问题
http://www.7-zip.org/sdk.html This site provide a LZMA SDK for compress/decompress files, I would like to give it a shot but I am lost.
Anyone got experience on this? Or a tutorial? Thanks.
回答1:
Short answer: don't
The 7zip sdk is old and unmaintained and it's just a JNI wrapper around the C++ library. A pure Java implementation on a modern JVM (1.7+) is as fast as a C++ one and has less dependecies and portability issues.
Have a look at http://tukaani.org/xz/java.html
XZ is a file format based on LZMA2 (an improved version of LZMA)
The guys that invented the XZ format build a pure java implementation of the XZ archive compression / extraction algorithms
The XZ file format is designed to store 1 file only. Thus you need to zip/tar the source folder(s) into a single uncompressed file first.
Using the java library is as easy as this:
FileInputStream inFile = new FileInputStream("src.tar");
FileOutputStream outfile = new FileOutputStream("src.tar.xz");
LZMA2Options options = new LZMA2Options();
options.setPreset(7); // play with this number: 6 is default but 7 works better for mid sized archives ( > 8mb)
XZOutputStream out = new XZOutputStream(outfile, options);
byte[] buf = new byte[8192];
int size;
while ((size = inFile.read(buf)) != -1)
out.write(buf, 0, size);
out.finish();
回答2:
Check out the LzmaAlone.java and LzmaBench.java files in the Java/SevenZip folder of the zip file from that link you posted.
回答3:
Use J7Zip. Its a java port of the LZMA SDK. You find it here:
http://sourceforge.net/projects/p7zip/files/J7Zip/
alternative
Use the lzmajio.jar with LzmaInputStream and LzmaOutputStream classes
you find it on github:
http://github.com/league/lzmajio/downloads
回答4:
You can use this library instead. It is "outdated" but still works fine.
Maven dependency
<dependency>
<groupId>com.github.jponge</groupId>
<artifactId>lzma-java</artifactId>
<version>1.2</version>
</dependency>
Utility class
import lzma.sdk.lzma.Decoder;
import lzma.streams.LzmaInputStream;
import lzma.streams.LzmaOutputStream;
import org.apache.commons.compress.utils.IOUtils;
import java.io.*;
import java.nio.file.Path;
public class LzmaCompressor
{
private Path rawFilePath;
private Path compressedFilePath;
public LzmaCompressor(Path rawFilePath, Path compressedFilePath)
{
this.rawFilePath = rawFilePath;
this.compressedFilePath = compressedFilePath;
}
public void compress() throws IOException
{
try (LzmaOutputStream outputStream = new LzmaOutputStream.Builder(
new BufferedOutputStream(new FileOutputStream(compressedFilePath.toFile())))
.useMaximalDictionarySize()
.useMaximalFastBytes()
.build();
InputStream inputStream = new BufferedInputStream(new FileInputStream(rawFilePath.toFile())))
{
IOUtils.copy(inputStream, outputStream);
}
}
public void decompress() throws IOException
{
try (LzmaInputStream inputStream = new LzmaInputStream(
new BufferedInputStream(new FileInputStream(compressedFilePath.toFile())),
new Decoder());
OutputStream outputStream = new BufferedOutputStream(
new FileOutputStream(rawFilePath.toFile())))
{
IOUtils.copy(inputStream, outputStream);
}
}
}
Firstly you have to create a file with content to start compressing. You can use this website to generate random text.
Example compression and decompression
Path rawFile = Paths.get("raw.txt");
Path compressedFile = Paths.get("compressed.lzma");
LzmaCompressor lzmaCompressor = new LzmaCompressor(rawFile, compressedFile);
lzmaCompressor.compress();
lzmaCompressor.decompress();
回答5:
Here are tested examples of using XZ Utils pure java library to pack and unpack XZ archives with LZMA2 compression algorithm with great ratio.
import org.tukaani.xz.*;
// CompressXz
public static void main(String[] args) throws Exception {
String from = args[0];
String to = args[1];
try (FileOutputStream fileStream = new FileOutputStream(to);
XZOutputStream xzStream = new XZOutputStream(
fileStream, new LZMA2Options(LZMA2Options.PRESET_MAX), BasicArrayCache.getInstance())) {
Files.copy(Paths.get(from), xzStream);
}
}
// DecompressXz
public static void main(String[] args) throws Exception {
String from = args[0];
String to = args[1];
try (FileInputStream fileStream = new FileInputStream(from);
XZInputStream xzStream = new XZInputStream(fileStream, BasicArrayCache.getInstance())) {
Files.copy(xzStream, Paths.get(to), StandardCopyOption.REPLACE_EXISTING);
}
}
// DecompressXzSeekable (partial)
public static void main(String[] args) throws Exception {
String from = args[0];
String to = args[1];
int offset = Integer.parseInt(args[2]);
int size = Integer.parseInt(args[3]);
try (SeekableInputStream fileStream = new SeekableFileInputStream(from);
SeekableXZInputStream xzStream = new SeekableXZInputStream(fileStream, BasicArrayCache.getInstance())) {
xzStream.seek(offset);
byte[] buf = new byte[size];
if (size != xzStream.read(buf)) {
xzStream.available(); // let it throw the last exception, if any
throw new IOException("Truncated stream");
}
Files.write(Paths.get(to), buf);
}
}
回答6:
https://mvnrepository.com/artifact/org.tukaani/xz/1.8
Kotlin code for Android:
fun initDatabase() {
var gisFile = this.getDatabasePath("china_gis.db");
if (!gisFile.exists()) {
if(!gisFile.parentFile.exists()) gisFile.parentFile.mkdirs();
var inStream = assets.open("china_gis_no_poly.db.xz")
inStream.use { input ->
val buf = ByteArray(1024)
XZInputStream(input).use { input ->
FileOutputStream(gisFile,true).use { output ->
var size: Int
while (true) {
size = input.read(buf);
if (size != -1) {
output.write(buf, 0, size)
} else {
break;
}
}
output.flush()
}
}
}
}
}
Java code:
byte[] buf = new byte[8192];
String name = "C:\\Users\\temp22\\Downloads\\2017-007-13\\china_gis_no_poly.db.xz";
try {
InputStream input = new FileInputStream(name);
FileOutputStream output= new FileOutputStream(name+".db");
try {
// Since XZInputStream does some buffering internally
// anyway, BufferedInputStream doesn't seem to be
// needed here to improve performance.
// in = new BufferedInputStream(in);
input = new XZInputStream(input);
int size;
while ((size = input.read(buf)) != -1)
output.write(buf, 0, size);
output.flush();
} finally {
// Close FileInputStream (directly or indirectly
// via XZInputStream, it doesn't matter).
input.close();
output.close();
}
} catch (FileNotFoundException e) {
System.err.println("XZDecDemo: Cannot open " + name + ": "
+ e.getMessage());
System.exit(1);
} catch (EOFException e) {
System.err.println("XZDecDemo: Unexpected end of input on "
+ name);
System.exit(1);
} catch (IOException e) {
System.err.println("XZDecDemo: Error decompressing from "
+ name + ": " + e.getMessage());
System.exit(1);
}
来源：https://stackoverflow.com/questions/5481487/how-to-use-lzma-sdk-to-compress-decompress-in-java