The spec generally just asks that feature ids be unique, and should be
stable across different representations. This gets tricky when the same
feature is represented in two different datasets - something we have never
(as an industry) figured out how to address.
As a quick way to prevent two featu
I have an interesting pull request
https://github.com/geotools/geotools/pull/431
This comes out of checking that the ECQL generated for Id filter can be
parsed ... turns out it could not. I went through and found several more
ECQL expressions to fix (EXIST, RELATE, IN).
--
Jody Garnett
--
Title: Message Title
Kevin Smith created an issue
Hello Devs,
in uDig we have an ticked [1] that featuretype names are take from filename
from the last Slash till the first point occurrence.
For most of the files this is fine but some file haven version numbers or
something else in the name, like
my_filename.whatever.v1.5.shp
The expectation i
Title: Message Title
Jeroen Dries created an issue
Niels I noticed that one of the pull requests mentions wfs-ng:
https://github.com/geotools/geotools/pull/406
Have you had a look at this one?
Jody Garnett
On Fri, Apr 25, 2014 at 6:42 PM, Niels Charlier wrote:
> Hello,
>
> While working on wfs-ng, I found the class AbstractDataStoreTest and I
I had wanted to set up a conformance test that datastore implementors,
justin ended up making this approach work for JDBC datastore
implementations.
For AbstractDataStore everyone made a copy! ... but since most of them have
died now wfs is the last one standing.
Jody Garnett
On Fri, Apr 25, 2
Hello,
While working on wfs-ng, I found the class AbstractDataStoreTest and I
was wondering what the deal is with that class:
- there is a copy inside the wfs-ng module, and one inside the main
module. The first appears to be a modified version of the latter.
Shouldn't the modified version move
