Example usage for org.apache.lucene.index SortedDocValues termsEnum

Introduction

In this page you can find the example usage for org.apache.lucene.index SortedDocValues termsEnum.

Prototype

public TermsEnum termsEnum() throws IOException 

Source Link

Document

Returns a TermsEnum over the values.

Usage

From source file:org.apache.solr.uninverting.TestFieldCache.java

License:Apache License

public void test() throws IOException {
    FieldCache cache = FieldCache.DEFAULT;
    NumericDocValues doubles = cache.getNumerics(reader, "theDouble", FieldCache.DOUBLE_POINT_PARSER);
    for (int i = 0; i < NUM_DOCS; i++) {
        assertEquals(i, doubles.nextDoc());
        assertEquals(Double.doubleToLongBits(Double.MAX_VALUE - i), doubles.longValue());
    }/*from   w  w  w .j  av a  2  s  . co  m*/

    NumericDocValues longs = cache.getNumerics(reader, "theLong", FieldCache.LONG_POINT_PARSER);
    for (int i = 0; i < NUM_DOCS; i++) {
        assertEquals(i, longs.nextDoc());
        assertEquals(Long.MAX_VALUE - i, longs.longValue());
    }

    NumericDocValues ints = cache.getNumerics(reader, "theInt", FieldCache.INT_POINT_PARSER);
    for (int i = 0; i < NUM_DOCS; i++) {
        assertEquals(i, ints.nextDoc());
        assertEquals(Integer.MAX_VALUE - i, ints.longValue());
    }

    NumericDocValues floats = cache.getNumerics(reader, "theFloat", FieldCache.FLOAT_POINT_PARSER);
    for (int i = 0; i < NUM_DOCS; i++) {
        assertEquals(i, floats.nextDoc());
        assertEquals(Float.floatToIntBits(Float.MAX_VALUE - i), floats.longValue());
    }

    Bits docsWithField = cache.getDocsWithField(reader, "theLong", FieldCache.LONG_POINT_PARSER);
    assertSame("Second request to cache return same array", docsWithField,
            cache.getDocsWithField(reader, "theLong", FieldCache.LONG_POINT_PARSER));
    assertTrue("docsWithField(theLong) must be class Bits.MatchAllBits",
            docsWithField instanceof Bits.MatchAllBits);
    assertTrue("docsWithField(theLong) Size: " + docsWithField.length() + " is not: " + NUM_DOCS,
            docsWithField.length() == NUM_DOCS);
    for (int i = 0; i < docsWithField.length(); i++) {
        assertTrue(docsWithField.get(i));
    }

    docsWithField = cache.getDocsWithField(reader, "sparse", FieldCache.INT_POINT_PARSER);
    assertSame("Second request to cache return same array", docsWithField,
            cache.getDocsWithField(reader, "sparse", FieldCache.INT_POINT_PARSER));
    assertFalse("docsWithField(sparse) must not be class Bits.MatchAllBits",
            docsWithField instanceof Bits.MatchAllBits);
    assertTrue("docsWithField(sparse) Size: " + docsWithField.length() + " is not: " + NUM_DOCS,
            docsWithField.length() == NUM_DOCS);
    for (int i = 0; i < docsWithField.length(); i++) {
        assertEquals(i % 2 == 0, docsWithField.get(i));
    }

    // getTermsIndex
    SortedDocValues termsIndex = cache.getTermsIndex(reader, "theRandomUnicodeString");
    for (int i = 0; i < NUM_DOCS; i++) {
        final String s;
        if (i > termsIndex.docID()) {
            termsIndex.advance(i);
        }
        if (i == termsIndex.docID()) {
            s = termsIndex.binaryValue().utf8ToString();
        } else {
            s = null;
        }
        assertTrue("for doc " + i + ": " + s + " does not equal: " + unicodeStrings[i],
                unicodeStrings[i] == null || unicodeStrings[i].equals(s));
    }

    int nTerms = termsIndex.getValueCount();

    TermsEnum tenum = termsIndex.termsEnum();
    for (int i = 0; i < nTerms; i++) {
        BytesRef val1 = BytesRef.deepCopyOf(tenum.next());
        final BytesRef val = termsIndex.lookupOrd(i);
        // System.out.println("i="+i);
        assertEquals(val, val1);
    }

    // seek the enum around (note this isn't a great test here)
    int num = atLeast(100);
    for (int i = 0; i < num; i++) {
        int k = random().nextInt(nTerms);
        final BytesRef val = BytesRef.deepCopyOf(termsIndex.lookupOrd(k));
        assertEquals(TermsEnum.SeekStatus.FOUND, tenum.seekCeil(val));
        assertEquals(val, tenum.term());
    }

    for (int i = 0; i < nTerms; i++) {
        final BytesRef val = BytesRef.deepCopyOf(termsIndex.lookupOrd(i));
        assertEquals(TermsEnum.SeekStatus.FOUND, tenum.seekCeil(val));
        assertEquals(val, tenum.term());
    }

    // test bad field
    termsIndex = cache.getTermsIndex(reader, "bogusfield");

    // getTerms
    BinaryDocValues terms = cache.getTerms(reader, "theRandomUnicodeString");
    for (int i = 0; i < NUM_DOCS; i++) {
        if (terms.docID() < i) {
            terms.nextDoc();
        }
        if (terms.docID() == i) {
            assertEquals(unicodeStrings[i], terms.binaryValue().utf8ToString());
        } else {
            assertNull(unicodeStrings[i]);
        }
    }

    // test bad field
    terms = cache.getTerms(reader, "bogusfield");

    // getDocTermOrds
    SortedSetDocValues termOrds = cache.getDocTermOrds(reader, "theRandomUnicodeMultiValuedField", null);
    int numEntries = cache.getCacheEntries().length;
    // ask for it again, and check that we didnt create any additional entries:
    termOrds = cache.getDocTermOrds(reader, "theRandomUnicodeMultiValuedField", null);
    assertEquals(numEntries, cache.getCacheEntries().length);

    for (int i = 0; i < NUM_DOCS; i++) {
        // This will remove identical terms. A DocTermOrds doesn't return duplicate ords for a docId
        List<BytesRef> values = new ArrayList<>(new LinkedHashSet<>(Arrays.asList(multiValued[i])));
        for (BytesRef v : values) {
            if (v == null) {
                // why does this test use null values... instead of an empty list: confusing
                break;
            }
            if (i > termOrds.docID()) {
                assertEquals(i, termOrds.nextDoc());
            }
            long ord = termOrds.nextOrd();
            assert ord != SortedSetDocValues.NO_MORE_ORDS;
            BytesRef scratch = termOrds.lookupOrd(ord);
            assertEquals(v, scratch);
        }
        if (i == termOrds.docID()) {
            assertEquals(SortedSetDocValues.NO_MORE_ORDS, termOrds.nextOrd());
        }
    }

    // test bad field
    termOrds = cache.getDocTermOrds(reader, "bogusfield", null);
    assertTrue(termOrds.getValueCount() == 0);

    FieldCache.DEFAULT.purgeByCacheKey(reader.getCoreCacheKey());
}

---

From source file:org.apache.solr.uninverting.TestFieldCacheVsDocValues.java

License:Apache License

private void assertEquals(int maxDoc, SortedDocValues expected, SortedDocValues actual) throws Exception {
    // can be null for the segment if no docs actually had any SortedDocValues
    // in this case FC.getDocTermsOrds returns EMPTY
    if (actual == null) {
        assertEquals(expected.getValueCount(), 0);
        return;/*from  ww w  .j av  a  2  s  .  c  om*/
    }
    assertEquals(expected.getValueCount(), actual.getValueCount());

    // compare ord lists
    while (true) {
        int docID = expected.nextDoc();
        if (docID == NO_MORE_DOCS) {
            assertEquals(NO_MORE_DOCS, actual.nextDoc());
            break;
        }
        assertEquals(docID, actual.nextDoc());
        assertEquals(expected.ordValue(), actual.ordValue());
        assertEquals(expected.binaryValue(), actual.binaryValue());
    }

    // compare ord dictionary
    for (long i = 0; i < expected.getValueCount(); i++) {
        final BytesRef expectedBytes = BytesRef.deepCopyOf(expected.lookupOrd((int) i));
        final BytesRef actualBytes = actual.lookupOrd((int) i);
        assertEquals(expectedBytes, actualBytes);
    }

    // compare termsenum
    assertEquals(expected.getValueCount(), expected.termsEnum(), actual.termsEnum());
}

---