[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
|
Re: [udig-devel] uDIG, GeoTools and MapInfo Tab
|
Balint Persics ha scritto:
Hi,
I am working on a GIS software based on uDIG 1.2 and I have several questions.
I need to implement a MapInfo TAB reader since we have a lot of files
in that format.
I set up the following plan:
0. Use GDAL/OGR Java bindings to read the data.
1. Implement the TAB reader as a GeoTools Data plugin (implementing
org.geotools.data.FeatureReader etc.) Maybe I will port the
unsupported module from GT 2.4 or implement it form scratch, it is not
decided yet.
2. Create a uDIG plugin, implementing a catalog service.
The following questions arose:
1. I want to try out the new optimized renderer in GT 2.7-SNAPSHOT (I
have files with several hundred megabytes and maybe hundreds of
thousands of features).
I think you're reading a bit too much into the renderer improvements.
I made modifications to the streaming renderer and the shapefile data
store so that there is no more a need for a shapefile specific renderer,
but that did not magically made it faster for all other formats: a
data store must implement the support for certain query hints in
a very efficient manner in order to be fast, and it's not always
possible to do so. For example, one of the hints requires the store
to be able to efficiently compute the bbox of a geometry before
reading the geometry itself: that is possible in shapefile, but not
in other data formats.
How to integrate GT 2.7-SNAPSHOT into uDIG
1.2? Simply replace the jars in net.refractions.udig.libs?
I don't think it's so easy, but I personally don't know.
2. How do I add GDAL Java bindings as a dependency to my GeoTools
module? Can I use GDAL 1.6 binaries downloaded from
http://vbkto.dyndns.org/sdk? I noticed that GDAL Java bindings are a
great PITA since the API is not backwards compatible between versions
(eg. ogr.Open(String) vs. ogr.Open(String,int). Is there any API docs
for correctly using GDAL Java bindings?
The module in unsupported was rebuilt months ago against the GDAL 1.7
bindings, but we're still waiting for imageio-ext to upgrade there
to make any actual usage of it, at the moment they are still using
GDAL 1.4.x.
Generally speaking writing a pure java module will give you better
control, better stability, superior performance if you know how
to do high performance java coding (and especially if you know well
NIO), but it's going to be a lot more work too.
Cheers
Andrea
--
Andrea Aime
OpenGeo - http://opengeo.org
Expert service straight from the developers.
