As soon as I upgradeed to 1.6.0b4 (and later to 1.6.0rc1), I noticed the use of "*" as propertyname, as in:
http://localhost:8080/geoserver/wfs?
request=GetFeature
&version=1.0.0
&propertyname=*
&featureversion=1
&typename=sdm:GEOREGIONISTAZIONI
triggered a "java.lang.IndexOutOfBoundsException: Index: 0, Size: 0
java.lang.IndexOutOfBoundsException: Index: 0, Size: 0", which didn't occur with 1.5.x; if I use an empty value for the propertyname, it works fine.
I've searched Jira but this issue seems non to be in the list: what should I do ?
Regards,
--------------------
Luca Morandini
www.lucamorandini.it
--------------------
Is propertyname=* part of the spec? It's just supposed to retrieve all?
Whey don't you just not use the propertyname parameter?
Chris
Luca Morandini wrote:
As soon as I upgradeed to 1.6.0b4 (and later to 1.6.0rc1), I noticed the use of "*" as propertyname, as in:
http://localhost:8080/geoserver/wfs?
request=GetFeature
&version=1.0.0
&propertyname=*
&featureversion=1
&typename=sdm:GEOREGIONISTAZIONI
triggered a "java.lang.IndexOutOfBoundsException: Index: 0, Size: 0
java.lang.IndexOutOfBoundsException: Index: 0, Size: 0", which didn't occur with 1.5.x; if I use an empty value for the propertyname, it works fine.
I've searched Jira but this issue seems non to be in the list: what should I do ?
Regards,
--------------------
Luca Morandini
www.lucamorandini.it
--------------------
-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Geoserver-users mailing list
Geoserver-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/geoserver-users
!DSPAM:4005,4742007777351628642973!
Chris Holmes wrote:
Is propertyname=* part of the spec? It's just supposed to retrieve all?
<<A list of properties may be specified for
each feature type that is being queried.
Refer to subclause 14.2.2 on how to form
lists of parameters. A "*" character can be
used to indicate that all properties should
be retrieved. There is a 1:1 mapping
between each element in a FEATUREID or
TYPENAME list and the
PROPERTYNAME list. The absense of a
value also indicates that all properties
should be fetched.>>
From the WFS 1.1.0 spec, page 99.
Whey don't you just not use the propertyname parameter?
Sure I can... but the use of "*" should be allowed too, if I understand the spec correctly.
Regards,
--------------------
Luca Morandini
www.lucamorandini.it
--------------------
Chris Holmes wrote:
Is propertyname=* part of the spec? It's just supposed to retrieve all?
<<A list of properties may be specified for
each feature type that is being queried.
Refer to subclause 14.2.2 on how to form
lists of parameters. A "*" character can be
used to indicate that all properties should
be retrieved. There is a 1:1 mapping
between each element in a FEATUREID or
TYPENAME list and the
PROPERTYNAME list. The absense of a
value also indicates that all properties
should be fetched.>>
From the WFS 1.1.0 spec, page 99.
Whey don't you just not use the propertyname parameter?
Sure I can... but the use of "*" should be allowed too, if I understand the spec correctly.
Regards,
--------------------
Luca Morandini
www.lucamorandini.it
--------------------
Luca Morandini wrote:
Chris Holmes wrote:
Is propertyname=* part of the spec? It's just supposed to retrieve all?
<<A list of properties may be specified for
each feature type that is being queried.
Refer to subclause 14.2.2 on how to form
lists of parameters. A "*" character can be
used to indicate that all properties should
be retrieved. There is a 1:1 mapping
between each element in a FEATUREID or
TYPENAME list and the
PROPERTYNAME list. The absense of a
value also indicates that all properties
should be fetched.>>
From the WFS 1.1.0 spec, page 99.
Huh. I thought I knew the WFS spec and don't remember that one, though maybe it wasn't in 1.0? CITE doesn't test it, or else it'd definitely be in there.
Could you make a bug report for it? Schedule against RC2.
thanks,
Chris
Whey don't you just not use the propertyname parameter?
Sure I can... but the use of "*" should be allowed too, if I understand the spec correctly.
Regards,
--------------------
Luca Morandini
www.lucamorandini.it
--------------------
-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Geoserver-users mailing list
Geoserver-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/geoserver-users
!DSPAM:4005,47420c86107401015089218!
Chris Holmes wrote:
Huh. I thought I knew the WFS spec and don't remember that one,
Glad an extra pair of eyeballs may help sometimes 
though maybe it wasn't in 1.0? CITE doesn't test it, or else it'd definitely be in there.
It was in 1.0 too, at page 65.
Could you make a bug report for it? Schedule against RC2.
Done, GEOS-1525.
Regards,
--------------------
Luca Morandini
www.lucamorandini.it
--------------------