Hi Nuno,
The changes make sense, unfortunately the patch still has issues. The patch
contains largely formatting changes. Looks like spaces were converted to
tabs at some point. This unfortunately makes it nearly impossible to figure
out what actually changed in that class. Seems to be mostly
the PostGISDialect class, but also the test classes contain misformatting
in terms of tabs vs spaces.
Not sure what IDE you are using but if you are using eclipse there are
existing formatting templates that will help with this. See this for more
details.
http://docs.geotools.org/latest/developer/conventions/code/style.html
Again, to reiterate, for any patches that are proposed it is imperative
that they only contain relevant changes, and no formatting changes.
-Justin
On Wed, May 2, 2012 at 8:22 AM, Nuno Miguel Carvalho Oliveira <
[email protected]> wrote:
> Hi,
>
> (Sorry for the late reply, i had a very very busy week).
>
> First, thank you for your very complete answer.
> I create a new patch based on your advices (in annex).
> Basically to read 3D coordinates from a PostGIS data
> base we need to put a hint in the query:
>
> Hints hints = query.getHints();
> hints.put(Hints.FEATURE_2D, Boolean.FALSE);
> query.setHints(hints);
>
> This hint will remove 'ST_Force_2D' from the final querry
> and set the encode to 'ST_AsEWKB' instead of 'ST_AsBinary'.
>
> I run the tests and all works fine (PostGIS3DTest).
>
> To resume what i have changed:
>
> - SQLDialect.java and PostGISDialect.java: I extend the
> 'encodeGeometryColumn' function to take also the 'Hints'
> as a parameter.
> - JDBCDataStore.java: I give the 'Hints' as a parameter
> to the 'encodeGeometryColumn' function.
> - PostGIS3DTest and PostGIS3DTestSetup: I add 3D tests
> to PostGIS.
>
>
> Any suggestions are very welcome.
>
> Regards,
>
> Nuno Oliveira
>
>
> 2012/4/21 Andrea Aime <[email protected]>
>
>> On Thu, Apr 19, 2012 at 7:01 AM, Nuno Miguel Carvalho Oliveira <
>> [email protected]> wrote:
>>
>>> Hi,
>>>
>>> Follows in anexe two patchs, one to integrate W3DS as a community module
>>> in GeoServer and another to GeoTools who permits the acess of 3D data from
>>> a PostGIS data base. Related to the acess of 3D data in GeoTools maybe it
>>> exists a better way to do this, any idea is welcome.
>>>
>>>
>> Hi,
>> let's start discussing the Geotools patch, since it's really short.
>> The patch is almost correct, but it's mis-using the hints somehow.
>>
>> Hints are used in the query as a way to tell the store that we'd
>> like the returned objects to have some particular property, such as being
>> built with a particular geometry factory, or being flat (2D).
>> The geometry column user data is instead used to carry some extra info
>> that describes it, such as the fact the column is a "geography" instead
>> of a
>> "geometry".
>>
>> Query hints are created per request and can vary over time, for example
>> WMS will always ask for 2D geometries, whilst W3DS or WFS won't.
>> Geometry descriptor attributes instead are structural, once set they
>> describe
>> something about the geometry column "forever".
>>
>> As such you should not mix them, otherwise one request can force the
>> geometry to be 2d by attaching the hint and the next request looking for
>> 3d data won't get it (because the hint is already in place).
>> To do what you want is better to extend the encodeGeometryColumn in
>> SQLDialect to take also the Hints as a parameter: it is an API break that
>> I believe we can take on trunk (Justin, thoughts?).
>>
>> The other thing that's missing is a test.
>> The JDBC testing framework is shared between all the databases and is
>> setup so that there is a base abstract class which needs to be subclasses
>> and return a "test setup" object that knows how to setup the expected
>> spatial tables with the DDL specific of that particular database.
>> You can extend JDBC3DTest and JDBC3DTestSetup, which are
>> now implemented only for Oracle and DB2, and create an implementation
>> for PostGIS.
>>
>> I'll try to have a look at the W3DS community module too, but as long as
>> it's a
>> community module you don't really need to pass any review.
>> One thing is important tough, in order for GeoServer to include your
>> module
>> in the future you, or the company/authority/adminstration that employs you
>> have to sign and send a "copyright assignment" contract that gives
>> Openplans,
>> the no-profit managing GeoServer, rights on the source code that you
>> created.
>> This is necessary to keep the ownership in one place and defend GeoServer
>> in
>> court should a license violation arise.
>>
>> Cheers
>> Andrea
>>
>> --
>> -------------------------------------------------------
>> Ing. Andrea Aime
>> GeoSolutions S.A.S.
>> Tech lead
>>
>> Via Poggio alle Viti 1187
>> 55054 Massarosa (LU)
>> Italy
>>
>> phone: +39 0584 962313
>> fax: +39 0584 962313
>> mob: +39 339 8844549
>>
>> http://www.geo-solutions.it
>> http://geo-solutions.blogspot.com/
>> http://www.youtube.com/user/GeoSolutionsIT
>> http://www.linkedin.com/in/andreaaime
>> http://twitter.com/geowolf
>>
>> -------------------------------------------------------
>>
>
>
--
Justin Deoliveira
OpenGeo - http://opengeo.org
Enterprise support for open source geospatial.
------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Geoserver-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/geoserver-devel
