The centroid of the geometry is used for the point parameter value.
This makes it easier to write expressions for the value of point
parameters, since all the QGIS expression functions for working
with geometry types return QgsGeometry value themselves (e.g.
make_point, centroid, ...). In this case it's much nicer to
allow expression values like `make_point(3,4)` within a precalculated
expression based value in a Processing model.
In QGIS Server 2.* the WMS GetPrint request could genrate JPEG image. QGIS Server 3.4 has lost this capabilities.
```
<ServiceExceptionReport xmlns="http://www.opengis.net/ogc" version="1.3.0" capture-installed="true">
<ServiceException code="InvalidFormat">
Output format jpg is not supported by the GetPrint request
</ServiceException>
</ServiceExceptionReport>
```
To fix this regression, it is necessary to accept JPEG output format.
And to avoid this regression to come back, the QGIS Server tests has been updated to accept jpg image test.
This algorithm creates a polygon layer containing the extent
of a print layout map item, with attributes specifying the map
size (in layout units), scale and rotatation.
The main use case is when you want to create an advanced overview
indicator and the inbuilt layout tools to do this don't suffice.
are displayed
Turns out there is NO reliable way to create a proxy model
which adds new rows to a model, so just bite the bullet and
do it in the underlying model (yuck)
[feature] [mesh] #20731 add option to format time in mesh layer
There is new settings dialog that can be opened by button next to time slider. User can set reference time and time format (e.g. 2019-03-21 22:01:11).
This adds a new parameter type specifically for map scales,
QgsProcessingParameterScale. The values are evaluated using
self.parameterAsDouble, which returns the map scale
denominator (matching the standard in other parts of the
QGIS API).
Scale parameters are displayed to users using the standard
QgsScaleWidget, which includes the combo box of predefined
scales and a shortcut button to match the current map scale.
These algorithms calculate the boolean OR or AND for a set of input
rasters. For AND, if all of the input rasters have a non-zero value
for a pixel, that pixel will be set to 1 in the output raster, otherwise
it will be set to 0. For OR, if ANY of the input rasters have a non-zero
value for a pixel, that pixel will be set to 1 in the output raster,
else 0.
A reference layer parameter specifies an existing raster layer to use
as a reference when creating the output raster. The output raster will
have the same extent, CRS, and pixel dimensions as this layer
By default, a nodata pixel in ANY of the input layers will result in
a nodata pixel in the output raster. If the 'Treat nodata values
as false' option is checked, then nodata inputs will be treated the
same as a 0 input value.
Makes for much simpler raster boolean logic calculation without
the complexity of using the raster calculator (and that's not
always possible to do anyway, e.g. when ANY of the input rasters
has a nodata pixel). It's also scalable dynamic to any number of
input rasters (unlike raster calc), so is more flexible when
used within models.
