Line and polygon pattern symbolizers need to support the render time scale_factor, and both pattern symbolizers and the point symbolizer need to support the transform.scale property (which only right now works with svg symbols).
This is as followup on #1274 (and #695)
See also #952
@springmeyer - we should support affine_transforms for images. AGG has got all required pieces.
@artemp - yes, sounds good - though I think just supporting .scale() should be considered if full affine transforms are hard to get done before release, given that scaling is the only commonly requested feature for bitmap images.
+ affine transform for raster markers - #1279
195b8ad - affine transform for raster markers (PointSymbolizer atm)
@artemp Hi, I'm getting raster markers that seem te be scaled twice when using scale_factor. I think this is related to commit 195b8ad, but I can't exactly tell what's wrong.
@JeroenVu - could you post .xml style, thanks ?
The problem could be that process(point_symbolizer) calls render_marker (src/agg/process_point_symbolizer.cpp +97) with pre-scaled transformation matrix (src/agg/process_point_symbolizer.cpp +72), while other calls pass raw image-transform matrix.
@lightmare @JeroenVu - yep, I can replicate it
@JeroenVu @lightmare - should be fixed in 774e1b1
Give it a try and let me know.
@artemp Thank you for looking at this. I built mapnik from source and regenerated the map, but apparently this isn't fixed yet.
@JeroenVu - are you using MarkersSymbolizer or PointSymbolizer ? The fix is for MarkersSymbolizer.
Pls, post what are you feeding into Mapnik so I can investigate, thanks.
@artemp Sorry, I just noticed I said "raster marker" in my first comment. I'm indeed using PointSymbolizer.
@JeroenVu - no worries, you should be able to achieve the same functionality using MarkersSymbolizer. We're planning to merge PointSymbolizer and MarkersSymbolizer in 2.1 . Anyways, I'll fix PointSymbolizer for now, too.
looks like PointSymbolizer was done in 195b8ad. This is working good, a followup now is: #1364 and #952