Java 大文件压缩

原博客：https://blog.csdn.net/ahau10/article/details/52550430/

压缩效率

win10 环境。linux 会更快

代码

1. 引入 commons-compress 坐标

<dependency>
    <groupId>org.apache.commons</groupId>
    <artifactId>commons-compress</artifactId>
    <version>1.20</version>
</dependency>

2. 创建 ScatterSample.java

public class ScatterSample {
    private String rootPath;

    ParallelScatterZipCreator scatterZipCreator = new ParallelScatterZipCreator();

    /**
     * ParallelScatterZipCreator api says:
     * 注意这个类不保证写入到输出文件的顺序。需要保持特定顺序的（manifests，文件夹）必须使用这个类的客户类进行处理
     * 通常的做法是 在调用这个类的writeTo方法前把这些东西写入到ZipArchiveOutputStream
     */
    ScatterZipOutputStream dirs = ScatterZipOutputStream
            .fileBased(File.createTempFile("whatever-preffix", ".whatever"));

    public ScatterSample(String rootPath) throws IOException {
        this.rootPath = rootPath;
    }

    public void addEntry(final ZipArchiveEntry zipArchiveEntry, final InputStreamSupplier streamSupplier)
            throws IOException {
        if (zipArchiveEntry.isDirectory() && !zipArchiveEntry.isUnixSymlink()) {
            dirs.addArchiveEntry(ZipArchiveEntryRequest.createZipArchiveEntryRequest(zipArchiveEntry, streamSupplier));
        } else {
            scatterZipCreator.addArchiveEntry(zipArchiveEntry, streamSupplier);
        }
    }

    public void writeTo(final ZipArchiveOutputStream zipArchiveOutputStream)
            throws IOException, ExecutionException, InterruptedException {
        dirs.writeTo(zipArchiveOutputStream);
        dirs.close();
        scatterZipCreator.writeTo(zipArchiveOutputStream);
    }

    public String getRootPath() {
        return rootPath;
    }

    public void setRootPath(String rootPath) {
        this.rootPath = rootPath;
    }
}

3. 创建压缩工具类

public class ZipUtilParallelScatter {

    /**
     * 压缩
     *
     * @param source  将要被压缩的文件
     * @param zipFile 压缩后的文件
     * @return 压缩后的文件路径
     */
    public String doZipFile(File source, File zipFile) {
        try {
            File zipFileParent = new File(zipFile.getParent());
            // 若 压缩文件的路径 不存在，则创建出来
            if (!zipFileParent.exists()) {
                zipFileParent.mkdirs();
            }
            ScatterSample scatterSample = new ScatterSample(source.getAbsolutePath());
            compressCurrentDirectory(scatterSample, source);
            ZipArchiveOutputStream zipArchiveOutputStream = new ZipArchiveOutputStream(zipFile);
            scatterSample.writeTo(zipArchiveOutputStream);
            zipArchiveOutputStream.close();

            return zipFile.getAbsolutePath();
        } catch (IOException | InterruptedException | ExecutionException e) {
            throw new RuntimeException(e);
        }
    }

    private void compressCurrentDirectory(ScatterSample scatterSample, File source) throws IOException {
        if (source == null) {
            throw new IOException("源路径不能为空！");
        }
        String relativePath = "";
        if (source.isFile()) {
            relativePath = source.getName();
            addEntry(relativePath, source, scatterSample);
            return;
        }

        // 空文件夹
        if (Objects.requireNonNull(source.listFiles()).length == 0) {
            relativePath = source.getAbsolutePath().replace(scatterSample.getRootPath(), "");
            addEntry(relativePath + File.separator, source, scatterSample);
            return;
        }
        for (File f : Objects.requireNonNull(source.listFiles())) {
            if (f.isDirectory()) {
                compressCurrentDirectory(scatterSample, f);
            } else {
                relativePath = f.getParent().replace(scatterSample.getRootPath(), "");
                addEntry(relativePath + File.separator + f.getName(), f, scatterSample);
            }
        }
    }

    private void addEntry(String entryName, File currentFile, ScatterSample scatterSample) throws IOException {
        ZipArchiveEntry archiveEntry = new ZipArchiveEntry(entryName);
        archiveEntry.setMethod(ZipEntry.DEFLATED);
        final InputStreamSupplier supp = new CustomInputStreamSupplier(currentFile);
        scatterSample.addEntry(archiveEntry, supp);
    }

    class CustomInputStreamSupplier implements InputStreamSupplier {
        private File currentFile;

        public CustomInputStreamSupplier(File currentFile) {
            this.currentFile = currentFile;
        }

        @Override
        public InputStream get() {
            try {
                // InputStreamSupplier api says:
                // 返回值：输入流。永远不能为Null,但可以是一个空的流
                return currentFile.isDirectory() ? new NullInputStream(0) : new FileInputStream(currentFile);
            } catch (FileNotFoundException e) {
                e.printStackTrace();
            }
            return null;
        }
    }
}

4. 测试

public class ZipUtilsTest {

    private File source = new File("F:\\upload");
    String targetPath = "F:/upload_" + System.currentTimeMillis() + ".zip";
    private File target = new File(targetPath);

    @Test
    public void testZipUtilParallelScatter() {
        long start = System.currentTimeMillis();
        ZipUtilParallelScatter zipUtil = new ZipUtilParallelScatter();
        final String zipPath = zipUtil.doZipFile(source, target);
        System.out.println("耗时：" + (System.currentTimeMillis() - start) + "；压缩文件路径：" + zipPath);
    }
}

5. 扩展
   可以将文件分批，改成多线程并发 执行 压缩方法

参考：
Java 高效压缩zip
使用Java ParallelScatterZipCreator快速压缩文件夹(Fast zipping folder using java ParallelScatterZipCreator)