Path -> URI -> Path round trip fails for lazily created zip file systems

[Xueming Shen]

On 12/28/13 11:42 AM, Philippe Marschall wrote:
> On Fri, Dec 27, 2013 at 8:20 PM, Xueming Shen <xueming.shen at oracle.com> wrote:
>> Hi Philippe,
>>
>> I've file a new issue into the bug system.
>> https://bugs.openjdk.java.net/browse/JDK-8031077
>>
>> The "realPath" need to be check and set, if null (in case of "create new"),
>> before to be added
>> into the map "filesystems". The check is to obtain the real path to check
>> whether or not the
>> filesystem (may represented by different url) has been open already, throw
>> FSAEE if yet, as
>> the API spec requests.
> Cool thanks. So if the file doesn't yet exist should it be created
> there instead of the ZipFileSystem constructor?
The spec does not appears to care where the new file should be created. 
I would take this
as an implementation details.  Either way is fine, but inside 
ZipFileSystem's constructor
would be more "convenient", personally.

-Sherman
> Cheers
> Philippe