I have noticed the following behavior that seems strange to me.
I am using the assembly plugin to create a tar.gz archive. I have
several filesets that I use, and they all specify file and directory
permissions. Some filesets create a single target directory from a
single source directory and move in all the files in that directory.
There is no directory recursion involved and it works fine, using
<include>**</include>. The archive contains an entry for that
directory, with the proper permissions.
One directory has a subdirectory and I also want to make a fileset out
of that. In that directory I was using <include>**/*</include>. All
the files that needed to be included were included with the right
permissions and there was an entry in the archive for the subdirectory
with the right permissions.
However, there was no entry for the top directory of this fileset and so
when the archive was extracted, the directory created had the default
rather than the desired permissions.
I fixed this (bandaided it) with
<include>**</include>
<include>**/*</include>
Should this be necessary? What am I missing? Why isn't a directory
created for the root of the fileset in the archive when only
<include>**/*</include> is specified?
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscr...@maven.apache.org
For additional commands, e-mail: users-h...@maven.apache.org
