commons-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Dmitri Plotnikov <>
Subject Re: [clazz] draft reflect implementation
Date Sun, 10 Nov 2002 16:38:12 GMT

----- Original Message -----
From: "Victor Volle" <>
To: "Jakarta Commons Developers List" <>
Sent: Sunday, November 10, 2002 7:46 AM
Subject: Re: [clazz] draft reflect implementation

> Dmitri
> Dmitri wrote:
> > Victor Volle wrote:
> > > PS: I am hesitating to offer my help, yet (besides criticism :-))
> > > because I am still not sure that I can use it for my own project
> > > and I would like to wait for the "dust to settle".
> > Critisizm is in fact extremely valuable help.  What I want to know
> > is what you think needs to be done in clazz for it to be useful for your
> > project.
> What I am working at is a simple but generic code generator.
> It creates a model from different sources (reflection, XML, XMI,
> direct integration with IDEs and so on).
Sounds great.

> So I could use clazz in to ways:
>    1) Simply use its advanced reflection features to load
>        my own meta model AND use it when loading from
>        XML (where I currently use my own hack of BeanUtils code)
Makes sense.

>    2) Additionally reuse the clazz meta model but then I would
>        have to add a whole bunch of things: Packages as model
>        elements (because they meight have Attributes),
>        InnerClasses, CompilationUnits (for Import handling
>        and type resolving (currently integrated in my MetaClass)),
>        MetaTypes (again for (lazy) type resolving).
In my opinion this is too Java-specific for Clazz.  I see Clazz as a
customizable version of Introspector, more than a precise model of Java
classes.  I have worked with a couple of such precise models: Borland's Jot,
Eclipse Jdt.  Those models are very powerful, but also very specific.  I am
not convinced that [clazz] is good for that purpose.

> I am quite sure that I will use clazz for the first purpose,
> because I am convinced that clazz is on a good way.
> But since the focus of clazz is not (yet) the meta model
> but the advanced reflection, I am not sure that the different
> use cases could or should be aligned.
[clazz] will soon have the "meta model" features, but whether to make it as
detailed as needed for your purposes is open for discussion.

> Victor
- Dmitri

To unsubscribe, e-mail:   <>
For additional commands, e-mail: <>

View raw message