[Mapserver-dev] layerObj resultcache?

[Howard Butler]

536 has included some enhancements as well as fixing the OBJECTID/SE_ROW_ID 
bug.  Enhancements include supporting version-specific queries through the 
connection string (it will use default if not specified) and using new 
constructs of the SDE C API for querying (se_queryinfo instead of 
SE_SQL_CONSTRUCT).  The benefits of the new construct is that SDE will be 
able to support joins with this method, and it is a bit faster (SDE 
pre-joins the feature and attribute tables in the database instead of 
querying them individually and joining them client-side).  ESRI has not 
actively developed the SE_SQL_CONSTRUCT stuff since SDE 3.0 days as far as 
I know.

The problem right now with the new changes is that SDE for Coverages is 
causing all kinds of breakage because it looks like a duck but barks like a 
dog.  I am changing some of the calls that I make with the new code to be 
sensitive to SDE for Coverages and work around it, but when I updated to 
the newest mapscript Friday, all of my mapscript test code broke.  I rolled 
back to pre-change SDE code and am still getting the same errors in my test 
code, so I suspect that something has changed.

Sean's latest mapscript docs 
<http://users.frii.com/sgillies/projects/mapscript/mapscript.html> say 
nothing about resultcache or resultcache object any more, and that is where 
my code is hanging (query the layer for an attribute and set the map extent 
to the bounds of the resultcache).


At 11:20 AM 4/12/2004, you wrote:
>Howard Butler wrote:
>>I upgraded to CVS version of mapscript to test my SDE changes for the 
>>next release.  Some of my past Python code depended on 
>>layer.resultcache  Did this get renamed?  My inspection of mapscript.i 
>>didn't show me much.
>
>You probably have to use the getResult(int i) method of the layerObj. I 
>don't know if it's new or not in the SWIG MapScript, but that's always 
>been the way to go with PHP MapScript.
>
>You talk about "SDE changes for the next release". Are you talking about 
>bug 536 or about new changes?
>http://mapserver.gis.umn.edu/bugs/show_bug.cgi?id=536
>Should we mark this bug with the "4.2 release" target milestone, Steve?
>
>Daniel
>--
>------------------------------------------------------------
>  Daniel Morissette               dmorissette at dmsolutions.ca
>  DM Solutions Group              http://www.dmsolutions.ca/
>------------------------------------------------------------