archive/blender-archive
Archived
1
1
Fork 0
Code Pull Requests Packages Activity
This repository has been archived on 2023-10-09. You can view files and clone it, but cannot push or open issues or pull requests.
Files
a262ea8c47f8106f66e16935c556f383ef306861
blender-archive/source/blender/compositor/intern/COM_NodeConverter.h

122 lines
4.3 KiB
C++
Raw Normal View History

Structural cleanup and improvements for the compositor. Many parts of the compositor are unnecessarily complicated. This patch aims at reducing the complexity of writing nodes and making the code more transparent. == Separating Nodes and Operations == Currently these are both mixed in the same graph, even though they have very different purposes and are used at distinct stages in the compositing process. The patch introduces dedicated graph classes for nodes and for operations. This removes the need for a lot of special case checks (isOperation etc.) and explicit type casts. It simplifies the code since it becomes clear at every stage what type of node we are dealing with. The compiler can use static typing to avoid common bugs from mixing up these types and fewer runtime sanity checks are needed. == Simplified Node Conversion == Converting nodes to operations was previously based on "relinking", i.e. nodes would start with by mirroring links in the Blender DNA node trees, then add operations and redirect these links to them. This was very hard to follow in many cases and required a lot of attention to avoid invalid states. Now there is a helper class called the NodeConverter, which is passed to nodes and implements a much simpler API for this process. Nodes can add operations and explicit connections as before, but defining "external" links to the inputs/outputs of the original node now uses mapping instead of directly modifying link data. Input data (node graph) and result (operations graph) are cleanly separated. == Removed Redundant Data Structures == A few redundant data structures have been removed, notably the SocketConnection. These are only needed temporarily during graph construction. For executing the compositor operations it is perfectly sufficient to store only the direct input link pointers. A common pointer indirection is avoided this way (which might also give a little performance improvement). == Avoid virtual recursive functions == Recursive virtual functions are evil. They are very hard to follow during debugging. At least in the parts this patch is concerned with these functions have been replaced by a non-virtual recursive core function (which might then call virtual non-recursive functions if needed). See for example NodeOperationBuilder::group_operations.
2014-04-15 16:06:12 +02:00
/*
* Copyright 2013, Blender Foundation.
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation; either version 2
* of the License, or (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software Foundation,
* Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
*
Cleanup: trailing whitespace (comment blocks) Strip unindented comment blocks - mainly headers to avoid conflicts.
2018-06-01 18:19:39 +02:00
* Contributor:
Structural cleanup and improvements for the compositor. Many parts of the compositor are unnecessarily complicated. This patch aims at reducing the complexity of writing nodes and making the code more transparent. == Separating Nodes and Operations == Currently these are both mixed in the same graph, even though they have very different purposes and are used at distinct stages in the compositing process. The patch introduces dedicated graph classes for nodes and for operations. This removes the need for a lot of special case checks (isOperation etc.) and explicit type casts. It simplifies the code since it becomes clear at every stage what type of node we are dealing with. The compiler can use static typing to avoid common bugs from mixing up these types and fewer runtime sanity checks are needed. == Simplified Node Conversion == Converting nodes to operations was previously based on "relinking", i.e. nodes would start with by mirroring links in the Blender DNA node trees, then add operations and redirect these links to them. This was very hard to follow in many cases and required a lot of attention to avoid invalid states. Now there is a helper class called the NodeConverter, which is passed to nodes and implements a much simpler API for this process. Nodes can add operations and explicit connections as before, but defining "external" links to the inputs/outputs of the original node now uses mapping instead of directly modifying link data. Input data (node graph) and result (operations graph) are cleanly separated. == Removed Redundant Data Structures == A few redundant data structures have been removed, notably the SocketConnection. These are only needed temporarily during graph construction. For executing the compositor operations it is perfectly sufficient to store only the direct input link pointers. A common pointer indirection is avoided this way (which might also give a little performance improvement). == Avoid virtual recursive functions == Recursive virtual functions are evil. They are very hard to follow during debugging. At least in the parts this patch is concerned with these functions have been replaced by a non-virtual recursive core function (which might then call virtual non-recursive functions if needed). See for example NodeOperationBuilder::group_operations.
2014-04-15 16:06:12 +02:00
* Lukas Toenne
*/
#ifndef _COM_NodeCompiler_h
#define _COM_NodeCompiler_h
Building with C++ guarded alloc works again
2014-04-16 02:22:49 +10:00
#ifdef WITH_CXX_GUARDEDALLOC
# include "MEM_guardedalloc.h"
#endif
Structural cleanup and improvements for the compositor. Many parts of the compositor are unnecessarily complicated. This patch aims at reducing the complexity of writing nodes and making the code more transparent. == Separating Nodes and Operations == Currently these are both mixed in the same graph, even though they have very different purposes and are used at distinct stages in the compositing process. The patch introduces dedicated graph classes for nodes and for operations. This removes the need for a lot of special case checks (isOperation etc.) and explicit type casts. It simplifies the code since it becomes clear at every stage what type of node we are dealing with. The compiler can use static typing to avoid common bugs from mixing up these types and fewer runtime sanity checks are needed. == Simplified Node Conversion == Converting nodes to operations was previously based on "relinking", i.e. nodes would start with by mirroring links in the Blender DNA node trees, then add operations and redirect these links to them. This was very hard to follow in many cases and required a lot of attention to avoid invalid states. Now there is a helper class called the NodeConverter, which is passed to nodes and implements a much simpler API for this process. Nodes can add operations and explicit connections as before, but defining "external" links to the inputs/outputs of the original node now uses mapping instead of directly modifying link data. Input data (node graph) and result (operations graph) are cleanly separated. == Removed Redundant Data Structures == A few redundant data structures have been removed, notably the SocketConnection. These are only needed temporarily during graph construction. For executing the compositor operations it is perfectly sufficient to store only the direct input link pointers. A common pointer indirection is avoided this way (which might also give a little performance improvement). == Avoid virtual recursive functions == Recursive virtual functions are evil. They are very hard to follow during debugging. At least in the parts this patch is concerned with these functions have been replaced by a non-virtual recursive core function (which might then call virtual non-recursive functions if needed). See for example NodeOperationBuilder::group_operations.
2014-04-15 16:06:12 +02:00
class NodeInput;
class NodeOutput;
class NodeOperation;
class NodeOperationInput;
class NodeOperationOutput;
class NodeOperationBuilder;
Fix T40986: crash on using the viewer node inside of group nodes. Viewers were activated both inside the active group as well as the top level tree (the latter being a quick fix for getting a fallback viewer). This caused a race condition on the shared viewer image. Now the active viewer is defined at node conversion time in the converter so that only one can be active at a time without each node having to follow complicated rules for exclusion.
2014-07-08 12:48:41 +02:00
class ViewerOperation;
Structural cleanup and improvements for the compositor. Many parts of the compositor are unnecessarily complicated. This patch aims at reducing the complexity of writing nodes and making the code more transparent. == Separating Nodes and Operations == Currently these are both mixed in the same graph, even though they have very different purposes and are used at distinct stages in the compositing process. The patch introduces dedicated graph classes for nodes and for operations. This removes the need for a lot of special case checks (isOperation etc.) and explicit type casts. It simplifies the code since it becomes clear at every stage what type of node we are dealing with. The compiler can use static typing to avoid common bugs from mixing up these types and fewer runtime sanity checks are needed. == Simplified Node Conversion == Converting nodes to operations was previously based on "relinking", i.e. nodes would start with by mirroring links in the Blender DNA node trees, then add operations and redirect these links to them. This was very hard to follow in many cases and required a lot of attention to avoid invalid states. Now there is a helper class called the NodeConverter, which is passed to nodes and implements a much simpler API for this process. Nodes can add operations and explicit connections as before, but defining "external" links to the inputs/outputs of the original node now uses mapping instead of directly modifying link data. Input data (node graph) and result (operations graph) are cleanly separated. == Removed Redundant Data Structures == A few redundant data structures have been removed, notably the SocketConnection. These are only needed temporarily during graph construction. For executing the compositor operations it is perfectly sufficient to store only the direct input link pointers. A common pointer indirection is avoided this way (which might also give a little performance improvement). == Avoid virtual recursive functions == Recursive virtual functions are evil. They are very hard to follow during debugging. At least in the parts this patch is concerned with these functions have been replaced by a non-virtual recursive core function (which might then call virtual non-recursive functions if needed). See for example NodeOperationBuilder::group_operations.
2014-04-15 16:06:12 +02:00
/** Interface type for converting a \a Node into \a NodeOperation.
* This is passed to \a Node::convertToOperation methods and allows them
* to register any number of operations, create links between them,
* and map original node sockets to their inputs or outputs.
*/
class NodeConverter {
public:
NodeConverter(NodeOperationBuilder *builder);
Cleanup: trailing space for compositor
2018-06-17 17:05:29 +02:00
Structural cleanup and improvements for the compositor. Many parts of the compositor are unnecessarily complicated. This patch aims at reducing the complexity of writing nodes and making the code more transparent. == Separating Nodes and Operations == Currently these are both mixed in the same graph, even though they have very different purposes and are used at distinct stages in the compositing process. The patch introduces dedicated graph classes for nodes and for operations. This removes the need for a lot of special case checks (isOperation etc.) and explicit type casts. It simplifies the code since it becomes clear at every stage what type of node we are dealing with. The compiler can use static typing to avoid common bugs from mixing up these types and fewer runtime sanity checks are needed. == Simplified Node Conversion == Converting nodes to operations was previously based on "relinking", i.e. nodes would start with by mirroring links in the Blender DNA node trees, then add operations and redirect these links to them. This was very hard to follow in many cases and required a lot of attention to avoid invalid states. Now there is a helper class called the NodeConverter, which is passed to nodes and implements a much simpler API for this process. Nodes can add operations and explicit connections as before, but defining "external" links to the inputs/outputs of the original node now uses mapping instead of directly modifying link data. Input data (node graph) and result (operations graph) are cleanly separated. == Removed Redundant Data Structures == A few redundant data structures have been removed, notably the SocketConnection. These are only needed temporarily during graph construction. For executing the compositor operations it is perfectly sufficient to store only the direct input link pointers. A common pointer indirection is avoided this way (which might also give a little performance improvement). == Avoid virtual recursive functions == Recursive virtual functions are evil. They are very hard to follow during debugging. At least in the parts this patch is concerned with these functions have been replaced by a non-virtual recursive core function (which might then call virtual non-recursive functions if needed). See for example NodeOperationBuilder::group_operations.
2014-04-15 16:06:12 +02:00
/** Insert a new operation into the operations graph.
* The operation must be created by the node.
*/
void addOperation(NodeOperation *operation);
Cleanup: trailing space for compositor
2018-06-17 17:05:29 +02:00
Structural cleanup and improvements for the compositor. Many parts of the compositor are unnecessarily complicated. This patch aims at reducing the complexity of writing nodes and making the code more transparent. == Separating Nodes and Operations == Currently these are both mixed in the same graph, even though they have very different purposes and are used at distinct stages in the compositing process. The patch introduces dedicated graph classes for nodes and for operations. This removes the need for a lot of special case checks (isOperation etc.) and explicit type casts. It simplifies the code since it becomes clear at every stage what type of node we are dealing with. The compiler can use static typing to avoid common bugs from mixing up these types and fewer runtime sanity checks are needed. == Simplified Node Conversion == Converting nodes to operations was previously based on "relinking", i.e. nodes would start with by mirroring links in the Blender DNA node trees, then add operations and redirect these links to them. This was very hard to follow in many cases and required a lot of attention to avoid invalid states. Now there is a helper class called the NodeConverter, which is passed to nodes and implements a much simpler API for this process. Nodes can add operations and explicit connections as before, but defining "external" links to the inputs/outputs of the original node now uses mapping instead of directly modifying link data. Input data (node graph) and result (operations graph) are cleanly separated. == Removed Redundant Data Structures == A few redundant data structures have been removed, notably the SocketConnection. These are only needed temporarily during graph construction. For executing the compositor operations it is perfectly sufficient to store only the direct input link pointers. A common pointer indirection is avoided this way (which might also give a little performance improvement). == Avoid virtual recursive functions == Recursive virtual functions are evil. They are very hard to follow during debugging. At least in the parts this patch is concerned with these functions have been replaced by a non-virtual recursive core function (which might then call virtual non-recursive functions if needed). See for example NodeOperationBuilder::group_operations.
2014-04-15 16:06:12 +02:00
/** Map input socket of the node to an operation socket.
* Links between nodes will then generate equivalent links between
* the mapped operation sockets.
Cleanup: trailing space for compositor
2018-06-17 17:05:29 +02:00
*
Structural cleanup and improvements for the compositor. Many parts of the compositor are unnecessarily complicated. This patch aims at reducing the complexity of writing nodes and making the code more transparent. == Separating Nodes and Operations == Currently these are both mixed in the same graph, even though they have very different purposes and are used at distinct stages in the compositing process. The patch introduces dedicated graph classes for nodes and for operations. This removes the need for a lot of special case checks (isOperation etc.) and explicit type casts. It simplifies the code since it becomes clear at every stage what type of node we are dealing with. The compiler can use static typing to avoid common bugs from mixing up these types and fewer runtime sanity checks are needed. == Simplified Node Conversion == Converting nodes to operations was previously based on "relinking", i.e. nodes would start with by mirroring links in the Blender DNA node trees, then add operations and redirect these links to them. This was very hard to follow in many cases and required a lot of attention to avoid invalid states. Now there is a helper class called the NodeConverter, which is passed to nodes and implements a much simpler API for this process. Nodes can add operations and explicit connections as before, but defining "external" links to the inputs/outputs of the original node now uses mapping instead of directly modifying link data. Input data (node graph) and result (operations graph) are cleanly separated. == Removed Redundant Data Structures == A few redundant data structures have been removed, notably the SocketConnection. These are only needed temporarily during graph construction. For executing the compositor operations it is perfectly sufficient to store only the direct input link pointers. A common pointer indirection is avoided this way (which might also give a little performance improvement). == Avoid virtual recursive functions == Recursive virtual functions are evil. They are very hard to follow during debugging. At least in the parts this patch is concerned with these functions have been replaced by a non-virtual recursive core function (which might then call virtual non-recursive functions if needed). See for example NodeOperationBuilder::group_operations.
2014-04-15 16:06:12 +02:00
* \note A \a Node input can be mapped to multiple \a NodeOperation inputs.
*/
void mapInputSocket(NodeInput *node_socket, NodeOperationInput *operation_socket);
/** Map output socket of the node to an operation socket.
* Links between nodes will then generate equivalent links between
* the mapped operation sockets.
Cleanup: trailing space for compositor
2018-06-17 17:05:29 +02:00
*
Structural cleanup and improvements for the compositor. Many parts of the compositor are unnecessarily complicated. This patch aims at reducing the complexity of writing nodes and making the code more transparent. == Separating Nodes and Operations == Currently these are both mixed in the same graph, even though they have very different purposes and are used at distinct stages in the compositing process. The patch introduces dedicated graph classes for nodes and for operations. This removes the need for a lot of special case checks (isOperation etc.) and explicit type casts. It simplifies the code since it becomes clear at every stage what type of node we are dealing with. The compiler can use static typing to avoid common bugs from mixing up these types and fewer runtime sanity checks are needed. == Simplified Node Conversion == Converting nodes to operations was previously based on "relinking", i.e. nodes would start with by mirroring links in the Blender DNA node trees, then add operations and redirect these links to them. This was very hard to follow in many cases and required a lot of attention to avoid invalid states. Now there is a helper class called the NodeConverter, which is passed to nodes and implements a much simpler API for this process. Nodes can add operations and explicit connections as before, but defining "external" links to the inputs/outputs of the original node now uses mapping instead of directly modifying link data. Input data (node graph) and result (operations graph) are cleanly separated. == Removed Redundant Data Structures == A few redundant data structures have been removed, notably the SocketConnection. These are only needed temporarily during graph construction. For executing the compositor operations it is perfectly sufficient to store only the direct input link pointers. A common pointer indirection is avoided this way (which might also give a little performance improvement). == Avoid virtual recursive functions == Recursive virtual functions are evil. They are very hard to follow during debugging. At least in the parts this patch is concerned with these functions have been replaced by a non-virtual recursive core function (which might then call virtual non-recursive functions if needed). See for example NodeOperationBuilder::group_operations.
2014-04-15 16:06:12 +02:00
* \note A \a Node output can only be mapped to one \a NodeOperation output.
* Any existing operation output mapping will be replaced.
*/
void mapOutputSocket(NodeOutput *node_socket, NodeOperationOutput *operation_socket);
Cleanup: trailing space for compositor
2018-06-17 17:05:29 +02:00
Structural cleanup and improvements for the compositor. Many parts of the compositor are unnecessarily complicated. This patch aims at reducing the complexity of writing nodes and making the code more transparent. == Separating Nodes and Operations == Currently these are both mixed in the same graph, even though they have very different purposes and are used at distinct stages in the compositing process. The patch introduces dedicated graph classes for nodes and for operations. This removes the need for a lot of special case checks (isOperation etc.) and explicit type casts. It simplifies the code since it becomes clear at every stage what type of node we are dealing with. The compiler can use static typing to avoid common bugs from mixing up these types and fewer runtime sanity checks are needed. == Simplified Node Conversion == Converting nodes to operations was previously based on "relinking", i.e. nodes would start with by mirroring links in the Blender DNA node trees, then add operations and redirect these links to them. This was very hard to follow in many cases and required a lot of attention to avoid invalid states. Now there is a helper class called the NodeConverter, which is passed to nodes and implements a much simpler API for this process. Nodes can add operations and explicit connections as before, but defining "external" links to the inputs/outputs of the original node now uses mapping instead of directly modifying link data. Input data (node graph) and result (operations graph) are cleanly separated. == Removed Redundant Data Structures == A few redundant data structures have been removed, notably the SocketConnection. These are only needed temporarily during graph construction. For executing the compositor operations it is perfectly sufficient to store only the direct input link pointers. A common pointer indirection is avoided this way (which might also give a little performance improvement). == Avoid virtual recursive functions == Recursive virtual functions are evil. They are very hard to follow during debugging. At least in the parts this patch is concerned with these functions have been replaced by a non-virtual recursive core function (which might then call virtual non-recursive functions if needed). See for example NodeOperationBuilder::group_operations.
2014-04-15 16:06:12 +02:00
/** Create a proxy operation for a node input.
* This operation will be removed later and replaced
* by direct links between the connected operations.
*/
Fix T41067: Muted nodes still do data type conversions. Proxy operations from muted nodes would still create conversion operations where the datatypes don't match, which creates unexpected behavior. Arguably datatype conversion could still happen even when the main operation is muted, but this would be a design change and so is disabled now.
2014-07-15 10:55:49 +02:00
NodeOperationOutput *addInputProxy(NodeInput *input, bool use_conversion);
Structural cleanup and improvements for the compositor. Many parts of the compositor are unnecessarily complicated. This patch aims at reducing the complexity of writing nodes and making the code more transparent. == Separating Nodes and Operations == Currently these are both mixed in the same graph, even though they have very different purposes and are used at distinct stages in the compositing process. The patch introduces dedicated graph classes for nodes and for operations. This removes the need for a lot of special case checks (isOperation etc.) and explicit type casts. It simplifies the code since it becomes clear at every stage what type of node we are dealing with. The compiler can use static typing to avoid common bugs from mixing up these types and fewer runtime sanity checks are needed. == Simplified Node Conversion == Converting nodes to operations was previously based on "relinking", i.e. nodes would start with by mirroring links in the Blender DNA node trees, then add operations and redirect these links to them. This was very hard to follow in many cases and required a lot of attention to avoid invalid states. Now there is a helper class called the NodeConverter, which is passed to nodes and implements a much simpler API for this process. Nodes can add operations and explicit connections as before, but defining "external" links to the inputs/outputs of the original node now uses mapping instead of directly modifying link data. Input data (node graph) and result (operations graph) are cleanly separated. == Removed Redundant Data Structures == A few redundant data structures have been removed, notably the SocketConnection. These are only needed temporarily during graph construction. For executing the compositor operations it is perfectly sufficient to store only the direct input link pointers. A common pointer indirection is avoided this way (which might also give a little performance improvement). == Avoid virtual recursive functions == Recursive virtual functions are evil. They are very hard to follow during debugging. At least in the parts this patch is concerned with these functions have been replaced by a non-virtual recursive core function (which might then call virtual non-recursive functions if needed). See for example NodeOperationBuilder::group_operations.
2014-04-15 16:06:12 +02:00
/** Create a proxy operation for a node output.
* This operation will be removed later and replaced
* by direct links between the connected operations.
*/
Fix T41067: Muted nodes still do data type conversions. Proxy operations from muted nodes would still create conversion operations where the datatypes don't match, which creates unexpected behavior. Arguably datatype conversion could still happen even when the main operation is muted, but this would be a design change and so is disabled now.
2014-07-15 10:55:49 +02:00
NodeOperationInput *addOutputProxy(NodeOutput *output, bool use_conversion);
Cleanup: trailing space for compositor
2018-06-17 17:05:29 +02:00
Fix T39799: Backdrop (compositor) ignores alpha. This issue is because of a somewhat "special" behavior in old code, which got lost during rB09874df: There was a variant of the `relinkConnections` function which would leave the socket completely unconnected. This is not a valid state really (given that each unconnected input must otherwise connected to a constant `Set` type node), but was used as a way to distinguish connected alpha/depth sockets in composite and viewer output nodes. https://developer.blender.org/diffusion/B/browse/master/source/blender/compositor/intern/COM_InputSocket.cpp;28a829893c702918afc5ac1945a06eaefa611594$69 After the large cleanup patch ({D309}) every socket is now automatically connected to a constant, such that `getInputSocketReader` will never return a NULL pointer. This breaks the previous test method, which needs to be replaced by more explicit flags. Luckily this was done only for very few output nodes (Composite, Viewer, Output-File). These now use the regular SetValueOperation default in case "use alpha" is disabled, but set this to an explicit 1.0 value instead of mapping to the node socket.
2014-04-25 12:00:35 +02:00
/** Define a constant input value. */
void addInputValue(NodeOperationInput *input, float value);
/** Define a constant input color. */
void addInputColor(NodeOperationInput *input, const float value[4]);
/** Define a constant input vector. */
void addInputVector(NodeOperationInput *input, const float value[3]);
Cleanup: trailing space for compositor
2018-06-17 17:05:29 +02:00
Structural cleanup and improvements for the compositor. Many parts of the compositor are unnecessarily complicated. This patch aims at reducing the complexity of writing nodes and making the code more transparent. == Separating Nodes and Operations == Currently these are both mixed in the same graph, even though they have very different purposes and are used at distinct stages in the compositing process. The patch introduces dedicated graph classes for nodes and for operations. This removes the need for a lot of special case checks (isOperation etc.) and explicit type casts. It simplifies the code since it becomes clear at every stage what type of node we are dealing with. The compiler can use static typing to avoid common bugs from mixing up these types and fewer runtime sanity checks are needed. == Simplified Node Conversion == Converting nodes to operations was previously based on "relinking", i.e. nodes would start with by mirroring links in the Blender DNA node trees, then add operations and redirect these links to them. This was very hard to follow in many cases and required a lot of attention to avoid invalid states. Now there is a helper class called the NodeConverter, which is passed to nodes and implements a much simpler API for this process. Nodes can add operations and explicit connections as before, but defining "external" links to the inputs/outputs of the original node now uses mapping instead of directly modifying link data. Input data (node graph) and result (operations graph) are cleanly separated. == Removed Redundant Data Structures == A few redundant data structures have been removed, notably the SocketConnection. These are only needed temporarily during graph construction. For executing the compositor operations it is perfectly sufficient to store only the direct input link pointers. A common pointer indirection is avoided this way (which might also give a little performance improvement). == Avoid virtual recursive functions == Recursive virtual functions are evil. They are very hard to follow during debugging. At least in the parts this patch is concerned with these functions have been replaced by a non-virtual recursive core function (which might then call virtual non-recursive functions if needed). See for example NodeOperationBuilder::group_operations.
2014-04-15 16:06:12 +02:00
/** Define a constant output value. */
void addOutputValue(NodeOutput *output, float value);
/** Define a constant output color. */
void addOutputColor(NodeOutput *output, const float value[4]);
/** Define a constant output vector. */
void addOutputVector(NodeOutput *output, const float value[3]);
Cleanup: trailing space for compositor
2018-06-17 17:05:29 +02:00
Structural cleanup and improvements for the compositor. Many parts of the compositor are unnecessarily complicated. This patch aims at reducing the complexity of writing nodes and making the code more transparent. == Separating Nodes and Operations == Currently these are both mixed in the same graph, even though they have very different purposes and are used at distinct stages in the compositing process. The patch introduces dedicated graph classes for nodes and for operations. This removes the need for a lot of special case checks (isOperation etc.) and explicit type casts. It simplifies the code since it becomes clear at every stage what type of node we are dealing with. The compiler can use static typing to avoid common bugs from mixing up these types and fewer runtime sanity checks are needed. == Simplified Node Conversion == Converting nodes to operations was previously based on "relinking", i.e. nodes would start with by mirroring links in the Blender DNA node trees, then add operations and redirect these links to them. This was very hard to follow in many cases and required a lot of attention to avoid invalid states. Now there is a helper class called the NodeConverter, which is passed to nodes and implements a much simpler API for this process. Nodes can add operations and explicit connections as before, but defining "external" links to the inputs/outputs of the original node now uses mapping instead of directly modifying link data. Input data (node graph) and result (operations graph) are cleanly separated. == Removed Redundant Data Structures == A few redundant data structures have been removed, notably the SocketConnection. These are only needed temporarily during graph construction. For executing the compositor operations it is perfectly sufficient to store only the direct input link pointers. A common pointer indirection is avoided this way (which might also give a little performance improvement). == Avoid virtual recursive functions == Recursive virtual functions are evil. They are very hard to follow during debugging. At least in the parts this patch is concerned with these functions have been replaced by a non-virtual recursive core function (which might then call virtual non-recursive functions if needed). See for example NodeOperationBuilder::group_operations.
2014-04-15 16:06:12 +02:00
/** Add an explicit link between two operations. */
void addLink(NodeOperationOutput *from, NodeOperationInput *to);
Cleanup: trailing space for compositor
2018-06-17 17:05:29 +02:00
Structural cleanup and improvements for the compositor. Many parts of the compositor are unnecessarily complicated. This patch aims at reducing the complexity of writing nodes and making the code more transparent. == Separating Nodes and Operations == Currently these are both mixed in the same graph, even though they have very different purposes and are used at distinct stages in the compositing process. The patch introduces dedicated graph classes for nodes and for operations. This removes the need for a lot of special case checks (isOperation etc.) and explicit type casts. It simplifies the code since it becomes clear at every stage what type of node we are dealing with. The compiler can use static typing to avoid common bugs from mixing up these types and fewer runtime sanity checks are needed. == Simplified Node Conversion == Converting nodes to operations was previously based on "relinking", i.e. nodes would start with by mirroring links in the Blender DNA node trees, then add operations and redirect these links to them. This was very hard to follow in many cases and required a lot of attention to avoid invalid states. Now there is a helper class called the NodeConverter, which is passed to nodes and implements a much simpler API for this process. Nodes can add operations and explicit connections as before, but defining "external" links to the inputs/outputs of the original node now uses mapping instead of directly modifying link data. Input data (node graph) and result (operations graph) are cleanly separated. == Removed Redundant Data Structures == A few redundant data structures have been removed, notably the SocketConnection. These are only needed temporarily during graph construction. For executing the compositor operations it is perfectly sufficient to store only the direct input link pointers. A common pointer indirection is avoided this way (which might also give a little performance improvement). == Avoid virtual recursive functions == Recursive virtual functions are evil. They are very hard to follow during debugging. At least in the parts this patch is concerned with these functions have been replaced by a non-virtual recursive core function (which might then call virtual non-recursive functions if needed). See for example NodeOperationBuilder::group_operations.
2014-04-15 16:06:12 +02:00
/** Add a preview operation for a operation output. */
void addPreview(NodeOperationOutput *output);
/** Add a preview operation for a node input. */
void addNodeInputPreview(NodeInput *input);
Cleanup: trailing space for compositor
2018-06-17 17:05:29 +02:00
Structural cleanup and improvements for the compositor. Many parts of the compositor are unnecessarily complicated. This patch aims at reducing the complexity of writing nodes and making the code more transparent. == Separating Nodes and Operations == Currently these are both mixed in the same graph, even though they have very different purposes and are used at distinct stages in the compositing process. The patch introduces dedicated graph classes for nodes and for operations. This removes the need for a lot of special case checks (isOperation etc.) and explicit type casts. It simplifies the code since it becomes clear at every stage what type of node we are dealing with. The compiler can use static typing to avoid common bugs from mixing up these types and fewer runtime sanity checks are needed. == Simplified Node Conversion == Converting nodes to operations was previously based on "relinking", i.e. nodes would start with by mirroring links in the Blender DNA node trees, then add operations and redirect these links to them. This was very hard to follow in many cases and required a lot of attention to avoid invalid states. Now there is a helper class called the NodeConverter, which is passed to nodes and implements a much simpler API for this process. Nodes can add operations and explicit connections as before, but defining "external" links to the inputs/outputs of the original node now uses mapping instead of directly modifying link data. Input data (node graph) and result (operations graph) are cleanly separated. == Removed Redundant Data Structures == A few redundant data structures have been removed, notably the SocketConnection. These are only needed temporarily during graph construction. For executing the compositor operations it is perfectly sufficient to store only the direct input link pointers. A common pointer indirection is avoided this way (which might also give a little performance improvement). == Avoid virtual recursive functions == Recursive virtual functions are evil. They are very hard to follow during debugging. At least in the parts this patch is concerned with these functions have been replaced by a non-virtual recursive core function (which might then call virtual non-recursive functions if needed). See for example NodeOperationBuilder::group_operations.
2014-04-15 16:06:12 +02:00
/** When a node has no valid data
* @note missing image / group pointer, or missing renderlayer from EXR
*/
NodeOperation *setInvalidOutput(NodeOutput *output);
Cleanup: trailing space for compositor
2018-06-17 17:05:29 +02:00
Fix T40986: crash on using the viewer node inside of group nodes. Viewers were activated both inside the active group as well as the top level tree (the latter being a quick fix for getting a fallback viewer). This caused a race condition on the shared viewer image. Now the active viewer is defined at node conversion time in the converter so that only one can be active at a time without each node having to follow complicated rules for exclusion.
2014-07-08 12:48:41 +02:00
/** Define a viewer operation as the active output, if possible */
void registerViewer(ViewerOperation *viewer);
/** The currently active viewer output operation */
ViewerOperation *active_viewer() const;
Cleanup: trailing space for compositor
2018-06-17 17:05:29 +02:00
Structural cleanup and improvements for the compositor. Many parts of the compositor are unnecessarily complicated. This patch aims at reducing the complexity of writing nodes and making the code more transparent. == Separating Nodes and Operations == Currently these are both mixed in the same graph, even though they have very different purposes and are used at distinct stages in the compositing process. The patch introduces dedicated graph classes for nodes and for operations. This removes the need for a lot of special case checks (isOperation etc.) and explicit type casts. It simplifies the code since it becomes clear at every stage what type of node we are dealing with. The compiler can use static typing to avoid common bugs from mixing up these types and fewer runtime sanity checks are needed. == Simplified Node Conversion == Converting nodes to operations was previously based on "relinking", i.e. nodes would start with by mirroring links in the Blender DNA node trees, then add operations and redirect these links to them. This was very hard to follow in many cases and required a lot of attention to avoid invalid states. Now there is a helper class called the NodeConverter, which is passed to nodes and implements a much simpler API for this process. Nodes can add operations and explicit connections as before, but defining "external" links to the inputs/outputs of the original node now uses mapping instead of directly modifying link data. Input data (node graph) and result (operations graph) are cleanly separated. == Removed Redundant Data Structures == A few redundant data structures have been removed, notably the SocketConnection. These are only needed temporarily during graph construction. For executing the compositor operations it is perfectly sufficient to store only the direct input link pointers. A common pointer indirection is avoided this way (which might also give a little performance improvement). == Avoid virtual recursive functions == Recursive virtual functions are evil. They are very hard to follow during debugging. At least in the parts this patch is concerned with these functions have been replaced by a non-virtual recursive core function (which might then call virtual non-recursive functions if needed). See for example NodeOperationBuilder::group_operations.
2014-04-15 16:06:12 +02:00
private:
/** The internal builder for storing the results of the graph construction. */
NodeOperationBuilder *m_builder;
#ifdef WITH_CXX_GUARDEDALLOC
MEM_CXX_CLASS_ALLOC_FUNCS("COM:NodeCompiler")
#endif
};
#endif /* _COM_NodeCompiler_h */
Copy Permalink