity(similarity));
Document doc = new Document();
String value = IntStream.range(0, length).mapToObj(i -> "b").collect(Collectors.joining(" "));
doc.add(new TextField("foo", value, Store.NO));
w.addDocument(doc);
IndexReader reader = DirectoryReader.open(w);
IndexSearcher searcher = newSearcher(reader);
searcher.setSimilarity(similarity);
Term term = new Term("foo", "b");
TermContext context = TermContext.build(reader.getContext(), term);
SimWeight simWeight = similarity.computeWeight(1f, searcher.collectionStatistics("foo"), searcher.termStatistics(term, context));
SimilarityBase.BasicSimScorer simScorer = (SimilarityBase.BasicSimScorer) similarity.simScorer(simWeight, reader.leaves().get(0));
float docLength = simScorer.getLengthValue(0);
assertEquals(length, (int) docLength);
w.close();
reader.close();
dir.close();
}
}
}
Aggregations