This is related to commit e0d38ba3f912b7ae6b34e66a44e529ec37394458. In case
we have to deal with an inconsitent raster where the nodata value is set to
a value not representable in the data type, ignore it. Otherwise, a NaN value
in a Byte raster would be cast as 0.
QGIS had two problems:
1) It was using the specified legend URL only if its mime type was matching
the layer's mime type. There is no reason for that.
2) When QGIS was using the default layer (empty string), it was not even
trying to find out in what style to pick the legend URL.
When there is a layer group with several sub-layers, the group has a
"default" style and the sub-layers each have a "default" layer. QGIS
was showing two "default" styles for the sub-layers, which would be
confusing to the user and could pick the wrong legend for the
sub-layer if the user picked the wrong entry (the first one).
Had to create a static lib for wmsprovider in order to unittest it.
expecting geometry inputs, then use that feature's geometry
instead of throwing an error
eg
buffer(@atlas_feature,100)
or
buffer(get_feature('my_layer','id=54'))
This commit makes composer automatically georeference outputs
(where output format makes this possible, eg TIF and PDF).
The existing option to create a world file has been separated
from the map selection for georeferencing. The new behaviour
is to always georeference outputs, and only create the
separate world file if that option is checked.
Currently if the source raster has no nodata value, the writer will still write
nan as the output nodata value, ignoring the flag that specifies if there is a
nodata value or not. On a raster with byte data type, this will cause
confusion on reading since the nodata value will be somehow cast as 0.
Make QgsRasterChecker check for nodata consistency between source and target, and
add a test file that shows the issue.
This commit simplifies the geometry just before to paint it in the
current QPainter device.
This commit allows to override the '>>' operators of the QgsWkbPtr
class. Also, It defines a new QgsWkbSimplifierPtr class to automatically
simplify the input point stream.
Integration tests for WFS-T provider and WFS-T server.
The tests perform WFS-T CRUD operations on a QGIS Server
instance through WFS-T data provider.
Funded by Boundless
For other providers features will automatically be made valid
through the proces of adding to the provider's storage itself
and then later retrieving via iterators. But the memory
provider uses a direct copy of the feature, so if we don't
explicitly mark features as valid the provider may be
returning features incorrectly marked as invalid.
...and add tests to ensure all providers always return valid
features
This commit adds a new mode to the attribute table dialog for searching
and filtering features. When activated (using a button on the toolbar
or by pressng CTRL+F), the dialog will switch to form view and all
widgets are replaced with their search widget wrapper variant.
Alongside each widget is a tool button with options for controlling
the search/filter behaviour for that field, eg "equal to", "not equal
to", "is null", "greater than", etc.., with the options presented
matching themselves to the corresponding field and widget type.
New buttons appear at the bottom of the form for either selecting
matching features (with options for add to selection/remove from
selection/select within current selection) or filtering features
in the table (with options for adding features to a current filter
or further restricting a current filter).
Sponsored by SIGE
- add selectByRect( QgsRectangle&, SelectBehaviour) and
selectByIds( QgsFeatureIds, SelectBehaviour) for selecting
by rect and ids respectively, with options to add to selection/
remove from selection/intersect with current selection
- deprecate select( QgsRectangle ) and setSelectedFeatures in
favour of new methods
- add unit tests
Makes it simple for scripts to select by expression. The method
also accepts a parameter which dictates whether matching features
are added to an existing selection, removed from the selection
or intersected with the current selection.
The existing code from the select by expression dialog has been
moved to QgsVectorLayer, and optimised for maximum possible speed.
Also added unit tests.
- The group is renamed as 'Select fields to export and their export options'
- It has no longer a checkbox. It is just collapsible. So
QgsVectorLayerSaveAsDialog::attributeSelection() is deprecated and always
return true.
- For most formats, all attributes are selected by default
- For CSV/XLSX/ODS, if they have edit widgets, the corresponding checkbox in
"Replace with displayed values" column is also checked by default.
- For DXF, keep existing behaviour: attributes are unchecked and not checkable
- For KML, keep existing behaviour: attributes are unselected by default
This commit adds a number of different forms of aggregates to
the expression engine.
1. Aggregates within the current layer, eg sum("passengers")
Supports sub expressions (ie sum("passengers"/2) ), group by
( sum("passengers", group_by:="line_segment") ), and optional
filters ( sum("passengers", filter:= "station_class" > 3 ) )
2. Relational aggregates, which calculate an aggregate over
all matching child features from a relation, eg
relation_aggregate( 'my_relation', 'mean', "some_child_field" )
3. A summary aggregate function, for calculating aggregates
on other layers. Eg aggregate('rail_station_layer','sum',"passengers")
The summary aggregate function supports an optional filter,
making it possible to calculate things like:
aggregate('rail_stations','sum',"passengers",
intersects(@atlas_geometry, $geometry ) )
for calculating the total number of passengers for the stations
inside the current atlas feature
In all cases the calculations are cached inside the expression
context, so they only need to be calculated once for each
set of expression evaluations.
Sponsored by Kanton of Zug, Switzerland
Can be used to store the results of expensive sub-expression
calculations (eg layer aggregates), so that future expression
evaluation using the same context does not have to recalculate
the cached values.
