rewrote renderers where necessary for new node structure
Largely conforming renderers to new NodeValue system. Code seems a lot cleaner this way which is a nice advantage. Likely non-functional as this won't compile just yet and still needs probably another day or two of testing to get it back to where it was before.
This commit is contained in:
@@ -27,12 +27,12 @@ public slots:
|
||||
|
||||
void Render(NodeDependency path);
|
||||
|
||||
virtual QVariant RenderAsSibling(NodeDependency dep) = 0;
|
||||
NodeValueTable RenderAsSibling(NodeDependency dep);
|
||||
|
||||
signals:
|
||||
void RequestSibling(NodeDependency path);
|
||||
|
||||
void CompletedCache(NodeDependency dep);
|
||||
void CompletedCache(NodeDependency dep, NodeValueTable data);
|
||||
|
||||
protected:
|
||||
/**
|
||||
@@ -64,22 +64,22 @@ protected:
|
||||
|
||||
virtual void CloseInternal() = 0;
|
||||
|
||||
virtual void RenderInternal(const NodeDependency& path);
|
||||
virtual NodeValueTable RenderInternal(const NodeDependency& path);
|
||||
|
||||
virtual bool OutputIsAccelerated(NodeOutput *output) = 0;
|
||||
|
||||
virtual QVariant RunNodeAccelerated(NodeOutput *output) = 0;
|
||||
virtual NodeValueTable RunNodeAccelerated(NodeOutput *output) = 0;
|
||||
|
||||
StreamPtr ResolveStreamFromInput(NodeInput* input);
|
||||
DecoderPtr ResolveDecoderFromInput(NodeInput* input);
|
||||
|
||||
virtual FramePtr RetrieveFromDecoder(DecoderPtr decoder, const TimeRange& range) = 0;
|
||||
|
||||
QList<NodeInput*> ProcessNodeInputsForTime(Node* n, const TimeRange& time);
|
||||
|
||||
virtual QVariant FrameToValue(FramePtr frame) = 0;
|
||||
|
||||
QVariant ProcessNodeNormally(const NodeDependency &dep);
|
||||
NodeValueTable ProcessNodeNormally(const NodeDependency &dep);
|
||||
|
||||
virtual NodeValueTable RenderBlock(NodeOutput *output, const TimeRange& range) = 0;
|
||||
|
||||
DecoderCache* decoder_cache();
|
||||
|
||||
|
||||
Reference in New Issue
Block a user