Please refer to JDK-8250968<https://bugs.openjdk.org/browse/JDK-8250968> which why the field in question was added. If you are interested in additional support, you can open an enhancement request for consideration.
Best Lance On Dec 12, 2023, at 4:15 PM, Alan Snyder <javali...@cbfiddle.com<mailto:javali...@cbfiddle.com>> wrote: On Dec 12, 2023, at 12:57 PM, Eirik Bjørsnøs <eir...@gmail.com<mailto:eir...@gmail.com>> wrote: Alan, ZipOutputStream is a relatively high-level API for producing ZIP files. There are many shapes and corner cases allowed by APPNOTE.TXT which ZipOutputStream has no means to produce. So while exposing the "external file attributes" field (which was incorrectly named "extraAttributes" in ZipEntry) is perhaps not harmful in itself, the question is: Where do we stop? Right now, I’m more interested in reading zip files, but to answer your question: we can stop at any point. It's worth noting that the ZipFileSystem API has the "enablePosixFileAttributes" env option which lets you set POSIX permissions on ZIP file entries using Files.setPosixFilePermissions() I’m more interested in the symlink bit, and do not currently have any need to use ZipFileSystem. I’m wondering, though, why it is OK to access information via ZipFileSystem and not directly via ZipEntry? The following OpenJDK test can serve as an example: https://github.com/openjdk/jdk/blob/master/test/jdk/sun/security/tools/jarsigner/PosixPermissionsTest.java#L151 Cheers, Eirik, On Tue, Dec 12, 2023 at 9:17 PM Alan Snyder <javali...@cbfiddle.com<mailto:javali...@cbfiddle.com>> wrote: ZipEntry is a public class and I am aware that it is used outside the JDK. Presumably that is not a problem. I’m wondering why the class stores the external file attributes field but does not provide public accessors for it. I would find it useful to have access to this field. I would rather not have to switch to using a third party library just for this reason. I understand that the interpretation of the external file attributes is not standardized, but the field itself is defined in the standard cited by the package info page, so I don’t see the harm in making it accessible without interpretation. Alan [cid:E1C4E2F0-ECD0-4C9D-ADB4-B16CA7BCB7FC@home] Lance Andersen | Principal Member of Technical Staff | +1.781.442.2037 Oracle Java Engineering 1 Network Drive Burlington, MA 01803 lance.ander...@oracle.com<mailto:lance.ander...@oracle.com>
