On Sat, 17 Sep 2022 08:05:37 GMT, Alan Bateman <al...@openjdk.org> wrote:
> It's a bit better but I think we can make it clearer and also link the JAR
> Manifest section of the JAR file spec. Can you try this:
>
> ```
> * <p> The {@link #getManifest() getManifest} method is used to read the
> * <a href="{@docRoot}/../specs/jar/jar.html#jar-manifest">JAR Manifest</a>
> * from the entry {@code META-INF/MANIFEST.MF} when it is the first entry
> * in the stream (or the second entry in the case that the fist entry is
> * {@code META-INF/} and the second entry {@code META-INF/MANIFEST.MF}).
> *
> * <p> The {@link #getNextJarEntry()} and {@link #getNextEntry()} methods are
> * used to read JAR file entries from the stream. These methods skip over the
> * manifest ({@code META-INF/MANIFEST.MF}) when it is at the beginning of the
> * stream. In other words, these methods do not return an entry for the
> manifest
> * when the manifest is the first entry in the stream. If the first entry is
> * {@code META-INF/} and the second entry is the manifest then both are
> skipped
> * over by these methods. Whether these methods skip over the manifest when it
> * appears later in the stream is not specified.
> ```
>
Revised using the above
> I think we also have to update getManifest method to align with the above as
> doesn't say anything about the manifest needing to be at the beginning of the
> stream.
Ok, I added some verbiage similar to what I had originally before we decided
to update the class description. Please let me know if this is what you had in
mind
Thank you again for your feedback
-------------
PR: https://git.openjdk.org/jdk/pull/10045
