Cool, yeah I think the TransactionPlugin could have some interesting uses
for sure, I think it’s kind of under utilized.

Re the user data attributes I think personally think it could be general
enough. The whole user data idea about shuffling attributes along that you
don’t want to model explicitly I think is pretty useful. And extending that
out to exchange formats like GML and JSON I think is just a natural
extension of that. Interested to hear how other devs feel.

On Fri, Apr 8, 2016 at 9:33 AM Jim Hughes <jn...@ccri.com> wrote:

> Hi Justin,
>
> Thanks.  I forgot to mention that I had seen the TransactionPlugin
> extension point; that one is awesome and I see how to use it support my
> second suggestion.
>
> Would changes in the GML2/3ParsingUtils to inject attributes into a
> SimpleFeature's userData be sensible / more broadly useful?  I'm a bit shy
> about pushing hard for such a change since I am not sure about any other
> ramifications.
>
> Thanks,
>
> Jim
>
>
> On 04/08/2016 10:45 AM, Justin Deoliveira wrote:
>
> Hey Jim,
>
> The wfs native element might be the cleanest way to handle this. I’ve seen
> folks use it before for security and validation type stuff with some
> success. In GeoServer you can plug in WFS transactions callbacks pretty
> easily using the TransactionPlugin extension point. It has direct access
> to the transaction object which has the native content on it.
>
> If you do want to go with approach (1): during GML parsing I don’t believe
> there is any way to tag a GML attribute as something you want to store in
> user data. Folks might correct me on that one though. To support it I
> wonder if we could do something simple like a custom attribute on the gml
> attribute element. Something like:
>
>  <ns:someAttribute userData=“true”>foo</ns:someAttribute>
>
> One potential downside might be validation… however in GeoServer last I
> checked we don’t validate features on the way in… mostly because it’s a
> pain to have to specify all the schema references properly so a parser can
> actually validate. It may also be there in the gml schema there is some
> attribute we might be able to use /abuse for this purpose… I can’t think of
> one off hand though.
>
> As far as I know the feature parsing logic still lives in the
> GML3ParsingUtils and GML2ParsingUtils classes. To add some support for
> putting an attribute in user data rather than a normal attribute I think
> you would be looking at modifying this method.
>
> As you’ll notice this only applies to simple features. If working with
> complex features i am actually not sure where that logic lives anymore. One
> of the app-schema folks can point you at that if need be.
>
> Hope that helps!
>
> -Justin
>
>
> On Fri, Apr 8, 2016 at 8:09 AM Jim Hughes <jn...@ccri.com> wrote:
>
>> Hi all,
>>
>> Apologies for the cross-post, but my question hits a little of how
>> GeoServer handles WFS transactions and a little bit of how GeoTools may
>> handle WFS/GML parsing.
>>
>> My high-level goal is to find a way to add security info to a WFS
>> transaction.  I can see three big options:  1. add some bits to the GML
>> feature / WFS call which would set a SimpleFeature's userData or
>> otherwise provide additional data, 2. leverage a WFS 'Native' tag, or 3.
>> store the security info in the SimpleFeature.
>>
>> I've got a pretty good handle on the last two options.  The first option
>> would provide the greatest flexibility, so I'm wondering if anyone can
>> help point me in the right direction there.  (Or otherwise say that this
>> is impossible without re-writing GT/GS XML handling...)
>>
>> Thanks in advance,
>>
>> Jim
>>
>>
>> ------------------------------------------------------------------------------
>> _______________________________________________
>> GeoTools-Devel mailing list
>> GeoTools-Devel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/geotools-devel
>>
>
>
------------------------------------------------------------------------------
_______________________________________________
GeoTools-Devel mailing list
GeoTools-Devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/geotools-devel

Reply via email to