Examples with ExtractedTextCache org.apache.jackrabbit.oak.plugins.index.search.ExtractedTextCache used on opensource projects

Example 1 with ExtractedTextCache

use of org.apache.jackrabbit.oak.plugins.index.search.ExtractedTextCache in project jackrabbit-oak by apache.

the class LuceneIndexEditorTest method multiplexingWriter.

@Test
public void multiplexingWriter() throws Exception {
    newLucenePropertyIndex("lucene", "foo");
    MountInfoProvider mip = Mounts.newBuilder().mount("foo", "/libs", "/apps").build();
    EditorHook hook = new EditorHook(new IndexUpdateProvider(new LuceneIndexEditorProvider(null, new ExtractedTextCache(0, 0), null, mip)));
    NodeState indexed = hook.processCommit(EMPTY_NODE, builder.getNodeState(), CommitInfo.EMPTY);
    builder = indexed.builder();
    NodeState before = indexed;
    builder.child("content").child("en").setProperty("foo", "bar");
    builder.child("libs").child("install").setProperty("foo", "bar");
    NodeState after = builder.getNodeState();
    indexed = hook.processCommit(before, after, CommitInfo.EMPTY);
    builder = indexed.builder();
    assertEquals(1, numDocs(mip.getMountByName("foo")));
    assertEquals(1, numDocs(mip.getDefaultMount()));
}

Example 2 with ExtractedTextCache

use of org.apache.jackrabbit.oak.plugins.index.search.ExtractedTextCache in project jackrabbit-oak by apache.

the class ElasticIndexImporterSupport method createElasticEditorProvider.

private IndexEditorProvider createElasticEditorProvider() {
    final ElasticConnection.Builder.BuildStep buildStep = ElasticConnection.newBuilder().withIndexPrefix(indexPrefix).withConnectionParameters(scheme, host, port);
    final ElasticConnection connection;
    if (apiKeyId != null && apiSecretId != null) {
        connection = buildStep.withApiKeys(apiKeyId, apiSecretId).build();
    } else {
        connection = buildStep.build();
    }
    closer.register(connection);
    ElasticIndexTracker indexTracker = new ElasticIndexTracker(connection, new ElasticMetricHandler(StatisticsProvider.NOOP));
    ElasticIndexEditorProvider editorProvider = new ElasticIndexEditorProvider(indexTracker, connection, new ExtractedTextCache(10 * FileUtils.ONE_MB, 100));
    return editorProvider;
}

Example 3 with ExtractedTextCache

use of org.apache.jackrabbit.oak.plugins.index.search.ExtractedTextCache in project jackrabbit-oak by apache.

the class ElasticOutOfBandIndexer method createElasticEditorProvider.

private IndexEditorProvider createElasticEditorProvider() {
    final ElasticConnection.Builder.BuildStep buildStep = ElasticConnection.newBuilder().withIndexPrefix(indexPrefix).withConnectionParameters(scheme, host, port);
    final ElasticConnection connection;
    if (apiKeyId != null && apiSecretId != null) {
        connection = buildStep.withApiKeys(apiKeyId, apiSecretId).build();
    } else {
        connection = buildStep.build();
    }
    closer.register(connection);
    ElasticIndexTracker indexTracker = new ElasticIndexTracker(connection, new ElasticMetricHandler(StatisticsProvider.NOOP));
    return new ElasticIndexEditorProvider(indexTracker, connection, new ExtractedTextCache(10 * FileUtils.ONE_MB, 100));
}

Example 4 with ExtractedTextCache

use of org.apache.jackrabbit.oak.plugins.index.search.ExtractedTextCache in project jackrabbit-oak by apache.

the class ActiveDeletedBlobCollectorMBeanImplTest method clonedNSWithSharedDS.

@Test
public void clonedNSWithSharedDS() throws Exception {
    MemoryBlobStore bs = new MemoryBlobStore();
    bs.setBlockSizeMin(48);
    MemoryDocumentStore mds1 = new MemoryDocumentStore();
    DocumentNodeStore dns1 = builderProvider.newBuilder().setDocumentStore(mds1).setBlobStore(bs).build();
    // Create initial repo with InitialContent. It has enough data to create blobs
    LuceneIndexEditorProvider editorProvider = new LuceneIndexEditorProvider();
    ContentRepository repository = new Oak(dns1).with(new InitialContent()).with(new OpenSecurityProvider()).with(editorProvider).with(new PropertyIndexEditorProvider()).with(new NodeTypeIndexProvider()).createContentRepository();
    ContentSession session = repository.login(null, null);
    Root root = session.getLatestRoot();
    TestUtil.createFulltextIndex(root.getTree("/"), "testIndex");
    root.commit();
    // pause active deletion
    IndexPathService indexPathService = new IndexPathServiceImpl(dns1);
    AsyncIndexInfoService asyncIndexInfoService = MockRegistrar.getAsyncIndexInfoService(newArrayList(new IndexMBeanInfoSupplier("foo-async", () -> STATUS_DONE, () -> 2L)));
    ActiveDeletedBlobCollectorMBeanImpl bean = new ActiveDeletedBlobCollectorMBeanImpl(ActiveDeletedBlobCollectorFactory.NOOP, wb, dns1, indexPathService, asyncIndexInfoService, new MemoryBlobStore(), sameThreadExecutor());
    bean.clock = clock;
    bean.flagActiveDeletionUnsafeForCurrentState();
    // we try here to create some churn and we want some files to get created at dns1
    // BUT get deleted at dns2. "segments_1" is one such file.
    // since our "creation" of churn is assumed, we should assert that dns1 has "segments_1"
    // (and later dns2 doesn't have it)
    root = session.getLatestRoot();
    assertTrue("First pass indexing should generate segments_1", root.getTree("/oak:index/testIndex/:data/segments_1").exists());
    // shutdown first instance
    dns1.dispose();
    // clone
    MemoryDocumentStore mds2 = mds1.copy();
    DocumentNodeStore dns2 = builderProvider.newBuilder().setDocumentStore(mds2).setBlobStore(bs).build();
    // create some churn to delete some index files - using clone store
    // we'd setup lucene editor with active deletion collector
    DeletedFileTrackingADBC deletedFileTrackingADBC = new DeletedFileTrackingADBC(new File(temporaryFolder.getRoot(), "adbc-workdir"));
    editorProvider = new LuceneIndexEditorProvider(null, null, new ExtractedTextCache(0, 0), null, Mounts.defaultMountInfoProvider(), deletedFileTrackingADBC, null, null);
    repository = new Oak(dns2).with(new OpenSecurityProvider()).with(editorProvider).with(new PropertyIndexEditorProvider()).with(new NodeTypeIndexProvider()).createContentRepository();
    session = repository.login(null, null);
    root = session.getLatestRoot();
    Tree rootTree = root.getTree("/");
    for (int i = 0; i < 20; i++) {
        Tree child = rootTree.addChild("a" + i);
        for (int j = 0; j < 20; j++) {
            child.setProperty("foo" + j, "bar" + j);
        }
    }
    root.commit();
    // since our index is not async, we are unable to track commit progress automatically.
    // OR, iow, we need to play the rold of AsyncIndexUpdate explicitly
    deletedFileTrackingADBC.blobDeletionCallback.commitProgress(COMMIT_SUCCEDED);
    deletedFileTrackingADBC.purgeBlobsDeleted(Clock.SIMPLE.getTime() + TimeUnit.SECONDS.toMillis(1), bs);
    root = session.getLatestRoot();
    assertFalse("Churn created via dns2 should delete segments_1", root.getTree("/oak:index/testIndex/:data/segments_1").exists());
    dns2.dispose();
    // validate index using dns1 which should still have valid index data even
    // after dns2's churn
    dns1 = builderProvider.newBuilder().setDocumentStore(mds1).setBlobStore(bs).build();
    IndexConsistencyChecker checker = new IndexConsistencyChecker(dns1.getRoot(), "/oak:index/testIndex", new File(temporaryFolder.getRoot(), "checker-workdir"));
    IndexConsistencyChecker.Result result = checker.check(IndexConsistencyChecker.Level.BLOBS_ONLY);
    assertFalse("Nodestore1 can't read blobs: " + result.missingBlobIds + " while reading index", result.missingBlobs);
}

Example 5 with ExtractedTextCache

use of org.apache.jackrabbit.oak.plugins.index.search.ExtractedTextCache in project jackrabbit-oak by apache.

the class ElasticFacetTest method createRepository.

private void createRepository() throws RepositoryException {
    ElasticConnection connection = elasticRule.useDocker() ? elasticRule.getElasticConnectionForDocker() : elasticRule.getElasticConnectionFromString();
    ElasticIndexTracker indexTracker = new ElasticIndexTracker(connection, new ElasticMetricHandler(StatisticsProvider.NOOP));
    ElasticIndexEditorProvider editorProvider = new ElasticIndexEditorProvider(indexTracker, connection, new ExtractedTextCache(10 * FileUtils.ONE_MB, 100));
    ElasticIndexProvider indexProvider = new ElasticIndexProvider(indexTracker);
    NodeStore nodeStore = new MemoryNodeStore(INITIAL_CONTENT);
    Oak oak = new Oak(nodeStore).with(editorProvider).with(indexTracker).with(indexProvider);
    Jcr jcr = new Jcr(oak);
    Repository repository = jcr.createRepository();
    adminSession = repository.login(new SimpleCredentials("admin", "admin".toCharArray()), null);
    // we'd always query anonymously
    anonymousSession = repository.login(new GuestCredentials(), null);
    anonymousSession.refresh(true);
    anonymousSession.save();
    qe = anonymousSession.getWorkspace().getQueryManager();
}