[jira] [Commented] (COMPRESS-254) ZipArchiveOutputStream try to switch to Zip64 when entries count over 64K entries

["Stefan Bodewig (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/COMPRESS-254?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13870485#comment-13870485 ] 

Stefan Bodewig commented on COMPRESS-254:
-----------------------------------------

When not using Zip64 you are limited to 65535 entries or you create invalid ZIP archives.  See section 4.3.16 in http://www.pkware.com/documents/casestudies/APPNOTE.TXT where there are only two bytes for "total number of entries in the central directory".  So when we introduced Zip64 support we plugged a hole that allowed Compress to create invalid archives.


> ZipArchiveOutputStream try to switch to Zip64 when entries count over 64K entries
> ---------------------------------------------------------------------------------
>
>                 Key: COMPRESS-254
>                 URL: https://issues.apache.org/jira/browse/COMPRESS-254
>             Project: Commons Compress
>          Issue Type: Bug
>          Components: Archivers
>    Affects Versions: 1.3, 1.4, 1.5, 1.6
>         Environment: Unzip version < 6.0 (without Zip64 support)
>            Reporter: Taras Ledkov
>
> Since version 1.3 we cannot create zip archive that contains over then 64K entries & without Zip64 extension.
> In case Zip64Mode.AsNeed is used target zip file cannot be open by unzip older then version 6.0.
> In case Zip64Mode.Never is used exception is thrown.
> Why so strong restriction as (ZipConstants,ZIP64_MAGIC_SHORT = 0xFFFF) was happened? 
> To reproduce you can use dummy test below & [old unzip binaries|ftp://ftp.info-zip.org/pub/infozip/unix/]: 
> {code:title=TestHuge}
> import org.apache.commons.compress.archivers.zip.ZipArchiveEntry;
> import org.apache.commons.compress.archivers.zip.ZipArchiveOutputStream;
> import java.io.File;
> import java.io.FileOutputStream;
> import java.io.IOException;
> import java.io.OutputStream;
> public class TestHuge {
>     public static void main(String[] args) throws IOException {
>         try(OutputStream os = new FileOutputStream(new File("test.zip"));
>             ZipArchiveOutputStream zos = new ZipArchiveOutputStream(os)) {
>             zos.setLevel(0);
>             for (int dirCount = 0; dirCount < 10; ++dirCount) {
>                 ZipArchiveEntry entryDir = new ZipArchiveEntry("dir" + dirCount + "/");
>                 zos.putArchiveEntry(entryDir);
>                 zos.closeArchiveEntry();
>                 for (int i = 0; i < 0x4000; ++i) {
>                     ZipArchiveEntry entryFile = new ZipArchiveEntry("dir" + dirCount + "/" + "file" + i);
>                     zos.putArchiveEntry(entryFile);
>                     zos.write(("" + dirCount + " " + i).getBytes());
>                     zos.closeArchiveEntry();
>                 }
>             }
>             zos.finish();
>         }
>     }
> }
> {code}



--
This message was sent by Atlassian JIRA
(v6.1.5#6160)