Patel, Ronak Avinash (US SSA) ha scritto:
Hello,
I was going over the documentation about how to create a custom
Datastore from the following links:
http://docs.codehaus.org/display/GEOTDOC/DataStore+Developers+Guide
However, I'm still very confused as to what gets returned in each of the
various methods and how they are used.
You should look at the GeoTools javadocs
Is there an overall architecture diagram/explaination of geoserver and
how it uses its datastores available?
Nope there is none, everything evolved smoothly over time, we don't have
anyone interesting in constantly keeping up to date that kind of
diagram as the code is a moving target.
There is something here but it's very very high level:
http://geoserver.org/display/GEOSDOC/1+GeoServer+Architecture
Is there also better documentation that someone can point me to?
Not that I know of. Anyways, roughly speaking GeoServer gets
the feature type name list and the schema from the DataStore,
then grabs a FeatureSource for all its reading needs, and that
mostly getting a FeatureCollection and a FeatureIterator.
For writing needs we go thru FeatureStore/FeatureLocking instead,
so you need to implement the methods of those only if you need
WFS-T. As far as I know we never access directly the readers
and the writers exposed by datastore, but I may be mistaken.
In GeoTools you'll find two sets of base classes to create a
new data store, AbstractDataStore (in main) and ContentDataStore
(in data), the tutorial illustrates how to use the first, thought
the latter has probably a better design (dev doc wise we're seriously
un-manned, if you are interested in contributing in that direction
it would be most welcomed).
Hope this helps
Cheers
Andrea
--
Andrea Aime
OpenGeo - http://opengeo.org
Expert service straight from the developers.