[geotk] NullPointerException with WFSFeatureStore

[johann sorel]

Hello,

I think you can increase the timeout to 60s and throw an exception. I am 
currently occupied with other tasks so if you have a few minutes to 
spare, you should be able to make the changes.

Johann


On 10/12/2014 07:02, Martin Desruisseaux wrote:
> Hello Emmanuel
>
> I just had a look in the code. I saw a hard-coded timeout of 10 
> seconds in the WebFeatureClient class (line 154 and 158). Would it be 
> possible that the server queried by your application take more than 10 
> seconds to answer?
>
> Hello Johann
>
> I case of timeout, the WebFeatureClient.getCapabilities() method at 
> line 159 logs a warning and returns null. However the Javadoc does not 
> said that this method may return null, and indeed the 
> WFSFeatureStore.checkTypeExists() method at line 114 does not verify 
> if the WFSCapabilities object is null, which causes the 
> NullPointerException. What would be the appropriate behaviour?
>
>   * In case of failure, let WebFeatureClient throws an exception? (my
>     preferred choice, since otherwise we get a NullPointerException
>     anyway, which is much less informative).
>   * In case of failure, have WebFeatureClient to return an empty
>     WFSCapabilities? (I don't think it would be safe, since the caller
>     does not know that there is a problem).
>   * Document that WebFeatureClient.getCapabilities() may return null
>     in case of failure, and revisit all caller code for making them
>     null-safe?
>
>
>     Martin
>
>
>
> Le 09/12/14 07:18, Emmanuel Blondel a écrit :
>> Hello,
>>
>> I'm currently assessing the WFS client from Geotoolkit. We would like 
>> to use it in a new project (also because we would like to use Apache 
>> SIS progressively), with the purpose to discover featuretypes and 
>> fetch the featureCollection data, relying on the official GeoAPI.
>>
>> I've been trying to use Geotoolkit 4.0-M2 (which AFAIK is the latest 
>> milestone release), and the geotk-wfs-client.
>>
>> I get the following stack trace :
>>
>> /déc. 08, 2014 11:11:58 PM org.geotoolkit.data.wfs.WebFeatureClient 
>> getCapabilities//
>> //AVERTISSEMENT: TimeOut error, the server takes too much time to 
>> answer. //
>> //Exception in thread "main" java.lang.NullPointerException//
>> //    at 
>> org.geotoolkit.data.wfs.WFSFeatureStore.checkTypeExist(WFSFeatureStore.java:114)//
>> //    at 
>> org.geotoolkit.data.wfs.WFSFeatureStore.<init>(WFSFeatureStore.java:107)//
>> //    at org.fao.fi.wfstesting.App.main(App.java:33)/
>>
>>
>> for this code:
>>
>>         final WebFeatureClient wfsServer = new WebFeatureClient(new 
>> URL("http://www.fao.org/figis/geoserver/ows"), "1.1.0");
>>         final GetCapabilitiesRequest getCapa = 
>> wfsServer.createGetCapabilities();
>>
>>         final GetFeatureRequest getFeature = 
>> wfsServer.createGetFeature();
>>         //simplify WFS like a datastore
>>         final FeatureStore store = new WFSFeatureStore(wfsServer);
>>
>>         String[] typeNames = store.getTypeNames();
>>         for(String typeName : typeNames){
>>             FeatureType ft = store.getFeatureType(typeName);
>>             System.out.println(ft.getName());
>>             System.out.println("==");
>>         }
>>
>> I admit i'm lost. I had a look to the code for which i identified the 
>> timeout warning.
>> But nothing about the exception. Any idea?
>>
>> Thanks in advance for your help
>> Emmanuel
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/geotoolkit/attachments/20141210/891e9016/attachment.html>