Previously, translucent png images are not displayed properly
due to alpha channels premultiplication.
This patch implements that missing part to support it properly
by introducing the Surface data between canvas engine & rasterizer
@Issue: https://github.com/Samsung/thorvg/issues/655
This reverts commit cd5116b053.
Ah this breaks the Stress example due to Picture::duplicate() is not available...
Need to consider and come back again.
* common: added colorSpace() function
This patch introduces colorSpace() function for SW and GL engine.
* infra: change LoadModule:read() into LoadModule:read(uint32_t colorspace)
This patch changes LoadModule:read() into LoadModule:read(uint32_t colorspace)
* picture: implement passing colorspace into loader
This patch implements passing colorspace into loaders.
Loader->read is now called on the first update.
* external_jpg_loader: support colorspaces
* external_png_loader: support colorspaces
We have encountered that multi-threading usage that user creates,
multiple canvases owned by multiple user threads.
Current sw_engine memory pool has been considered only for multi-threads,
spawned by tvg task scheduler.
In this case it's safe but when user threads introduced, it can occur race-condition.
Thus, Here is a renewal policy that non-threading tvg(initialized threads with zero),
takes care of multiple user threads bu changing its policy,
each of canvases should have individual memory pool to guarantee mutual-exclusion.
@API additions
enum MempoolPolicy
{
Default = 0, ///< Default behavior that ThorVG is designed to.
Shareable, ///< Memory Pool is shared among the SwCanvases.
Individual ///< Allocate designated memory pool that is only used by current instance.
};
Result SwCanvas::mempool(MempoolPolicy policy) noexcept;
All in all, if user calls multiple threads, set memory pool policy to Individual.
If ClipPath is a singular rectangle,
we don't need to apply this to all children nodes to adjust rle span regions.
Rather than its regular sequence,
we can adjust render region as merging viewport that is introduced internally,
All in all,
If a Paint has a single ClipPath that is Rectangle,
it sets viewport with Rectangle area that viewport is applied to
raster engine to cut off the rendering boundary.
In the normal case it brings trivial effects.
but when use SVGs which has a viewbox, it could increase the performance
up to 10% (profiled with 200 svgs rendering at the same time)
Note that, this won't be applied if the Paint has affine or rotation transform.
@Issues: 294
Renamed internal interfaces.
We need both blender & compositor interfaces.
Renamed SwCompositor -> SwBlender which is for pixel joining methods.
Added (SwCompositor, Compositor) which is designed for compositing images.
Splited out ClipPath routine from other pixel compositions'
since yet it's unlikely compatible...
Also revise internal engine interfaces to be simpler.
This is a step forward to enhance masking feature.
Introduce RendererMethod::renderRegion() to return acutal drawing region info.
That is used by scene composition to composite actual partial drawing region
for better performance.
@Issues: 173
Apply tvg Array instead of std::vector
Also Fixed to compList in update() to passed by reference, not copying.
Also Fixed Composition Target memory leak
Here is the binary size result:
[libthorvg.so] 1785376 >> 1607416
[text] 121255 >> 118277
[data] 7792 >> 7736
[dec] 129119 >> 126085
this is an additional enhancement of af8c278c5e
Now scene opacity composition is supported.
Also, this implementaion fixes an incorrect scene bounding box computation.
Plus, adding stroking feathering to shape bounding box size.
Move the prepare stage of shape & stroking composition stage to a separate function
this returns SwImage to use in composite stage.
Also clear partial region buffer since we know composition area.
This implementation supports shape + stroke opacity composition.
Currently, tvg shape provides individual alpha values for filling & stroking
These alpha values are working individually, meaning that if stroking is half translucent,
user can see that translucent stroking is crossed the shape outlines.
Sometimes this result can be expected but user also expects the shape filling is invisible
behind of translucent stroking.
For this reason, Paint provides an additional api opacity()
that applies opacity value to whole paint attributes.
This is a little expensive job, please consider if you can possibly avoid that usage.
See Opacity example.
@Issues: 94
Add RawLoader class that loads and display raw images,
and adds a Rasterizer for image data.
Image data can be loaded via picture.
Loaded image supports Composition, Transformation and Alpha blending.
New API
Result load(uint32_t* data, uint32_t width, uint32_t height, bool isCopy) noexcept;
we introduced shared memory pool for avoiding reallocate memory
while it process the stroke outlines, It experimentally increase
the outline data if we use the allocated memory for multiples shape strokes,
we don't need to alloc/free memory during the process.
This shared outline memory is allocated for threads count
so that we don't interrupt memory access during the tasks.
@Issues: 75
We introduced separate opacity interface to adjust alpha value by paint.
This opacity will affect to whole paint image if paint is a group of paints.
Also, this opacity is to multipy with fill/stroke alpha values.
This means if the opacity is valid, the paint might deal with a composition step,
which is very expensive due to additional rendering step.
One tip is, if you want to toggle on/off for a certian paint,
you can set opacity to 255 or 0.
@API Additions:
Result Paint::opacity(uint8_t o) noexcept;
uint8_t Paint::opacity() const noexcept;
@Examples: examples/Opacity
@Issues: 94
common sw_engine: Implement ClipPath feature
Paint object can composite by using composite API.
ClipPath composite is clipping by path unit of paint.
The following cases are supported.
Shape->composite(Shape);
Scene->composite(Shape);
Picture->composite(Shape);
Add enum
enum CompMethod { None = 0, ClipPath };
Add APIs
Result composite(std::unique_ptr<Paint> comp, CompMethod method) const noexcept;
* Example: Added testClipPath
we should avoid code insertion during file dependencies,
such as #include "xxx.h" which has implementations.
This could increase binary size, we can avoid it as possible.
Current patch improves binary size like this:
From: file(2059008) = text(120360) data(8096) bss(80) dec(128536)
To : file(1921832) = text(118429) data(7872) bss(56) dec(126357)
More additional patches will come in to optmize binary size.
tvg canvas must draw retained shapes for every draw call
even though user missed call update() for shapes.
that case canvs must draw shapes without update,
it means drawing them within previous condition.
previous implementation didn't consider multiple canvases,
multiple canvas shared one renderer engine that brought corrupted contexts.
Thus, each canvas instances should have designated renderer engine instances.
Now fixed.
Now, we have 2 points for asynchronous behaviors.
1. update shapes:
Each shape update will be performed by async when you push shape to canvas.
Meaning, if you have time gap between update and rendering in process main-loop,
you can have a benefit by this.
2. rasterization by canvas:
Canvas.draw() will be performed asynchnously until you call canvas.sync();
Meaing, if you can trigger tvg rendering eariler than composition time.
You can have a benefit by this.
If these 1, 2 points might not work for your program,
You can just toggle off async by setting threads number zero at initialization.
Or if you could apply either point of them for your program,
It might be good for performance.
But the best approach is to make both async properly.
Though this might need to fine-grained tuning integration between your program & tvg,
You could achieve the best peformance by parallelzing tasks as possible without any jobs delaying.
Change-Id: I04f9a61ebb426fd897624f5b24c83841737e6b5b
We can use RGBA colorspace rather ARGB for pixel data.
This would be better for many rendering system,
since it's more widely preferred than ARGB including opengl.
Change-Id: Ibbfe6a511d77bf0ef30ce261995467c11164d306
we can't control any threads count that could drop the performance.
remove async() and will come back with fine-tuned threading-pool.
Change-Id: I17c39792234acfce6db334abc0ce12da23978a9a
We prefer to build up a tiny compact engine at memory rather than compatibility,
this engine is not considerd for end-users but designed for middle-level framework
and some low-level users.
Thus, we won't consider 64bits data size,
use explicit 32 bits data until coming next upgrade...
Change-Id: I0704d5f1e0eb909cccc10922bc5972e115fbbcc0
Each scene could conserve its own transformation,
the origin could be the parent Paint (if they were beloned to)
Thus, you can compose the multi-tranformed scene group which
simultenoulsly working in own spaces.
See testTransform and testSceneTransform, how they work.
Change-Id: I51deb4d66f8fcd024f8dc7a1e1af57c398a9d7fe
Paint supports translate, rotate, scale functions for transformation
The origin of these transformation is center of the paint,
thus you have to consider the center-aligned vertices if you'd like to use
these transformation functions.
This policy has been considered for scene transformation.
Change-Id: I78b63d7965faec0ec5b9a98a7776993744534b54
Come to think of it, this optimized method is not so useful,
it could just bring the user misunderstanding and
not to efficient as I expected in the most cases.
So, changed policy for transformation behaviors.
it keeps the properties as others but leaves it to the backend implementation.
Plus, this change contains the correct RenderUpdateFlag.
You can refer the flag in the backend to figure out which kinds of properites has been updated
Change-Id: Ibe0494712598a8161950b9ae2e22ac45bed1c47b
PaintNode -> Paint
ShapeNode -> Shape
SceneNode -> Scene
We can keep clean and neat shorter names.
Change-Id: Ic8521d456d947985e5fbe1ba2bde06faa1f52469
