implicitly shared copy of an internal cache instead of recreating the
coordinate sequence again and again.
Improves performance of the nodetool on large features a lot (refs #13963)
Also introduce Qgs(Coordinate|Ring|Point)SequenceV2 typedefs.
- removes storage of bounding box from QgsPointV2 (gives significant
decrease in size of object)
- more efficient calculation of bounding box for linestrings
Additionally, this commit moves the bounding box invalidation to a
virtual QgsAbstractGeometryV2::clearCache() method, so that other
non-bounding box caches can also be cleared when the geometry
is modified.
- Setting exterior ring will force all interior rings to match
dimensionality of exterior ring
- add equality operator
- use correct WKB types for rings when creating polygon from WKB
- segmentize curves when setting exterior ring with curved segments
- fix export to GML2/3 (should use LinearRing, not LineString as
ring element tag)
- fixes for handling Polygon25D
- add QgsAbstractGeometryV2::convertTo( QgsWKBTypes::Type type ) for
easy conversion between geometry types
- fix crash when calculating perimeter with no exterior ring
- ensure that added rings respect dimensionality of polygon (avoids
issues such as polygons with z having a ring without z, or a
Polygon25D with LineStringZ rings)
- if a curved ring is added to a polygon then a segmentized version
of the ring is used (can't have a Polygon with a CircularString
ring)
- when calling setInteriorRings, make sure empty rings are skipped
and that all rings are converted to correct type for polygon
- don't crash when requesting or removing interior ring with
index < 0
Split length from perimeter calculation in geometry API, as
returning perimeter for length for polygons is misleading and
results in incorrect length/perimeter calculations for mixed
geometry collections.
Enable length & perimeter unit tests against reference geometries.
Now the length & perimeter values match those calculated by
PostGIS.
Rationale:
- there was a lot of large objects passed by value, so potentially
there's a speed bump from this
- even for implicitly shared classes like QString/QList there's still
a (small) cost for copying the objects when there's no reason to
- it's the right thing to do!
