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

Example 11 with TermDocs

use of org.apache.lucene.index.TermDocs in project zm-mailbox by Zimbra.

the class RemoteMailQueue method summarize.

private void summarize(SearchResult result, IndexReader indexReader) throws IOException {
    TermEnum terms = indexReader.terms();
    boolean hasDeletions = indexReader.hasDeletions();
    do {
        Term term = terms.term();
        if (term != null) {
            String field = term.field();
            if (field != null && field.length() > 0) {
                QueueAttr attr = QueueAttr.valueOf(field);
                if (attr == QueueAttr.addr || attr == QueueAttr.host || attr == QueueAttr.from || attr == QueueAttr.to || attr == QueueAttr.fromdomain || attr == QueueAttr.todomain || attr == QueueAttr.reason || attr == QueueAttr.received) {
                    List<SummaryItem> list = result.sitems.get(attr);
                    if (list == null) {
                        list = new LinkedList<SummaryItem>();
                        result.sitems.put(attr, list);
                    }
                    int count = 0;
                    if (hasDeletions) {
                        TermDocs termDocs = indexReader.termDocs(term);
                        while (termDocs.next()) {
                            if (!indexReader.isDeleted(termDocs.doc())) {
                                count++;
                            }
                        }
                    } else {
                        count = terms.docFreq();
                    }
                    if (count > 0) {
                        list.add(new SummaryItem(term.text(), count));
                    }
                }
            }
        }
    } while (terms.next());
}

Example 12 with TermDocs

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

the class AbstractExcerpt method getExcerpt.

/**
     * {@inheritDoc}
     */
public String getExcerpt(NodeId id, int maxFragments, int maxFragmentSize) throws IOException {
    IndexReader reader = index.getIndexReader();
    try {
        checkRewritten(reader);
        Term idTerm = TermFactory.createUUIDTerm(id.toString());
        TermDocs tDocs = reader.termDocs(idTerm);
        int docNumber;
        Document doc;
        try {
            if (tDocs.next()) {
                docNumber = tDocs.doc();
                doc = reader.document(docNumber);
            } else {
                // node not found in index
                return null;
            }
        } finally {
            tDocs.close();
        }
        Fieldable[] fields = doc.getFieldables(FieldNames.FULLTEXT);
        if (fields.length == 0) {
            log.debug("Fulltext field not stored, using {}", SimpleExcerptProvider.class.getName());
            SimpleExcerptProvider exProvider = new SimpleExcerptProvider();
            exProvider.init(query, index);
            return exProvider.getExcerpt(id, maxFragments, maxFragmentSize);
        }
        StringBuffer text = new StringBuffer();
        String separator = "";
        for (int i = 0; i < fields.length; i++) {
            if (fields[i].stringValue().length() == 0) {
                continue;
            }
            text.append(separator);
            text.append(fields[i].stringValue());
            separator = " ";
        }
        TermFreqVector tfv = reader.getTermFreqVector(docNumber, FieldNames.FULLTEXT);
        if (tfv instanceof TermPositionVector) {
            return createExcerpt((TermPositionVector) tfv, text.toString(), maxFragments, maxFragmentSize);
        } else {
            log.debug("No TermPositionVector on Fulltext field.");
            return null;
        }
    } finally {
        Util.closeOrRelease(reader);
    }
}

Example 13 with TermDocs

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

the class CachingIndexReader method getParent.

/**
     * Returns the <code>DocId</code> of the parent of <code>n</code> or
     * {@link DocId#NULL} if <code>n</code> does not have a parent
     * (<code>n</code> is the root node).
     *
     * @param n the document number.
     * @param deleted the documents that should be regarded as deleted.
     * @return the <code>DocId</code> of <code>n</code>'s parent.
     * @throws IOException if an error occurs while reading from the index.
     */
DocId getParent(int n, BitSet deleted) throws IOException {
    DocId parent;
    boolean existing = false;
    int parentDocNum = inSegmentParents[n];
    if (parentDocNum != -1) {
        parent = DocId.create(parentDocNum);
    } else {
        parent = foreignParentDocIds.get(n);
    }
    if (parent != null) {
        existing = true;
        // check if valid and reset if necessary
        if (!parent.isValid(deleted)) {
            if (log.isDebugEnabled()) {
                log.debug(parent + " not valid anymore.");
            }
            parent = null;
        }
    }
    if (parent == null) {
        int plainDocId = -1;
        Document doc = document(n, FieldSelectors.UUID_AND_PARENT);
        String[] parentUUIDs = doc.getValues(FieldNames.PARENT);
        if (parentUUIDs.length == 0 || parentUUIDs[0].length() == 0) {
            // root node
            parent = DocId.NULL;
        } else {
            if (shareableNodes.get(n)) {
                parent = DocId.create(parentUUIDs);
            } else {
                if (!existing) {
                    Term id = TermFactory.createUUIDTerm(parentUUIDs[0]);
                    TermDocs docs = termDocs(id);
                    try {
                        while (docs.next()) {
                            if (!deleted.get(docs.doc())) {
                                plainDocId = docs.doc();
                                parent = DocId.create(plainDocId);
                                break;
                            }
                        }
                    } finally {
                        docs.close();
                    }
                }
                // DocId was invalid. thus, only allowed to create DocId from uuid
                if (parent == null) {
                    parent = DocId.create(parentUUIDs[0]);
                }
            }
        }
        // finally put to cache
        if (plainDocId != -1) {
            // PlainDocId
            inSegmentParents[n] = plainDocId;
        } else {
            // UUIDDocId
            foreignParentDocIds.put(n, parent);
            if (existing) {
                // there was an existing parent reference in
                // inSegmentParents, which was invalid and is replaced
                // with a UUIDDocId (points to a foreign segment).
                // mark as unknown
                inSegmentParents[n] = -1;
            }
        }
    }
    return parent;
}

Example 14 with TermDocs

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

the class IndexingQueue method initialize.

/**
     * Initializes the indexing queue.
     *
     * @param index the multi index this indexing queue belongs to.
     * @throws IOException if an error occurs while reading from the index.
     */
void initialize(MultiIndex index) throws IOException {
    if (initialized) {
        throw new IllegalStateException("already initialized");
    }
    // check index for nodes that need to be reindexed
    CachingMultiIndexReader reader = index.getIndexReader();
    try {
        TermDocs tDocs = reader.termDocs(new Term(FieldNames.REINDEXING_REQUIRED, ""));
        try {
            while (tDocs.next()) {
                queueStore.addUUID(reader.document(tDocs.doc(), FieldSelectors.UUID).get(FieldNames.UUID));
            }
        } finally {
            tDocs.close();
        }
    } finally {
        reader.release();
    }
    String[] uuids = queueStore.getPending();
    for (String uuid : uuids) {
        try {
            Document doc = index.createDocument(new NodeId(uuid));
            pendingDocuments.put(uuid, doc);
            log.debug("added node {}. New size of indexing queue: {}", uuid, pendingDocuments.size());
        } catch (IllegalArgumentException e) {
            log.warn("Invalid UUID in indexing queue store: " + uuid);
        } catch (RepositoryException e) {
            // node does not exist anymore
            log.debug("Node with uuid {} does not exist anymore", uuid);
            queueStore.removeUUID(uuid);
        }
    }
    initialized = true;
}

Example 15 with TermDocs

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

the class SharedFieldCache method getValueIndex.

/**
     * Creates a <code>ValueIndex</code> for a <code>field</code> and a term
     * <code>prefix</code>. The term prefix acts as the property name for the
     * shared <code>field</code>.
     * <p>
     * This method is an adapted version of: <code>FieldCacheImpl.getStringIndex()</code>
     *
     * @param reader     the <code>IndexReader</code>.
     * @param field      name of the shared field.
     * @param prefix     the property name, will be used as term prefix.
     * @return a ValueIndex that contains the field values and order
     *         information.
     * @throws IOException if an error occurs while reading from the index.
     */
public ValueIndex getValueIndex(IndexReader reader, String field, String prefix) throws IOException {
    if (reader instanceof ReadOnlyIndexReader) {
        reader = ((ReadOnlyIndexReader) reader).getBase();
    }
    field = field.intern();
    ValueIndex ret = lookup(reader, field, prefix);
    if (ret == null) {
        final int maxDocs = reader.maxDoc();
        Comparable<?>[] retArray = new Comparable<?>[maxDocs];
        Map<Integer, Integer> positions = new HashMap<Integer, Integer>();
        boolean usingSimpleComparable = true;
        int setValues = 0;
        if (maxDocs > 0) {
            IndexFormatVersion version = IndexFormatVersion.getVersion(reader);
            boolean hasPayloads = version.isAtLeast(IndexFormatVersion.V3);
            TermDocs termDocs;
            byte[] payload = null;
            int type;
            if (hasPayloads) {
                termDocs = reader.termPositions();
                payload = new byte[1];
            } else {
                termDocs = reader.termDocs();
            }
            TermEnum termEnum = reader.terms(new Term(field, prefix));
            try {
                if (termEnum.term() == null) {
                    throw new RuntimeException("no terms in field " + field);
                }
                do {
                    Term term = termEnum.term();
                    if (term.field() != field || !term.text().startsWith(prefix)) {
                        break;
                    }
                    final String value = termValueAsString(term, prefix);
                    termDocs.seek(term);
                    while (termDocs.next()) {
                        int termPosition = 0;
                        type = PropertyType.UNDEFINED;
                        if (hasPayloads) {
                            TermPositions termPos = (TermPositions) termDocs;
                            termPosition = termPos.nextPosition();
                            if (termPos.isPayloadAvailable()) {
                                payload = termPos.getPayload(payload, 0);
                                type = PropertyMetaData.fromByteArray(payload).getPropertyType();
                            }
                        }
                        setValues++;
                        Comparable<?> v = getValue(value, type);
                        int doc = termDocs.doc();
                        Comparable<?> ca = retArray[doc];
                        if (ca == null) {
                            if (usingSimpleComparable) {
                                // put simple value on the queue
                                positions.put(doc, termPosition);
                                retArray[doc] = v;
                            } else {
                                retArray[doc] = new ComparableArray(v, termPosition);
                            }
                        } else {
                            if (ca instanceof ComparableArray) {
                                ((ComparableArray) ca).insert(v, termPosition);
                            } else {
                                // Comparable to ComparableArray
                                for (int pos : positions.keySet()) {
                                    retArray[pos] = new ComparableArray(retArray[pos], positions.get(pos));
                                }
                                positions = null;
                                usingSimpleComparable = false;
                                ComparableArray caNew = (ComparableArray) retArray[doc];
                                retArray[doc] = caNew.insert(v, termPosition);
                            }
                        }
                    }
                } while (termEnum.next());
            } finally {
                termDocs.close();
                termEnum.close();
            }
        }
        ValueIndex value = new ValueIndex(retArray, setValues);
        store(reader, field, prefix, value);
        return value;
    }
    return ret;
}