On Thu, Mar 12, 2015 at 3:27 PM, Andrea Aime <andrea.a...@geo-solutions.it>
wrote:
> On Thu, Mar 12, 2015 at 7:00 PM, Martin Davis <mtncl...@gmail.com> wrote:
>
>> Now who's talking crazy! :)
>>
>> +0 on using a full templating language. Seems perhaps a bit of overkill
>> for the immediate need? The "macro language" already discussed would
>> suffice for the immediate need I think?
>>
>>
> The thing is, we have no macro language, we have currently just a string
> replacing, and plain assume people can write variable names
> such that they will avoid troubles/overlaps with the rest of the query.
> So how do we scale up to something smarter? Templating feels easier
> because the language, escaping, parsing and whatnot is done for us.
> Of course it's not our language, so we would have to be able to mark that
> a certain query is the complex type, not the %ABC% one.
>
> My worry is to grow the current language that we have, without screwing
> up, and in a way that we can grow it later... so if you feel like you can
> code a solid "parser" for these extensions without using a ready made
> templating language, I guess I'm all for it.
>
It doesn't feel like the parser will be too hard to write, but the
"language execution" part might be a bit trickier....
>
>
>
>> Whatever the syntax used, it seems like the solution has to:
>>
>> - allow altering the geometry column expression based on whether the
>> context is a WMS rendering request or something else (data request,
>> preliminary scan for query schema, etc)
>> - avoid adding an extra spatial filter if one is already provided (or
>> alternatively, add it in such a way that the DB doesn't squawk)
>>
>
> We want to add an optimization that will query against the native fields,
> but the filter against the computed ones should stay, because
> we need this to work _also_ for WFS (the layer is published on both
> protocols).
>
I'm proposing a completely different SQL query for WMS rendering VS other
uses (WFS, layer schema evaluation). WFS should work as usual against the
"raw" dataset, but the WMS rendering query will be able to transform the
geometry into something more appropriate for the request scale.
>
>
>>
>>
------------------------------------------------------------------------------
Dive into the World of Parallel Programming The Go Parallel Website, sponsored
by Intel and developed in partnership with Slashdot Media, is your hub for all
things parallel software development, from weekly thought leadership blogs to
news, videos, case studies, tutorials and more. Take a look and join the
conversation now. http://goparallel.sourceforge.net/
_______________________________________________
Geoserver-users mailing list
Geoserver-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/geoserver-users
