mhugent
1a6840c571
implemented and tested QgsShapeFileProvider::getNextFeature(list<int>). This speeds up rendering of graduated symbol layers a lot (especially if there is much attribute data). However, for the commit, the old getNextFeature(bool) has been enabled again, because getNextFeature(list) has to be implemented in all data providers to not break the other providers.
...
git-svn-id: http://svn.osgeo.org/qgis/trunk@1423 c8812cc2-4d05-0410-92ff-de0c093fc19c
2004-05-17 20:14:01 +00:00
mhugent
b25ce35fb7
added a method getNextFeature(std::list<int>&) to the dataprovider classes (but didn't program it yet). The idea is, that besides copying no attributes (getNextFeature(false)) or all attributes (getNextFeature(true)), it should be possible to copy only the attributes with indexes contained in the passed list. This should improve the rendering efficiency of graduated renderers. I plan to implement this method for the shapefileprovider and make some efficiency tests. For the other data provider, it would be nice to have some help, because i am not very familiar with e.g. the postgresprovider, the delimited text provider or the grass provider
...
git-svn-id: http://svn.osgeo.org/qgis/trunk@1414 c8812cc2-4d05-0410-92ff-de0c093fc19c
2004-05-16 20:39:47 +00:00
mcoletti
4f120dc186
- plugged memory leak in QgsShapeFileProvider::getFirstFeature() whereby the
...
OGRFeature wasn't getting destroyed
- made getFirstFeature() a little more robust with regards to handling memory
related errors
- implementation files should include their own headers first, so moved its
#include accordingly to top; the meant creating class forward declarations
in header file.
git-svn-id: http://svn.osgeo.org/qgis/trunk@1105 c8812cc2-4d05-0410-92ff-de0c093fc19c
2004-03-26 22:29:54 +00:00
gsherman
15a79c50c2
Indentation changes, stub for new getNextFeature function that will be implemented to plug memory leaks
...
git-svn-id: http://svn.osgeo.org/qgis/trunk@1008 c8812cc2-4d05-0410-92ff-de0c093fc19c
2004-03-10 23:21:01 +00:00
mcoletti
ab3b85fdbb
- ran files through GNU indent to make it easier to read code
...
- isValid() was always returning true when, in fact, occasionally the
provider was not in a valid state
git-svn-id: http://svn.osgeo.org/qgis/trunk@978 c8812cc2-4d05-0410-92ff-de0c093fc19c
2004-03-04 20:35:14 +00:00
gsherman
80f901f76c
added intersect option to select to properly select features during select and identify operations. OGR does not support intersects so this is only implemented in the postgres provider
...
git-svn-id: http://svn.osgeo.org/qgis/trunk@739 c8812cc2-4d05-0410-92ff-de0c093fc19c
2004-02-07 19:07:50 +00:00
mhugent
eb6c570363
cleaned up a few things in the shapefile provider
...
git-svn-id: http://svn.osgeo.org/qgis/trunk@734 c8812cc2-4d05-0410-92ff-de0c093fc19c
2004-02-07 14:09:00 +00:00
gsherman
f87e2178d4
added validity checking at the data provider level. Modified logging in postgres provider
...
git-svn-id: http://svn.osgeo.org/qgis/trunk@591 c8812cc2-4d05-0410-92ff-de0c093fc19c
2004-01-25 07:41:36 +00:00
gsherman
47af50b1e3
merge with datasource branch
...
git-svn-id: http://svn.osgeo.org/qgis/trunk@519 c8812cc2-4d05-0410-92ff-de0c093fc19c
2004-01-19 18:17:13 +00:00
