io.camunda.zeebe.snapshots.impl.FileBasedSnapshotChunkReader Maven / Gradle / Ivy
The newest version!
/*
* Copyright Camunda Services GmbH and/or licensed to Camunda Services GmbH under
* one or more contributor license agreements. See the NOTICE file distributed
* with this work for additional information regarding copyright ownership.
* Licensed under the Camunda License 1.0. You may not use this file
* except in compliance with the Camunda License 1.0.
*/
package io.camunda.zeebe.snapshots.impl;
import io.camunda.zeebe.snapshots.SnapshotChunk;
import io.camunda.zeebe.snapshots.SnapshotChunkReader;
import java.io.IOException;
import java.io.RandomAccessFile;
import java.io.UncheckedIOException;
import java.nio.ByteBuffer;
import java.nio.file.Files;
import java.nio.file.Path;
import java.util.NavigableSet;
import java.util.TreeSet;
/**
* Implements a chunk reader where each chunk is a single file in a root directory. Chunks are then
* ordered lexicographically, and the files are assumed to be immutable, i.e. no more are added to
* the directory once this is created.
*/
public final class FileBasedSnapshotChunkReader implements SnapshotChunkReader {
private final Path directory;
private final NavigableSet chunks;
private long offset;
private NavigableSet chunksView;
private final int totalCount;
private final String snapshotID;
private long maximumChunkSize;
public FileBasedSnapshotChunkReader(final Path directory) throws IOException {
this(directory, Long.MAX_VALUE);
}
FileBasedSnapshotChunkReader(final Path directory, final long maximumChunkSize)
throws IOException {
this.directory = directory;
chunks = collectChunks(directory);
totalCount = chunks.size();
chunksView = new TreeSet<>(chunks);
snapshotID = directory.getFileName().toString();
this.maximumChunkSize = maximumChunkSize;
}
private NavigableSet collectChunks(final Path directory) throws IOException {
final var set = new TreeSet<>(CharSequence::compare);
try (final var stream = Files.list(directory).sorted()) {
stream.map(directory::relativize).map(Path::toString).forEach(set::add);
}
return set;
}
@Override
public void reset() {
chunksView = new TreeSet<>(chunks);
}
@Override
public void seek(final ByteBuffer id) {
if (id == null) {
return;
}
final var chunkId = new SnapshotChunkId(id);
offset = chunkId.offset();
chunksView = new TreeSet<>(chunks.tailSet(chunkId.fileName(), true));
}
@Override
public ByteBuffer nextId() {
if (chunksView.isEmpty()) {
return null;
}
return new SnapshotChunkId(chunksView.first().toString(), offset).id();
}
@Override
public void setMaximumChunkSize(final int maximumChunkSize) {
this.maximumChunkSize = maximumChunkSize;
}
@Override
public void close() {
chunks.clear();
chunksView.clear();
}
@Override
public boolean hasNext() {
return !chunksView.isEmpty();
}
@Override
public SnapshotChunk next() {
final var fileName = chunksView.first().toString();
final var filePath = directory.resolve(fileName).toString();
try (final var file = new RandomAccessFile(filePath, "r")) {
final var fileLength = file.length();
final var bytesToRead = Math.min(maximumChunkSize, fileLength - offset);
final byte[] buffer = new byte[(int) bytesToRead];
file.seek(offset);
file.readFully(buffer);
final var fileBlockPosition = offset;
offset += bytesToRead;
if (offset == fileLength) {
offset = 0;
chunksView.pollFirst();
}
return SnapshotChunkUtil.createSnapshotChunkFromFileChunk(
snapshotID, totalCount, fileName, buffer, fileBlockPosition, fileLength);
} catch (final IOException e) {
throw new UncheckedIOException(e);
}
}
}