feat: GZIPInputStream instead of GzipCompressorInputStream to read tgz

InseeFr · Dec 5, 2024 · 0b2e177 · 0b2e177
1 parent 11cdf55
commit 0b2e177
Show file tree

Hide file tree

Showing 2 changed files with 59 additions and 38 deletions.
diff --git a/...src/main/java/fr/insee/arc/core/service/p2chargement/archiveloader/TarGzDecompressor.java b/...src/main/java/fr/insee/arc/core/service/p2chargement/archiveloader/TarGzDecompressor.java
@@ -6,6 +6,7 @@
 import java.io.FileInputStream;
 import java.io.FileOutputStream;
 import java.io.IOException;
+import java.util.zip.GZIPInputStream;
 import java.util.zip.GZIPOutputStream;
 
 import org.apache.commons.compress.archivers.tar.TarArchiveEntry;
@@ -32,51 +33,70 @@ public class TarGzDecompressor implements IArchiveExtractor {
 	@Override
 	public void extract(File archiveFile) throws IOException {
 		StaticLoggerDispatcher.info(LOGGER, "decompress()" + archiveFile.getName());
-		File dir = new File(archiveFile + ".dir");
+
+		File archiveFileExtractTargetDirectory = new File(archiveFile + ".dir");
 
-		try (GzipCompressorInputStream gzipIn = new GzipCompressorInputStream(
-				new BufferedInputStream(new FileInputStream(archiveFile), CompressedUtils.READ_BUFFER_SIZE));
-				TarArchiveInputStream tarIn = new TarArchiveInputStream(gzipIn)) {
-			TarArchiveEntry entry;
+		try (FileInputStream fis = new FileInputStream(archiveFile)) {
+			try (BufferedInputStream bis = new BufferedInputStream(fis, CompressedUtils.READ_BUFFER_SIZE)) {
+				try (GZIPInputStream gzipIn = new GZIPInputStream(bis, CompressedUtils.READ_BUFFER_SIZE)) {
+					try (TarArchiveInputStream tarIn = new TarArchiveInputStream(gzipIn, CompressedUtils.READ_BUFFER_SIZE)) {
+						extract(tarIn, archiveFileExtractTargetDirectory);
+					}
+				}
+			}
+		}
+		StaticLoggerDispatcher.info(LOGGER, "Untar completed successfully!");
+	}
+
+
+	/**
+	 * extract file from the TarArchiveInputStream
+	 * @param tarIn
+	 * @param archiveFileExtractTargetDirectory
+	 * @throws IOException
+	 */
+	private void extract(TarArchiveInputStream tarIn, File archiveFileExtractTargetDirectory) throws IOException
+	{
+		TarArchiveEntry entry;
 
-			while ((entry = (TarArchiveEntry) tarIn.getNextEntry()) != null) {
-				/** If the entry is a directory, create the directory. **/
+		while ((entry = (TarArchiveEntry) tarIn.getNextEntry()) != null) {
+			/** If the entry is a directory, create the directory. **/
 
-				// directories if not empty are automatically read in tar entries list
-				if (!entry.isDirectory()) {
-					int count;
-					byte[] data = new byte[32738];
+			// directories if not empty are automatically read in tar entries list
+			if (!entry.isDirectory()) {
+				int count;
+				byte[] data = new byte[CompressedUtils.WRITE_BUFFER_SIZE];
 
-					// temporary name for the file being uncompress
-					try (FileOutputStream fos = new FileOutputStream(dir.getAbsolutePath() + File.separator
-							+ ManipString.redoEntryName(entry.getName()) + ".tmp", false);
-							BufferedOutputStream dest = new BufferedOutputStream(fos, 32738);
-							GZIPOutputStream zdest = new GZIPOutputStream(dest);) {
-						while ((count = tarIn.read(data, 0, 32738)) != -1) {
-							zdest.write(data, 0, count);
+				// temporary name for the file being uncompress
+				try (FileOutputStream fos = new FileOutputStream(archiveFileExtractTargetDirectory.getAbsolutePath() + File.separator
+						+ ManipString.redoEntryName(entry.getName()) + ".tmp", false))
+				{
+					try(BufferedOutputStream dest = new BufferedOutputStream(fos, CompressedUtils.WRITE_BUFFER_SIZE))
+					{
+						try(GZIPOutputStream zdest = new GZIPOutputStream(dest, CompressedUtils.WRITE_BUFFER_SIZE))
+						{
+							while ((count = tarIn.read(data, 0, CompressedUtils.READ_BUFFER_SIZE)) != -1) {
+								zdest.write(data, 0, count);
+							}								
 						}
 					}
+				}
 
-					// rename the file when over makes it thread safe and available for other
-					// threads waiting
-					try {
-						FileUtilsArc.renameTo(
-								new File(dir.getAbsolutePath() + File.separator + ManipString.redoEntryName(entry.getName())
-										+ ".tmp"),
-								new File(
-										dir.getAbsolutePath() + File.separator + ManipString.redoEntryName(entry.getName()))
-						//
-						);
-					} catch (ArcException e) {
-						throw new IOException(e);
-					}
+				// rename the file when over makes it thread safe and available for other
+				// threads waiting
+				try {
+					FileUtilsArc.renameTo(
+							new File(archiveFileExtractTargetDirectory.getAbsolutePath() + File.separator + ManipString.redoEntryName(entry.getName())
+									+ ".tmp"),
+							new File(
+									archiveFileExtractTargetDirectory.getAbsolutePath() + File.separator + ManipString.redoEntryName(entry.getName()))
+					//
+					);
+				} catch (ArcException e) {
+					throw new IOException(e);
 				}
 			}
-
 		}
-
-		StaticLoggerDispatcher.info(LOGGER, "Untar completed successfully!");
-
 	}
-
+	
 }
diff --git a/arc-utils/src/main/java/fr/insee/arc/utils/files/CompressedUtils.java b/arc-utils/src/main/java/fr/insee/arc/utils/files/CompressedUtils.java
@@ -31,8 +31,9 @@ public class CompressedUtils {
 
 	private static final Logger LOGGER = LogManager.getLogger(CompressedUtils.class);
 
-	public static final int READ_BUFFER_SIZE = 131072;
-
+	public static final int READ_BUFFER_SIZE = 81920;
+	public static final int WRITE_BUFFER_SIZE = 81920;
+
 	private CompressedUtils() {
 		throw new IllegalStateException("Utility class");
 	}