Examples with TermDocs org.apache.lucene.index.TermDocs used on opensource projects

Example 6 with TermDocs

use of org.apache.lucene.index.TermDocs in project jackrabbit by apache.

the class TermDocsCache method termDocs.

/**
     * Returns the {@link TermDocs} for the given term.
     *
     * @param t the term.
     * @return the term docs for the given term.
     * @throws IOException if an error occurs while reading from the index.
     */
public TermDocs termDocs(final Term t) throws IOException {
    if (t == null || t.field() != field) {
        return reader.termDocs(t);
    }
    String text = t.text();
    if (unknownValues.get(text) != null) {
        log.debug("EmptyTermDocs({},{})", field, text);
        return EmptyTermDocs.INSTANCE;
    }
    // maintain cache
    CacheEntry entry;
    synchronized (cache) {
        entry = cache.get(text);
        if (entry == null) {
            // check space
            if (cache.size() >= CACHE_SIZE) {
                // prune half of them and adjust the rest
                CacheEntry[] entries = cache.values().toArray(new CacheEntry[cache.size()]);
                Arrays.sort(entries);
                int threshold = entries[CACHE_SIZE / 2].numAccessed;
                for (Iterator<Map.Entry<String, CacheEntry>> it = cache.entrySet().iterator(); it.hasNext(); ) {
                    Map.Entry<String, CacheEntry> e = it.next();
                    if (e.getValue().numAccessed <= threshold) {
                        // prune
                        it.remove();
                    } else {
                        // adjust
                        CacheEntry ce = e.getValue();
                        ce.numAccessed = (int) Math.sqrt(ce.numAccessed);
                    }
                }
            }
            entry = new CacheEntry();
            cache.put(text, entry);
        } else {
            entry.numAccessed++;
        }
    }
    // that are read only irregularly.
    if (entry.numAccessed < 10) {
        if (log.isDebugEnabled()) {
            log.debug("#{} TermDocs({},{})", new Object[] { entry.numAccessed, field, text });
        }
        return reader.termDocs(t);
    }
    if (entry.bits == null) {
        // collect bits
        BitSet bits = null;
        TermDocs tDocs = reader.termDocs(t);
        try {
            while (tDocs.next()) {
                if (bits == null) {
                    bits = new BitSet(reader.maxDoc());
                }
                bits.set(tDocs.doc());
            }
        } finally {
            tDocs.close();
        }
        if (bits != null) {
            entry.bits = bits;
        }
    }
    if (entry.bits == null) {
        // none collected
        unknownValues.put(text, text);
        return EmptyTermDocs.INSTANCE;
    } else {
        if (log.isDebugEnabled()) {
            log.debug("CachedTermDocs({},{},{}/{})", new Object[] { field, text, entry.bits.cardinality(), reader.maxDoc() });
        }
        return new CachedTermDocs(entry.bits);
    }
}

Example 7 with TermDocs

use of org.apache.lucene.index.TermDocs in project jackrabbit by apache.

the class MatchAllScorer method calculateDocFilter.

/**
     * Calculates a BitSet filter that includes all the nodes
     * that have content in properties according to the field name
     * passed in the constructor of this MatchAllScorer.
     *
     * @throws IOException if an error occurs while reading from
     *                     the search index.
     */
@SuppressWarnings({ "unchecked" })
private void calculateDocFilter(PerQueryCache cache) throws IOException {
    Map<String, BitSet> readerCache = (Map<String, BitSet>) cache.get(MatchAllScorer.class, reader);
    if (readerCache == null) {
        readerCache = new HashMap<String, BitSet>();
        cache.put(MatchAllScorer.class, reader, readerCache);
    }
    // get BitSet for field
    docFilter = readerCache.get(field);
    if (docFilter != null) {
        // use cached BitSet;
        return;
    }
    // otherwise calculate new
    docFilter = new BitSet(reader.maxDoc());
    // we match all terms
    String namedValue = FieldNames.createNamedValue(field, "");
    TermEnum terms = reader.terms(new Term(FieldNames.PROPERTIES, namedValue));
    try {
        TermDocs docs = reader.termDocs();
        try {
            while (terms.term() != null && terms.term().field() == FieldNames.PROPERTIES && terms.term().text().startsWith(namedValue)) {
                docs.seek(terms);
                while (docs.next()) {
                    docFilter.set(docs.doc());
                }
                terms.next();
            }
        } finally {
            docs.close();
        }
    } finally {
        terms.close();
    }
    // put BitSet into cache
    readerCache.put(field, docFilter);
}

Example 8 with TermDocs

use of org.apache.lucene.index.TermDocs in project greplin-lucene-utils by Cue.

the class PrefixFilter method getDocIdSet.

@Override
public DocIdSet getDocIdSet(final IndexReader reader) throws IOException {
    FixedBitSet result = new FixedBitSet(reader.maxDoc());
    Term term = new Term(this.field, this.prefix);
    TermEnum te = reader.terms(term);
    TermDocs td = reader.termDocs();
    try {
        do {
            term = te.term();
            // noinspection StringEquality
            if (term == null || this.field != term.field() || !term.text().startsWith(this.prefix)) {
                break;
            }
            td.seek(te.term());
            while (td.next()) {
                result.set(td.doc());
            }
        } while (te.next());
    } finally {
        te.close();
        td.close();
    }
    return result;
}

Example 9 with TermDocs

use of org.apache.lucene.index.TermDocs in project greplin-lucene-utils by Cue.

the class TermsFilter method getDocIdSet.

@Override
public DocIdSet getDocIdSet(final IndexReader reader) throws IOException {
    FixedBitSet result = new FixedBitSet(reader.maxDoc());
    TermDocs td = reader.termDocs();
    try {
        for (Term term : this.terms) {
            td.seek(term);
            while (td.next()) {
                result.set(td.doc());
            }
        }
    } finally {
        td.close();
    }
    return result;
}

Example 10 with TermDocs

use of org.apache.lucene.index.TermDocs in project greplin-lucene-utils by Cue.

the class FilteredMultiReader method termDocs.

@Override
public TermDocs termDocs(final Term term) throws IOException {
    TermDocs result = termDocs();
    result.seek(term);
    return result;
}