This allows nice and simple, elegant construction of checks for
Python.
To use, Python based checks should use the decorator syntax:
from qgis.core import check
@check.register(type=QgsAbstractValidityCheck.TypeLayoutCheck)
def my_layout_check(context, feedback):
results = ...
return results
Or, a more complete example. This one throws a warning when attempting
to export a layout with a map item set to the Web Mercator projection:
@check.register(type=QgsAbstractValidityCheck.TypeLayoutCheck)
def layout_map_crs_choice_check(context, feedback):
layout = context.layout
results = []
for i in layout.items():
if isinstance(i, QgsLayoutItemMap) and i.crs().authid() == 'EPSG:3857':
res = QgsValidityCheckResult()
res.type = QgsValidityCheckResult.Warning
res.title='Map projection is misleading'
res.detailedDescription='The projection for the map item {} is set to <i>Web Mercator (EPSG:3857)</i> which misrepresents areas and shapes. Consider using an appropriate local projection instead.'.format(i.displayName())
results.append(res)
return results
Better to use QVector here, because the QgsLineString/QgsGeometry
methods all use QVector too, and we want to avoid unnecessary
list->vector conversions.
These algorithms allow users to convert z or m values present in feature
geometries to attributes in the layer. By default the z/m value from the
first vertex in the feature is extracted, but optionally statistics
can be calculated on ALL the z/m values from the geometry (e.g. calculating
mean/min/max/sum/etc of z values).
Like the vector zonal stats algorithm, but this one works with
the zones defined in another raster.
Iterates over the input rasters in blocks to be nice and
memory efficient.
From the algorithm help:
"This algorithm calculates statistics for a raster layer's
values, categorized by zones defined in another raster layer.
If the reference layer parameter is set to "Input layer",
then zones are determined by sampling the zone raster layer
value at the centroid of each pixel from the source raster
layer.
If the reference layer parameter is set to "Zones layer",
then the input raster layer will be sampled at the centroid
of each pixel from the zones raster layer.
If either the source raster layer or the zone raster layer
value is NODATA for a pixel, that pixel's value will be
skipped and not including in the calculated statistics."
