Search in sources :

Example 36 with Directory

use of org.apache.lucene.store.Directory in project elasticsearch by elastic.

the class StoreTests method testCheckIntegrity.

public void testCheckIntegrity() throws IOException {
    Directory dir = newDirectory();
    long luceneFileLength = 0;
    try (IndexOutput output = dir.createOutput("lucene_checksum.bin", IOContext.DEFAULT)) {
        int iters = scaledRandomIntBetween(10, 100);
        for (int i = 0; i < iters; i++) {
            BytesRef bytesRef = new BytesRef(TestUtil.randomRealisticUnicodeString(random(), 10, 1024));
            output.writeBytes(bytesRef.bytes, bytesRef.offset, bytesRef.length);
            luceneFileLength += bytesRef.length;
        }
        CodecUtil.writeFooter(output);
        luceneFileLength += CodecUtil.footerLength();
    }
    final long luceneChecksum;
    try (IndexInput indexInput = dir.openInput("lucene_checksum.bin", IOContext.DEFAULT)) {
        assertEquals(luceneFileLength, indexInput.length());
        luceneChecksum = CodecUtil.retrieveChecksum(indexInput);
    }
    dir.close();
}
Also used : ChecksumIndexInput(org.apache.lucene.store.ChecksumIndexInput) IndexInput(org.apache.lucene.store.IndexInput) IndexOutput(org.apache.lucene.store.IndexOutput) BytesRef(org.apache.lucene.util.BytesRef) Directory(org.apache.lucene.store.Directory) RAMDirectory(org.apache.lucene.store.RAMDirectory)

Example 37 with Directory

use of org.apache.lucene.store.Directory in project elasticsearch by elastic.

the class StoreTests method testVerifyingIndexInput.

public void testVerifyingIndexInput() throws IOException {
    Directory dir = newDirectory();
    IndexOutput output = dir.createOutput("foo.bar", IOContext.DEFAULT);
    int iters = scaledRandomIntBetween(10, 100);
    for (int i = 0; i < iters; i++) {
        BytesRef bytesRef = new BytesRef(TestUtil.randomRealisticUnicodeString(random(), 10, 1024));
        output.writeBytes(bytesRef.bytes, bytesRef.offset, bytesRef.length);
    }
    CodecUtil.writeFooter(output);
    output.close();
    // Check file
    IndexInput indexInput = dir.openInput("foo.bar", IOContext.DEFAULT);
    long checksum = CodecUtil.retrieveChecksum(indexInput);
    indexInput.seek(0);
    IndexInput verifyingIndexInput = new Store.VerifyingIndexInput(dir.openInput("foo.bar", IOContext.DEFAULT));
    readIndexInputFullyWithRandomSeeks(verifyingIndexInput);
    Store.verify(verifyingIndexInput);
    assertThat(checksum, equalTo(((ChecksumIndexInput) verifyingIndexInput).getChecksum()));
    IOUtils.close(indexInput, verifyingIndexInput);
    // Corrupt file and check again
    corruptFile(dir, "foo.bar", "foo1.bar");
    verifyingIndexInput = new Store.VerifyingIndexInput(dir.openInput("foo1.bar", IOContext.DEFAULT));
    readIndexInputFullyWithRandomSeeks(verifyingIndexInput);
    try {
        Store.verify(verifyingIndexInput);
        fail("should be a corrupted index");
    } catch (CorruptIndexException | IndexFormatTooOldException | IndexFormatTooNewException ex) {
    // ok
    }
    IOUtils.close(verifyingIndexInput);
    IOUtils.close(dir);
}
Also used : ChecksumIndexInput(org.apache.lucene.store.ChecksumIndexInput) CorruptIndexException(org.apache.lucene.index.CorruptIndexException) IndexOutput(org.apache.lucene.store.IndexOutput) IndexFormatTooOldException(org.apache.lucene.index.IndexFormatTooOldException) ChecksumIndexInput(org.apache.lucene.store.ChecksumIndexInput) IndexInput(org.apache.lucene.store.IndexInput) IndexFormatTooNewException(org.apache.lucene.index.IndexFormatTooNewException) BytesRef(org.apache.lucene.util.BytesRef) Directory(org.apache.lucene.store.Directory) RAMDirectory(org.apache.lucene.store.RAMDirectory)

Example 38 with Directory

use of org.apache.lucene.store.Directory in project elasticsearch by elastic.

the class StoreTests method testCanOpenIndex.

public void testCanOpenIndex() throws IOException {
    final ShardId shardId = new ShardId("index", "_na_", 1);
    IndexWriterConfig iwc = newIndexWriterConfig();
    Path tempDir = createTempDir();
    final BaseDirectoryWrapper dir = newFSDirectory(tempDir);
    assertFalse(Store.canOpenIndex(logger, tempDir, shardId, (id, l) -> new DummyShardLock(id)));
    IndexWriter writer = new IndexWriter(dir, iwc);
    Document doc = new Document();
    doc.add(new StringField("id", "1", random().nextBoolean() ? Field.Store.YES : Field.Store.NO));
    writer.addDocument(doc);
    writer.commit();
    writer.close();
    assertTrue(Store.canOpenIndex(logger, tempDir, shardId, (id, l) -> new DummyShardLock(id)));
    DirectoryService directoryService = new DirectoryService(shardId, INDEX_SETTINGS) {

        @Override
        public Directory newDirectory() throws IOException {
            return dir;
        }
    };
    Store store = new Store(shardId, INDEX_SETTINGS, directoryService, new DummyShardLock(shardId));
    store.markStoreCorrupted(new CorruptIndexException("foo", "bar"));
    assertFalse(Store.canOpenIndex(logger, tempDir, shardId, (id, l) -> new DummyShardLock(id)));
    store.close();
}
Also used : ShardId(org.elasticsearch.index.shard.ShardId) Path(java.nio.file.Path) IndexNotFoundException(org.apache.lucene.index.IndexNotFoundException) ShardId(org.elasticsearch.index.shard.ShardId) NoMergePolicy(org.apache.lucene.index.NoMergePolicy) NoSuchFileException(java.nio.file.NoSuchFileException) Arrays(java.util.Arrays) IndexFormatTooNewException(org.apache.lucene.index.IndexFormatTooNewException) IndexSettingsModule(org.elasticsearch.test.IndexSettingsModule) Date(java.util.Date) Term(org.apache.lucene.index.Term) AlreadyClosedException(org.apache.lucene.store.AlreadyClosedException) Random(java.util.Random) ChecksumIndexInput(org.apache.lucene.store.ChecksumIndexInput) CorruptIndexException(org.apache.lucene.index.CorruptIndexException) InputStreamStreamInput(org.elasticsearch.common.io.stream.InputStreamStreamInput) Document(org.apache.lucene.document.Document) Settings(org.elasticsearch.common.settings.Settings) ByteArrayInputStream(java.io.ByteArrayInputStream) AtomicInteger(java.util.concurrent.atomic.AtomicInteger) CodecUtil(org.apache.lucene.codecs.CodecUtil) Directory(org.apache.lucene.store.Directory) Map(java.util.Map) IOContext(org.apache.lucene.store.IOContext) Path(java.nio.file.Path) Matchers.notNullValue(org.hamcrest.Matchers.notNullValue) BytesRef(org.apache.lucene.util.BytesRef) DirectoryReader(org.apache.lucene.index.DirectoryReader) UUIDs(org.elasticsearch.common.UUIDs) SegmentInfos(org.apache.lucene.index.SegmentInfos) FileNotFoundException(java.io.FileNotFoundException) MockAnalyzer(org.apache.lucene.analysis.MockAnalyzer) Engine(org.elasticsearch.index.engine.Engine) IndexWriter(org.apache.lucene.index.IndexWriter) List(java.util.List) SortedDocValuesField(org.apache.lucene.document.SortedDocValuesField) IndexMetaData(org.elasticsearch.cluster.metadata.IndexMetaData) Matchers.equalTo(org.hamcrest.Matchers.equalTo) Matchers.greaterThan(org.hamcrest.Matchers.greaterThan) Matchers.is(org.hamcrest.Matchers.is) IndexWriterConfig(org.apache.lucene.index.IndexWriterConfig) Matchers.endsWith(org.hamcrest.Matchers.endsWith) ByteArrayOutputStream(java.io.ByteArrayOutputStream) StringField(org.apache.lucene.document.StringField) TestUtil(org.apache.lucene.util.TestUtil) RAMDirectory(org.apache.lucene.store.RAMDirectory) HashMap(java.util.HashMap) Index(org.elasticsearch.index.Index) Lucene(org.elasticsearch.common.lucene.Lucene) OutputStreamStreamOutput(org.elasticsearch.common.io.stream.OutputStreamStreamOutput) ArrayList(java.util.ArrayList) BaseDirectoryWrapper(org.apache.lucene.store.BaseDirectoryWrapper) NoDeletionPolicy(org.apache.lucene.index.NoDeletionPolicy) TimeValue(org.elasticsearch.common.unit.TimeValue) IndexSettings(org.elasticsearch.index.IndexSettings) ESTestCase(org.elasticsearch.test.ESTestCase) VersionUtils.randomVersion(org.elasticsearch.test.VersionUtils.randomVersion) IndexOutput(org.apache.lucene.store.IndexOutput) Matchers.empty(org.hamcrest.Matchers.empty) IndexInput(org.apache.lucene.store.IndexInput) Iterator(java.util.Iterator) SnapshotDeletionPolicy(org.apache.lucene.index.SnapshotDeletionPolicy) IndexFileNames(org.apache.lucene.index.IndexFileNames) IOUtils(org.apache.lucene.util.IOUtils) Matchers(org.hamcrest.Matchers) IOException(java.io.IOException) IndexFormatTooOldException(org.apache.lucene.index.IndexFormatTooOldException) ShardLock(org.elasticsearch.env.ShardLock) Version(org.apache.lucene.util.Version) ExceptionsHelper(org.elasticsearch.ExceptionsHelper) KeepOnlyLastCommitDeletionPolicy(org.apache.lucene.index.KeepOnlyLastCommitDeletionPolicy) Field(org.apache.lucene.document.Field) Translog(org.elasticsearch.index.translog.Translog) Collections.unmodifiableMap(java.util.Collections.unmodifiableMap) TextField(org.apache.lucene.document.TextField) DummyShardLock(org.elasticsearch.test.DummyShardLock) TransportNodesListShardStoreMetaData(org.elasticsearch.indices.store.TransportNodesListShardStoreMetaData) IndexWriter(org.apache.lucene.index.IndexWriter) StringField(org.apache.lucene.document.StringField) BaseDirectoryWrapper(org.apache.lucene.store.BaseDirectoryWrapper) CorruptIndexException(org.apache.lucene.index.CorruptIndexException) DummyShardLock(org.elasticsearch.test.DummyShardLock) Document(org.apache.lucene.document.Document) IndexWriterConfig(org.apache.lucene.index.IndexWriterConfig)

Example 39 with Directory

use of org.apache.lucene.store.Directory in project elasticsearch by elastic.

the class IndexSearcherWrapperTests method testReaderCloseListenerIsCalled.

public void testReaderCloseListenerIsCalled() throws IOException {
    Directory dir = newDirectory();
    IndexWriterConfig iwc = newIndexWriterConfig();
    IndexWriter writer = new IndexWriter(dir, iwc);
    Document doc = new Document();
    doc.add(new StringField("id", "1", random().nextBoolean() ? Field.Store.YES : Field.Store.NO));
    doc.add(new TextField("field", "doc", random().nextBoolean() ? Field.Store.YES : Field.Store.NO));
    writer.addDocument(doc);
    DirectoryReader open = ElasticsearchDirectoryReader.wrap(DirectoryReader.open(writer), new ShardId("foo", "_na_", 1));
    IndexSearcher searcher = new IndexSearcher(open);
    assertEquals(1, searcher.search(new TermQuery(new Term("field", "doc")), 1).totalHits);
    final AtomicInteger closeCalls = new AtomicInteger(0);
    IndexSearcherWrapper wrapper = new IndexSearcherWrapper() {

        @Override
        public DirectoryReader wrap(DirectoryReader reader) throws IOException {
            return new FieldMaskingReader("field", reader, closeCalls);
        }

        @Override
        public IndexSearcher wrap(IndexSearcher searcher) throws EngineException {
            return searcher;
        }
    };
    final int sourceRefCount = open.getRefCount();
    final AtomicInteger count = new AtomicInteger();
    final AtomicInteger outerCount = new AtomicInteger();
    try (Engine.Searcher engineSearcher = new Engine.Searcher("foo", searcher)) {
        final Engine.Searcher wrap = wrapper.wrap(engineSearcher);
        assertEquals(1, wrap.reader().getRefCount());
        ElasticsearchDirectoryReader.addReaderCloseListener(wrap.getDirectoryReader(), reader -> {
            if (reader == open) {
                count.incrementAndGet();
            }
            outerCount.incrementAndGet();
        });
        assertEquals(0, wrap.searcher().search(new TermQuery(new Term("field", "doc")), 1).totalHits);
        wrap.close();
        assertFalse("wrapped reader is closed", wrap.reader().tryIncRef());
        assertEquals(sourceRefCount, open.getRefCount());
    }
    assertEquals(1, closeCalls.get());
    IOUtils.close(open, writer, dir);
    assertEquals(1, outerCount.get());
    assertEquals(1, count.get());
    assertEquals(0, open.getRefCount());
    assertEquals(1, closeCalls.get());
}
Also used : IndexSearcher(org.apache.lucene.search.IndexSearcher) TermQuery(org.apache.lucene.search.TermQuery) DirectoryReader(org.apache.lucene.index.DirectoryReader) ElasticsearchDirectoryReader(org.elasticsearch.common.lucene.index.ElasticsearchDirectoryReader) FilterDirectoryReader(org.apache.lucene.index.FilterDirectoryReader) IndexSearcher(org.apache.lucene.search.IndexSearcher) Term(org.apache.lucene.index.Term) Document(org.apache.lucene.document.Document) IndexWriter(org.apache.lucene.index.IndexWriter) AtomicInteger(java.util.concurrent.atomic.AtomicInteger) StringField(org.apache.lucene.document.StringField) TextField(org.apache.lucene.document.TextField) Engine(org.elasticsearch.index.engine.Engine) Directory(org.apache.lucene.store.Directory) IndexWriterConfig(org.apache.lucene.index.IndexWriterConfig)

Example 40 with Directory

use of org.apache.lucene.store.Directory in project elasticsearch by elastic.

the class IndexSearcherWrapperTests method testIsCacheable.

public void testIsCacheable() throws IOException {
    Directory dir = newDirectory();
    IndexWriterConfig iwc = newIndexWriterConfig();
    IndexWriter writer = new IndexWriter(dir, iwc);
    Document doc = new Document();
    doc.add(new StringField("id", "1", random().nextBoolean() ? Field.Store.YES : Field.Store.NO));
    doc.add(new TextField("field", "doc", random().nextBoolean() ? Field.Store.YES : Field.Store.NO));
    writer.addDocument(doc);
    DirectoryReader open = ElasticsearchDirectoryReader.wrap(DirectoryReader.open(writer), new ShardId("foo", "_na_", 1));
    IndexSearcher searcher = new IndexSearcher(open);
    assertEquals(1, searcher.search(new TermQuery(new Term("field", "doc")), 1).totalHits);
    searcher.setSimilarity(iwc.getSimilarity());
    final AtomicInteger closeCalls = new AtomicInteger(0);
    IndexSearcherWrapper wrapper = new IndexSearcherWrapper() {

        @Override
        public DirectoryReader wrap(DirectoryReader reader) throws IOException {
            return new FieldMaskingReader("field", reader, closeCalls);
        }

        @Override
        public IndexSearcher wrap(IndexSearcher searcher) throws EngineException {
            return searcher;
        }
    };
    final ConcurrentHashMap<Object, TopDocs> cache = new ConcurrentHashMap<>();
    try (Engine.Searcher engineSearcher = new Engine.Searcher("foo", searcher)) {
        try (Engine.Searcher wrap = wrapper.wrap(engineSearcher)) {
            ElasticsearchDirectoryReader.addReaderCloseListener(wrap.getDirectoryReader(), reader -> {
                cache.remove(reader.getCoreCacheKey());
            });
            TopDocs search = wrap.searcher().search(new TermQuery(new Term("field", "doc")), 1);
            cache.put(wrap.reader().getCoreCacheKey(), search);
        }
    }
    assertEquals(1, closeCalls.get());
    assertEquals(1, cache.size());
    IOUtils.close(open, writer, dir);
    assertEquals(0, cache.size());
    assertEquals(1, closeCalls.get());
}
Also used : IndexSearcher(org.apache.lucene.search.IndexSearcher) TermQuery(org.apache.lucene.search.TermQuery) DirectoryReader(org.apache.lucene.index.DirectoryReader) ElasticsearchDirectoryReader(org.elasticsearch.common.lucene.index.ElasticsearchDirectoryReader) FilterDirectoryReader(org.apache.lucene.index.FilterDirectoryReader) IndexSearcher(org.apache.lucene.search.IndexSearcher) Term(org.apache.lucene.index.Term) Document(org.apache.lucene.document.Document) TopDocs(org.apache.lucene.search.TopDocs) IndexWriter(org.apache.lucene.index.IndexWriter) AtomicInteger(java.util.concurrent.atomic.AtomicInteger) StringField(org.apache.lucene.document.StringField) TextField(org.apache.lucene.document.TextField) ConcurrentHashMap(java.util.concurrent.ConcurrentHashMap) Engine(org.elasticsearch.index.engine.Engine) Directory(org.apache.lucene.store.Directory) IndexWriterConfig(org.apache.lucene.index.IndexWriterConfig)

Aggregations

Directory (org.apache.lucene.store.Directory)2194 Document (org.apache.lucene.document.Document)1374 MockAnalyzer (org.apache.lucene.analysis.MockAnalyzer)816 RandomIndexWriter (org.apache.lucene.index.RandomIndexWriter)669 IndexReader (org.apache.lucene.index.IndexReader)590 IndexSearcher (org.apache.lucene.search.IndexSearcher)383 BytesRef (org.apache.lucene.util.BytesRef)376 RAMDirectory (org.apache.lucene.store.RAMDirectory)360 Term (org.apache.lucene.index.Term)325 StringField (org.apache.lucene.document.StringField)313 IndexWriter (org.apache.lucene.index.IndexWriter)312 IndexWriterConfig (org.apache.lucene.index.IndexWriterConfig)274 TextField (org.apache.lucene.document.TextField)259 Field (org.apache.lucene.document.Field)257 NumericDocValuesField (org.apache.lucene.document.NumericDocValuesField)257 Test (org.junit.Test)237 FSDirectory (org.apache.lucene.store.FSDirectory)221 Analyzer (org.apache.lucene.analysis.Analyzer)193 DirectoryReader (org.apache.lucene.index.DirectoryReader)193 TopDocs (org.apache.lucene.search.TopDocs)174