This patch fixes a WMS server bug when "Short name" was
set in project settings for service capabilities.
When the short name was set, it was not possible
to load the whole WMS by selecting the root
layer named after the short name.
With tests for both cases (with and without short name).
Funded by Kanton Zug
(fix#20586).
Without this parameter it is not possible to remove collars surrounding
input raster which may overlap with other input rasters. As this is very
frequent case algorithm is useless without such parameter. To keep API
compatibility new parameter is optional and not used by default.
flag in a single call
This is much more efficient then making two calls, since the
QgsRasterBlock::isNoData() check internally calls QgsRasterBlock::value().
So by requiring API users to make the two separate calls individually,
we double the time this process takes...
This allows users to pass additional command-line arguments which are
not exposed in the algorithm definition. The most frequent use case is
enabling transparency and adding nodata values.
actually fetching the raster block data
This allows for efficient iteration over a "reference" layer, where
you require the block extent/origin/pixel size/etc (but not the
reference layer block data itself!), in order to fetch a block from
a DIFFERENT set of rasters (but keeping these pixel-aligned to the
reference raster).
Followup https://github.com/qgis/QGIS/pull/8830 that fixed
a regression with rewritten urls in the server, unfortunately
my original solution introduced a side-effect on the POST
request, with the new approach I'm introducing a new method
to retrieve the URL as seen by the web server: by default
this is the same URL seen by QGIS server, but in case
a rewrite module made some changes, the original URL will
be used as a base URL if not overridden by a config setting.
This PR comes with an extended set of tests that should
cover both (rewritten and unrewritten) cases for GET and
POST and for WFS/WFS/WCS and WMTS.
This is a small refactoring of 3D point symbol code, with three goals:
- use the new handler class just like line and polygon symbols already use
- just one feature loop instead of two and avoiding possibly slow query
that requests all features by their IDs
- make it possible to use point symbols with rule-based renderer
