When providing a project DTM/DEM, the default behavior for vector
feature with a z-component is to offset the $z coordinates from the
terrain elevation model in the elevation profile view. However, the
geometry coordinates are supposed to be absolute.
So it makes more sense, to only use the $z coordinate by default in
that case.
The different unit tests are updated accordingly.
Closes: https://github.com/qgis/QGIS/issues/59757
When providing a project DTM/DEM, the default behavior for vector
feature with a z-component is to offset the $z coordinates from the
terrain elevation model in the elevation profile view. However, the
geometry coordinates are supposed to be absolute.
So it makes more sense, to only use the $z coordinate by default in
that case.
Related: https://github.com/qgis/QGIS/issues/59757
This commit contains some changes created during rebasing onto current master, however they are insifignant and should be ignored, as they will match the changes already in master.
* split QgsPostgresSharedData and QgsPostgresUtils to separate file, to allow their use in qgspostgresrasterprovider
* add rasterColumnName
* move tableExists and columnExists to QgsPostgresUtils, add createStylesTable there
* functions to create, load and save styles in DB
* rename variable with typo
* fix function
* drop function that is not needed
* avoid clang warning
* use uri geometryColumn() instead of custom function
* fix docstring
* cannot return min max if triangular mesh does not exist
* only create QgsRenderedLayerStatistics if values changed to prevent loop in rendering
* this was set from two places, remove it from here
* Revert "this was set from two places, remove it from here"
This reverts commit 5d8fbcc321330f1054e5e07cad2ee70c0f9aa661.
* add test case for not existing triangular mesh
* only emit signals here, everything else is handled in renderer creation (avoids double rendering of the mesh layer)
* while updating the widget the component signals need to be blocked
* do not update render here
* add const
* 3D map should not update layer statistics
* update renderer settings from here, signals are emitted from mesh layer
* update test to simulate main canvas extent update
* do rendering in single pass
* emit renderer changed and remove emit of legendChanged
This allows to define a custom tolerance. If this tolerance is
enabled, then this tolerance is used instead of the one defined in the
elevation profile widget.
This custom tolerance is enabled and set to 0 by default for Lines and
Polygons. Indeed, most of the time, only want to use the tolerance for
points.
`MinScaleDenominator` and `MaxScaleDenominator` string are created
from a `QString::number` call with the default parameters. In that
case, the most concise format is used. This can result in the
scientific notation being used for very big or small numbers.
With this change, these parameters are now always displayed in a float
format. Reasons for this change:
- this might prevent a loss of precision for very small or big numbers
- this output is not supposed to be human readable so it does not
matter if is not concise
- this is the logic used by a lot of WMS servers, for example arcgis
server
PR #56369 introduced a regression for WFS layers, because stopping using
'false' and 'true' strings to compare with boolean fields in qgis
expressions, and forcing to use the keywords false and true, it
conflicted with the internal SQLite cache database of WFS layers.
This PR fixes it by storing internal boolean in SQLite cache database in
the right form compatible with QGIS internal functions (i.e. storing '0'
and '1' strings instead of 'true' and 'false' strings).
The typedef is only present in the .sip file, but without this
annotation SIP assumes it also exists in QGIS headers. This didn't
completely break the build, since we concatenate multiple SIP headers
into larger files, so the typedefs from one part "fixed" another. Run
sip-build without --concatenate to see the issue clearly.
This fix is currently sadly not fully effective due to a SIP bug:
https://github.com/Python-SIP/sip/issues/66
