[mapserver-dev] Re: INSPIRE Category Layer (may affect 6.0 release)

Martin Kofahl m.kofahl at gmx.net
Mon Jan 24 15:34:26 EST 2011


Thomas,

you're probably right expecting problems with sld. I could think the 
following solutions:

a) When using dummy layers as groups, the hidden layers for drawing may 
have the same NAME tag at the LAYER level. This might require reviewing 
of existing code in order to allow layers with identical name (May be it 
already works? I did not check.). SLD would have to be applied to 
multiple layers.

b) In stead of NAME the hidden layers for drawing may have a REFNAME or 
similar. All actions must apply to all layers which reference the 
requested layer name (may be easiest by adding a table with the 
ref-layer-id's to each real layer). Probably much coding work at 
different locations...


Regarding the drawing order, this will be a problem when a client 
requests a specific drawing order. This could be solved using REFNAME, 
too. Otherwise you may set the correct drawing order in the map file for 
theREQUIRES ... layers.


Regarding GetFeatureInfo most simple way is to have an overall dataset 
which is queryable by the group layer itself (but not drawn). With the 
above mentioned solution a) I see problems with HEADER and FOOTER. This 
would also suggest a solution like REFNAME.


Martin


On 24.01.2011 14:41, tellett wrote:
>
> Hi all,
>
> I like Martin and Barts idea on this. Assefa has already said it would be
> better to use an existing wms mechanism to implement the neccesary changes.
> Just one issue to mention is drawing rules. One of the main reasons for this
> change is to allow multigeometry layers (through nested layers), and when
> this is enabled its important that some drawing rules apply so that all
> features can be seen (polygons first, polylines second etc).
>
> I also think the use of dummy group and nested layers and the REQUIRES
> keyword might be a little problematic when it comes to the GetFeatureInfo
> call and applying SLD's, but I'm not 100% sure on this yet.
>
> Best regards




More information about the mapserver-dev mailing list