List of usage examples for org.apache.lucene.index PostingsEnum freq
public abstract int freq() throws IOException;
From source file:br.pucminas.ri.jsearch.queryexpansion.RocchioQueryExpansion.java
License:Open Source License
private List<Entry<String, Float>> getTermScoreList(Directory directory) throws CorruptIndexException, IOException { Map<String, Float> termScoreMap = new HashMap<>(); ConcreteTFIDFSimilarity sim = new ConcreteTFIDFSimilarity(); try (IndexReader idxReader = DirectoryReader.open(directory)) { idxReader.leaves().stream().map((leaf) -> leaf.reader()).forEach((reader) -> { try { Terms terms = reader.terms(Constants.DOC_CONTENT); TermsEnum termsEnum = terms.iterator(); PostingsEnum postings = null; int docsNum = idxReader.numDocs(); BytesRef text;/* www .j a v a2s . c o m*/ while ((text = termsEnum.next()) != null) { postings = termsEnum.postings(postings); while (postings.nextDoc() != PostingsEnum.NO_MORE_DOCS) { int freq = postings.freq(); float tf = sim.tf(freq); float idf = sim.idf(termsEnum.docFreq(), indexReader.numDocs()); termScoreMap.put(text.utf8ToString(), BETA * (tf * idf)); } } } catch (IOException ex) { Logger.getLogger(RocchioQueryExpansion.class.getName()).log(Level.SEVERE, null, ex); } finally { try { idxReader.close(); } catch (IOException ex) { Logger.getLogger(RocchioQueryExpansion.class.getName()).log(Level.SEVERE, null, ex); } } }); } return new ArrayList<>(termScoreMap.entrySet()); }
From source file:br.pucminas.ri.jsearch.queryexpansion.RocchioQueryExpansion.java
License:Open Source License
private float getScore(Directory directory, String term) throws CorruptIndexException, IOException { try (IndexReader idxReader = DirectoryReader.open(directory)) { ConcreteTFIDFSimilarity sim = new ConcreteTFIDFSimilarity(); for (LeafReaderContext context : idxReader.leaves()) { LeafReader reader = context.reader(); try { Terms terms = reader.terms(Constants.DOC_CONTENT); TermsEnum termsEnum = terms.iterator(); PostingsEnum postings = null; BytesRef text;/*from w w w . jav a2 s.co m*/ while ((text = termsEnum.next()) != null) { postings = termsEnum.postings(postings); if (text.utf8ToString().equalsIgnoreCase(term)) { while (postings.nextDoc() != PostingsEnum.NO_MORE_DOCS) { int freq = postings.freq(); float tf = sim.tf(freq); float idf = sim.idf(termsEnum.docFreq(), indexReader.numDocs()); return tf * idf; } } } } catch (IOException ex) { Logger.getLogger(RocchioQueryExpansion.class.getName()).log(Level.SEVERE, null, ex); } } } return 0; }
From source file:com.github.flaxsearch.resources.PositionsResource.java
License:Apache License
@GET public DocTermData getDocTermData(@QueryParam("segment") Integer segment, @PathParam("field") String field, @PathParam("term") String term, @PathParam("docId") int docId) throws Exception { TermsEnum te = readerManager.findTermPostings(segment, field, term); PostingsEnum pe = te.postings(null, PostingsEnum.ALL); if (pe.advance(docId) != docId) { String seg = segment == null ? "" : " in segment " + segment; String msg = String.format(Locale.ROOT, "No document %d%s in index", docId, seg); throw new WebApplicationException(msg, Response.Status.NOT_FOUND); }//w w w . ja v a 2 s. c o m List<PositionData> positions = new ArrayList<>(); int remaining = pe.freq(); while (remaining > 0) { remaining--; positions.add(new PositionData(pe)); } return new DocTermData(docId, positions); }
From source file:com.rocana.lucene.codec.v1.RocanaBasePostingsFormatTestCase.java
License:Apache License
@Override public void testInvertedWrite() throws Exception { Directory dir = newDirectory();/* w w w .j a va 2 s. c o m*/ MockAnalyzer analyzer = new MockAnalyzer(random()); analyzer.setMaxTokenLength(TestUtil.nextInt(random(), 1, IndexWriter.MAX_TERM_LENGTH)); IndexWriterConfig iwc = newIndexWriterConfig(analyzer); // Must be concurrent because thread(s) can be merging // while up to one thread flushes, and each of those // threads iterates over the map while the flushing // thread might be adding to it: final Map<String, TermFreqs> termFreqs = new ConcurrentHashMap<>(); final AtomicLong sumDocFreq = new AtomicLong(); final AtomicLong sumTotalTermFreq = new AtomicLong(); // TODO: would be better to use / delegate to the current // Codec returned by getCodec() iwc.setCodec(new AssertingCodec() { @Override public PostingsFormat getPostingsFormatForField(String field) { PostingsFormat p = getCodec().postingsFormat(); if (p instanceof PerFieldPostingsFormat) { p = ((PerFieldPostingsFormat) p).getPostingsFormatForField(field); } if (p instanceof RocanaPerFieldPostingsFormat) { p = ((RocanaPerFieldPostingsFormat) p).getPostingsFormatForField(field); } final PostingsFormat defaultPostingsFormat = p; final Thread mainThread = Thread.currentThread(); if (field.equals("body")) { // A PF that counts up some stats and then in // the end we verify the stats match what the // final IndexReader says, just to exercise the // new freedom of iterating the postings more // than once at flush/merge: return new PostingsFormat(defaultPostingsFormat.getName()) { @Override public FieldsConsumer fieldsConsumer(final SegmentWriteState state) throws IOException { final FieldsConsumer fieldsConsumer = defaultPostingsFormat.fieldsConsumer(state); return new FieldsConsumer() { @Override public void write(Fields fields) throws IOException { fieldsConsumer.write(fields); boolean isMerge = state.context.context == IOContext.Context.MERGE; // We only use one thread for flushing // in this test: assert isMerge || Thread.currentThread() == mainThread; // We iterate the provided TermsEnum // twice, so we excercise this new freedom // with the inverted API; if // addOnSecondPass is true, we add up // term stats on the 2nd iteration: boolean addOnSecondPass = random().nextBoolean(); //System.out.println("write isMerge=" + isMerge + " 2ndPass=" + addOnSecondPass); // Gather our own stats: Terms terms = fields.terms("body"); assert terms != null; TermsEnum termsEnum = terms.iterator(); PostingsEnum docs = null; while (termsEnum.next() != null) { BytesRef term = termsEnum.term(); // TODO: also sometimes ask for payloads/offsets? boolean noPositions = random().nextBoolean(); if (noPositions) { docs = termsEnum.postings(docs, PostingsEnum.FREQS); } else { docs = termsEnum.postings(null, PostingsEnum.POSITIONS); } int docFreq = 0; long totalTermFreq = 0; while (docs.nextDoc() != PostingsEnum.NO_MORE_DOCS) { docFreq++; totalTermFreq += docs.freq(); int limit = TestUtil.nextInt(random(), 1, docs.freq()); if (!noPositions) { for (int i = 0; i < limit; i++) { docs.nextPosition(); } } } String termString = term.utf8ToString(); // During merge we should only see terms // we had already seen during a // previous flush: assertTrue(isMerge == false || termFreqs.containsKey(termString)); if (isMerge == false) { if (addOnSecondPass == false) { TermFreqs tf = termFreqs.get(termString); if (tf == null) { tf = new TermFreqs(); termFreqs.put(termString, tf); } tf.docFreq += docFreq; tf.totalTermFreq += totalTermFreq; sumDocFreq.addAndGet(docFreq); sumTotalTermFreq.addAndGet(totalTermFreq); } else if (termFreqs.containsKey(termString) == false) { // Add placeholder (2nd pass will // set its counts): termFreqs.put(termString, new TermFreqs()); } } } // Also test seeking the TermsEnum: for (String term : termFreqs.keySet()) { if (termsEnum.seekExact(new BytesRef(term))) { // TODO: also sometimes ask for payloads/offsets? boolean noPositions = random().nextBoolean(); if (noPositions) { docs = termsEnum.postings(docs, PostingsEnum.FREQS); } else { docs = termsEnum.postings(null, PostingsEnum.POSITIONS); } int docFreq = 0; long totalTermFreq = 0; while (docs.nextDoc() != PostingsEnum.NO_MORE_DOCS) { docFreq++; totalTermFreq += docs.freq(); int limit = TestUtil.nextInt(random(), 1, docs.freq()); if (!noPositions) { for (int i = 0; i < limit; i++) { docs.nextPosition(); } } } if (isMerge == false && addOnSecondPass) { TermFreqs tf = termFreqs.get(term); assert tf != null; tf.docFreq += docFreq; tf.totalTermFreq += totalTermFreq; sumDocFreq.addAndGet(docFreq); sumTotalTermFreq.addAndGet(totalTermFreq); } //System.out.println(" term=" + term + " docFreq=" + docFreq + " ttDF=" + termToDocFreq.get(term)); assertTrue(docFreq <= termFreqs.get(term).docFreq); assertTrue(totalTermFreq <= termFreqs.get(term).totalTermFreq); } } // Also test seekCeil for (int iter = 0; iter < 10; iter++) { BytesRef term = new BytesRef( TestUtil.randomRealisticUnicodeString(random())); SeekStatus status = termsEnum.seekCeil(term); if (status == SeekStatus.NOT_FOUND) { assertTrue(term.compareTo(termsEnum.term()) < 0); } } } @Override public void close() throws IOException { fieldsConsumer.close(); } }; } @Override public FieldsProducer fieldsProducer(SegmentReadState state) throws IOException { return defaultPostingsFormat.fieldsProducer(state); } }; } else { return defaultPostingsFormat; } } }); RandomIndexWriter w = new RandomIndexWriter(random(), dir, iwc); LineFileDocs docs = new LineFileDocs(random()); int bytesToIndex = atLeast(100) * 1024; int bytesIndexed = 0; while (bytesIndexed < bytesToIndex) { Document doc = docs.nextDoc(); w.addDocument(doc); bytesIndexed += RamUsageTester.sizeOf(doc); } IndexReader r = w.getReader(); w.close(); Terms terms = MultiFields.getTerms(r, "body"); assertEquals(sumDocFreq.get(), terms.getSumDocFreq()); assertEquals(sumTotalTermFreq.get(), terms.getSumTotalTermFreq()); TermsEnum termsEnum = terms.iterator(); long termCount = 0; boolean supportsOrds = true; while (termsEnum.next() != null) { BytesRef term = termsEnum.term(); assertEquals(termFreqs.get(term.utf8ToString()).docFreq, termsEnum.docFreq()); assertEquals(termFreqs.get(term.utf8ToString()).totalTermFreq, termsEnum.totalTermFreq()); if (supportsOrds) { long ord; try { ord = termsEnum.ord(); } catch (UnsupportedOperationException uoe) { supportsOrds = false; ord = -1; } if (ord != -1) { assertEquals(termCount, ord); } } termCount++; } assertEquals(termFreqs.size(), termCount); r.close(); dir.close(); }
From source file:com.rocana.lucene.codec.v1.TestBlockPostingsFormat3.java
License:Apache License
/** * checks docs + freqs + positions + payloads, sequentially *///from ww w . j a va2 s .c om public void assertDocsAndPositionsEnum(PostingsEnum leftDocs, PostingsEnum rightDocs) throws Exception { assertNotNull(leftDocs); assertNotNull(rightDocs); assertEquals(-1, leftDocs.docID()); assertEquals(-1, rightDocs.docID()); int docid; while ((docid = leftDocs.nextDoc()) != DocIdSetIterator.NO_MORE_DOCS) { assertEquals(docid, rightDocs.nextDoc()); int freq = leftDocs.freq(); assertEquals(freq, rightDocs.freq()); for (int i = 0; i < freq; i++) { assertEquals(leftDocs.nextPosition(), rightDocs.nextPosition()); // we don't assert offsets/payloads, they are allowed to be different } } assertEquals(DocIdSetIterator.NO_MORE_DOCS, rightDocs.nextDoc()); }
From source file:com.rocana.lucene.codec.v1.TestBlockPostingsFormat3.java
License:Apache License
/** * checks advancing docs + positions//from w ww .ja va 2 s.com */ public void assertPositionsSkipping(int docFreq, PostingsEnum leftDocs, PostingsEnum rightDocs) throws Exception { if (leftDocs == null || rightDocs == null) { assertNull(leftDocs); assertNull(rightDocs); return; } int docid = -1; int averageGap = MAXDOC / (1 + docFreq); int skipInterval = 16; while (true) { if (random().nextBoolean()) { // nextDoc() docid = leftDocs.nextDoc(); assertEquals(docid, rightDocs.nextDoc()); } else { // advance() int skip = docid + (int) Math.ceil(Math.abs(skipInterval + random().nextGaussian() * averageGap)); docid = leftDocs.advance(skip); assertEquals(docid, rightDocs.advance(skip)); } if (docid == DocIdSetIterator.NO_MORE_DOCS) { return; } int freq = leftDocs.freq(); assertEquals(freq, rightDocs.freq()); for (int i = 0; i < freq; i++) { assertEquals(leftDocs.nextPosition(), rightDocs.nextPosition()); // we don't compare the payloads, it's allowed that one is empty etc } } }
From source file:com.shaie.annots.AnnotationSearchExample.java
License:Apache License
public static void main(String[] args) throws Exception { Directory dir = new RAMDirectory(); IndexWriterConfig conf = new IndexWriterConfig(new WhitespaceAnalyzer()); IndexWriter writer = new IndexWriter(dir, conf); // we need to add the annotation as a TokenStream field, therefore cannot use an Analyzer passed in the // IndexWriterConfig. Tokenizer tokenizer = new WhitespaceTokenizer(); tokenizer.setReader(new StringReader("quick brown fox ate the blue red chicken")); TeeSinkTokenFilter textStream = new TeeSinkTokenFilter(tokenizer); TokenStream colorAnnotationStream = new AnnotatingTokenFilter( textStream.newSinkTokenStream(new ColorsSinkFilter()), COLOR_ANNOT_TERM); Document doc = new Document(); doc.add(new TextField("text", textStream)); doc.add(new TextField("annot", colorAnnotationStream)); writer.addDocument(doc);//from w ww.j a v a2 s . c o m writer.close(); DirectoryReader reader = DirectoryReader.open(dir); LeafReader ar = reader.leaves().get(0).reader(); // we only have one segment printFieldTerms(ar, "text"); System.out.println(); final ByteArrayDataInput in = new ByteArrayDataInput(); PostingsEnum dape = ar.postings(new Term("annot", COLOR_ANNOT_TERM)); int docID = dape.nextDoc(); int freq = dape.freq(); System.out.println("Color annotation spans: doc=" + docID + ", freq=" + freq); for (int i = 0; i < freq; i++) { dape.nextPosition(); BytesRef payload = dape.getPayload(); in.reset(payload.bytes, payload.offset, payload.length); System.out.println(" start=" + in.readVInt() + ", length=" + in.readVInt()); } IndexSearcher searcher = new IndexSearcher(reader); System.out.println("\nsearching for 'red WITHIN color':"); Query q = new SpanWithinQuery(new SpanAnnotationTermQuery(new Term("annot", COLOR_ANNOT_TERM)), new SpanInclusivePositionTermQuery(new Term("text", "red"))); TopDocs td = searcher.search(q, 10); System.out.println(" num results: " + td.scoreDocs.length); System.out.println("\nsearching for 'ate WITHIN color':"); q = new SpanWithinQuery(new SpanAnnotationTermQuery(new Term("annot", COLOR_ANNOT_TERM)), new SpanInclusivePositionTermQuery(new Term("text", "ate"))); td = searcher.search(q, 10); System.out.println(" num results: " + td.scoreDocs.length); reader.close(); dir.close(); }
From source file:com.shaie.annots.AnnotationsUtils.java
License:Apache License
public static void printAnnotations(LeafReader reader, Term term) throws IOException { System.out.println("Annotations for " + term); final ByteArrayDataInput in = new ByteArrayDataInput(); final PostingsEnum postings = reader.postings(term, PostingsEnum.PAYLOADS); for (int docID = postings.nextDoc(); docID != DocIdSetIterator.NO_MORE_DOCS; docID = postings.nextDoc()) { final int freq = postings.freq(); System.out.println(" doc=" + docID + ", freq=" + freq); for (int i = 0; i < freq; i++) { postings.nextPosition();//from w ww. ja v a 2 s . co m final BytesRef payload = postings.getPayload(); in.reset(payload.bytes, payload.offset, payload.length); System.out.println(" start=" + in.readVInt() + ", length=" + in.readVInt()); } } }
From source file:com.shaie.utils.IndexUtils.java
License:Apache License
/** Prints the terms indexed under the given fields with full postings information. */ public static void printFieldTermsWithInfo(LeafReader reader, String... fields) throws IOException { for (final String field : fields) { System.out.println(format("Terms for field [%s], with positional info:", field)); final TermsEnum te = reader.terms(field).iterator(); BytesRef scratch;/*from w w w.j a v a 2 s . com*/ PostingsEnum postings = null; while ((scratch = te.next()) != null) { System.out.println(format(" %s", scratch.utf8ToString())); postings = te.postings(postings, PostingsEnum.ALL); for (postings.nextDoc(); postings.docID() != DocIdSetIterator.NO_MORE_DOCS; postings.nextDoc()) { final Map<Integer, BytesRef> positions = Maps.newTreeMap(); boolean addedPayload = false; for (int i = 0; i < postings.freq(); i++) { final int pos = postings.nextPosition(); final BytesRef payload = postings.getPayload(); if (payload != null) { positions.put(pos, BytesRef.deepCopyOf(payload)); addedPayload = true; } else { positions.put(pos, null); } } if (addedPayload) { System.out.println( format(" doc=%d, freq=%d", postings.docID(), postings.freq(), positions)); for (final Entry<Integer, BytesRef> e : positions.entrySet()) { System.out.println(format(" pos=%d, payload=%s", e.getKey(), e.getValue())); } } else { System.out.println(format(" doc=%d, freq=%d, pos=%s", postings.docID(), postings.freq(), positions.keySet())); } } } } }
From source file:edu.upenn.library.solrplugins.ProofOfConceptPayloadHandler.java
License:Apache License
private NamedList<Object> buildEntryValue(long count, PostingsEnum postings, Bits liveDocs) throws IOException { NamedList<Object> entry = new NamedList<>(); entry.add("count", count); int i = -1;//from w w w .j a v a 2s . c o m while (postings.nextDoc() != DocIdSetIterator.NO_MORE_DOCS) { if (!liveDocs.get(postings.docID())) { continue; } i++; NamedList<Object> documentEntry = new NamedList<>(); entry.add("doc" + i, documentEntry); for (int j = 0; j < postings.freq(); j++) { postings.nextPosition(); String extra = postings.getPayload().utf8ToString(); documentEntry.add("position" + j, extra); } } return entry; }