Please review, but don't commit (!) CXF-984 (aegis control of namespace prefixes)

[Benson Margulies <bi...@basistech.com>]

CXF-984 provides a new feature: allow services to control the use of
namespace prefixes.

 

Now, in theory, no one should want to do this. However, there are two
reasons why this has come up.

 

1)     wsdl readability.

2)     Primitive clients. If you are concocting a client in, say,
javascript or perl or someplace with a non-namespace-aware XML parser,
knowing what's going to show up in the prefixes can save a lot of
effort.

 

The proposed change here is rather lightweight. I claim that the
AegisDatabinding is the right place for this, since the prefixes
overarch the entire service. Pulling them out of .aegis.xml files can't
be made coherent.

 

The patch I posted needs more unit test, but I wanted to stop and see if
the rest of the group found this entire idea palatable before I pushed
it any further.

 

--benson

 

RE: Please review, but don't commit (!) CXF-984 (aegis control of namespace prefixes)

[Benson Margulies <bi...@basistech.com>]

Dan,

I just posted the larger patch. The first version neglected to include
the xsd:schema items produced by ReflectionServiceFactory for the
services. This one covers those.

I'd like to clean up the dueling 'AbstractDataBinding' issue, but that's
painful as a non-committer, I suspect. 

I'd also like to add more to the unit test. It would be fine with me if
you'd commit this much. I'll tee up more unit test later, and perhaps I
could refactor the AbstractDataBinding classes if I am transfigured
sometime soon.

--benson

RE: Please review, but don't commit (!) CXF-984 (aegis control of namespace prefixes)

[Benson Margulies <bi...@basistech.com>]

It turns out to be incomplete! I'll post a bigger one.

> -----Original Message-----
> From: Dan Diephouse [mailto:dan.diephouse@mulesource.com]
> Sent: Monday, September 10, 2007 8:59 AM
> To: cxf-dev@incubator.apache.org
> Subject: Re: Please review, but don't commit (!) CXF-984 (aegis
control of
> namespace prefixes)
> 
> Well actually your patch was much shorter than I had assumed because
it
> only changes the namespaces of the wsdl. So I took a look at it and it
> looks fine! I will see if I commit this later today.
> 
> - Dan
> 
> 
> Dan Diephouse wrote:
> > Hi Benson,
> >
> > I've got this on my agenda to look at. I may/may not get an email
out
> > today to you. But I'm pretty sure I can get you feedback by
tomorrow.
> > Thanks,
> > - Dan
> >
> > Benson Margulies wrote:
> >> CXF-984 provides a new feature: allow services to control the use
of
> >> namespace prefixes.
> >>
> >>
> >>
> >> Now, in theory, no one should want to do this. However, there are
two
> >> reasons why this has come up.
> >>
> >>
> >>
> >> 1)     wsdl readability.
> >>
> >> 2)     Primitive clients. If you are concocting a client in, say,
> >> javascript or perl or someplace with a non-namespace-aware XML
parser,
> >> knowing what's going to show up in the prefixes can save a lot of
> >> effort.
> >>
> >>
> >>
> >> The proposed change here is rather lightweight. I claim that the
> >> AegisDatabinding is the right place for this, since the prefixes
> >> overarch the entire service. Pulling them out of .aegis.xml files
can't
> >> be made coherent.
> >>
> >>
> >>
> >> The patch I posted needs more unit test, but I wanted to stop and
see
> if
> >> the rest of the group found this entire idea palatable before I
pushed
> >> it any further.
> >>
> >>
> >>
> >> --benson
> >>
> >>
> >>
> >>
> >>
> >
> >
> 
> 
> --
> Dan Diephouse
> MuleSource
> http://mulesource.com | http://netzooid.com/blog

Re: Please review, but don't commit (!) CXF-984 (aegis control of namespace prefixes)

[Dan Diephouse <da...@mulesource.com>]

Well actually your patch was much shorter than I had assumed because it 
only changes the namespaces of the wsdl. So I took a look at it and it 
looks fine! I will see if I commit this later today.

- Dan


Dan Diephouse wrote:
> Hi Benson,
>
> I've got this on my agenda to look at. I may/may not get an email out 
> today to you. But I'm pretty sure I can get you feedback by tomorrow. 
> Thanks,
> - Dan
>
> Benson Margulies wrote:
>> CXF-984 provides a new feature: allow services to control the use of
>> namespace prefixes.
>>
>>  
>>
>> Now, in theory, no one should want to do this. However, there are two
>> reasons why this has come up.
>>
>>  
>>
>> 1)     wsdl readability.
>>
>> 2)     Primitive clients. If you are concocting a client in, say,
>> javascript or perl or someplace with a non-namespace-aware XML parser,
>> knowing what's going to show up in the prefixes can save a lot of
>> effort.
>>
>>  
>>
>> The proposed change here is rather lightweight. I claim that the
>> AegisDatabinding is the right place for this, since the prefixes
>> overarch the entire service. Pulling them out of .aegis.xml files can't
>> be made coherent.
>>
>>  
>>
>> The patch I posted needs more unit test, but I wanted to stop and see if
>> the rest of the group found this entire idea palatable before I pushed
>> it any further.
>>
>>  
>>
>> --benson
>>
>>  
>>
>>
>>   
>
>


-- 
Dan Diephouse
MuleSource
http://mulesource.com | http://netzooid.com/blog

Re: Please review, but don't commit (!) CXF-984 (aegis control of namespace prefixes)

[Dan Diephouse <da...@mulesource.com>]

Hi Benson,

I've got this on my agenda to look at. I may/may not get an email out 
today to you. But I'm pretty sure I can get you feedback by tomorrow. 
Thanks,
- Dan

Benson Margulies wrote:
> CXF-984 provides a new feature: allow services to control the use of
> namespace prefixes.
>
>  
>
> Now, in theory, no one should want to do this. However, there are two
> reasons why this has come up.
>
>  
>
> 1)     wsdl readability.
>
> 2)     Primitive clients. If you are concocting a client in, say,
> javascript or perl or someplace with a non-namespace-aware XML parser,
> knowing what's going to show up in the prefixes can save a lot of
> effort.
>
>  
>
> The proposed change here is rather lightweight. I claim that the
> AegisDatabinding is the right place for this, since the prefixes
> overarch the entire service. Pulling them out of .aegis.xml files can't
> be made coherent.
>
>  
>
> The patch I posted needs more unit test, but I wanted to stop and see if
> the rest of the group found this entire idea palatable before I pushed
> it any further.
>
>  
>
> --benson
>
>  
>
>
>   


-- 
Dan Diephouse
MuleSource
http://mulesource.com | http://netzooid.com/blog