Re: XMLEncoder / XMLDecoder

From: Andrus Adamchik (andru..bjectstyle.org)
Date: Mon Jul 05 2010 - 07:46:26 UTC

  • Next message: Juan José Gil: "Mapping views with modeller"

    In the old implementation I am pretty sure there's no notion of object
    states stored in XML.

    In the one I haven't got there just yet, concentrating on serializing
    committed objects.

    Andrus

    On Jul 5, 2010, at 10:39 AM, Gary Jarrel wrote:

    > On Mon, Jul 5, 2010 at 5:04 PM, Andrus Adamchik <andru..bjectstyle.org
    > > wrote:
    >
    >> On a side note, I am not a big fan of our XML coding
    >> implementation. It is
    >> DOM based and is not very scaleable. I started writing an XStream-
    >> based
    >> replacement in sandbox:
    >>
    >> *
    >> http://svn.apache.org/viewvc/cayenne/sandbox/cayenne-serialization/?pathrev=814869
    >> * http://markmail.org/message/ljngasncx6jur6ud
    >>
    >> But then my own itch to solve it disappeared, so that was put on
    >> hold. IIRC
    >> it is usable for most object subgraphs, but not optimized for all
    >> scenarios.
    >>
    >
    > I've been looking at using XSteam as well, seems interesting, however
    > I have a need to serialize only a small number of objects so perhaps
    > the current implementation would do the trick as well. XStream seems
    > quite easy to use, may give it a god just for the fun of it.
    >
    > If I have decoded an object from XML is there an automated way to tell
    > that the decoded object has a representing record in the database.
    > Just having a look all the objects after being decoded have NEW as a
    > Persistence State. I tried to also serialize the primary key during
    > the serialization process, but when decoding it, it still has
    > persistence state new.
    >
    > Thank you
    >
    > Gary
    >



    This archive was generated by hypermail 2.0.0 : Mon Jul 05 2010 - 07:48:01 UTC