Files
test/source/blender/nodes/intern/node_tree_ref.cc

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

679 lines
22 KiB
C++
Raw Normal View History

/* SPDX-License-Identifier: GPL-2.0-or-later */
#include <mutex>
#include "NOD_node_tree_ref.hh"
#include "BLI_dot_export.hh"
#include "BLI_stack.hh"
#include "RNA_prototypes.h"
namespace blender::nodes {
NodeTreeRef::NodeTreeRef(bNodeTree *btree) : btree_(btree)
{
Map<bNode *, NodeRef *> node_mapping;
LISTBASE_FOREACH (bNode *, bnode, &btree->nodes) {
2021-03-07 14:46:48 +01:00
NodeRef &node = *allocator_.construct<NodeRef>().release();
node.tree_ = this;
node.bnode_ = bnode;
node.id_ = nodes_by_id_.append_and_get_index(&node);
LISTBASE_FOREACH (bNodeSocket *, bsocket, &bnode->inputs) {
2021-03-07 14:46:48 +01:00
InputSocketRef &socket = *allocator_.construct<InputSocketRef>().release();
socket.node_ = &node;
socket.index_ = node.inputs_.append_and_get_index(&socket);
socket.is_input_ = true;
socket.bsocket_ = bsocket;
socket.id_ = sockets_by_id_.append_and_get_index(&socket);
}
LISTBASE_FOREACH (bNodeSocket *, bsocket, &bnode->outputs) {
2021-03-07 14:46:48 +01:00
OutputSocketRef &socket = *allocator_.construct<OutputSocketRef>().release();
socket.node_ = &node;
socket.index_ = node.outputs_.append_and_get_index(&socket);
socket.is_input_ = false;
socket.bsocket_ = bsocket;
socket.id_ = sockets_by_id_.append_and_get_index(&socket);
}
LISTBASE_FOREACH (bNodeLink *, blink, &bnode->internal_links) {
2021-03-07 14:46:48 +01:00
InternalLinkRef &internal_link = *allocator_.construct<InternalLinkRef>().release();
internal_link.blink_ = blink;
for (InputSocketRef *socket_ref : node.inputs_) {
if (socket_ref->bsocket_ == blink->fromsock) {
internal_link.from_ = socket_ref;
break;
}
}
for (OutputSocketRef *socket_ref : node.outputs_) {
if (socket_ref->bsocket_ == blink->tosock) {
internal_link.to_ = socket_ref;
break;
}
}
Nodes: refactor node tree update handling Goals of this refactor: * More unified approach to updating everything that needs to be updated after a change in a node tree. * The updates should happen in the correct order and quadratic or worse algorithms should be avoided. * Improve detection of changes to the output to avoid tagging the depsgraph when it's not necessary. * Move towards a more declarative style of defining nodes by having a more centralized update procedure. The refactor consists of two main parts: * Node tree tagging and update refactor. * Generally, when changes are done to a node tree, it is tagged dirty until a global update function is called that updates everything in the correct order. * The tagging is more fine-grained compared to before, to allow for more precise depsgraph update tagging. * Depsgraph changes. * The shading specific depsgraph node for node trees as been removed. * Instead, there is a new `NTREE_OUTPUT` depsgrap node, which is only tagged when the output of the node tree changed (e.g. the Group Output or Material Output node). * The copy-on-write relation from node trees to the data block they are embedded in is now non-flushing. This avoids e.g. triggering a material update after the shader node tree changed in unrelated ways. Instead the material has a flushing relation to the new `NTREE_OUTPUT` node now. * The depsgraph no longer reports data block changes through to cycles through `Depsgraph.updates` when only the node tree changed in ways that do not affect the output. Avoiding unnecessary updates seems to work well for geometry nodes and cycles. The situation is a bit worse when there are drivers on the node tree, but that could potentially be improved separately in the future. Avoiding updates in eevee and the compositor is more tricky, but also less urgent. * Eevee updates are triggered by calling `DRW_notify_view_update` in `ED_render_view3d_update` indirectly from `DEG_editors_update`. * Compositor updates are triggered by `ED_node_composite_job` in `node_area_refresh`. This is triggered by calling `ED_area_tag_refresh` in `node_area_listener`. Removing updates always has the risk of breaking some dependency that no one was aware of. It's not unlikely that this will happen here as well. Adding back missing updates should be quite a bit easier than getting rid of unnecessary updates though. Differential Revision: https://developer.blender.org/D13246
2021-12-21 15:18:56 +01:00
BLI_assert(internal_link.from_ != nullptr);
BLI_assert(internal_link.to_ != nullptr);
node.internal_links_.append(&internal_link);
}
input_sockets_.extend(node.inputs_.as_span());
output_sockets_.extend(node.outputs_.as_span());
node_mapping.add_new(bnode, &node);
}
LISTBASE_FOREACH (bNodeLink *, blink, &btree->links) {
OutputSocketRef &from_socket = this->find_output_socket(
node_mapping, blink->fromnode, blink->fromsock);
InputSocketRef &to_socket = this->find_input_socket(
node_mapping, blink->tonode, blink->tosock);
2021-03-07 14:46:48 +01:00
LinkRef &link = *allocator_.construct<LinkRef>().release();
link.from_ = &from_socket;
link.to_ = &to_socket;
link.blink_ = blink;
links_.append(&link);
from_socket.directly_linked_links_.append(&link);
to_socket.directly_linked_links_.append(&link);
}
for (InputSocketRef *input_socket : input_sockets_) {
if (input_socket->is_multi_input_socket()) {
std::sort(input_socket->directly_linked_links_.begin(),
input_socket->directly_linked_links_.end(),
[&](const LinkRef *a, const LinkRef *b) -> bool {
int index_a = a->blink()->multi_input_socket_index;
int index_b = b->blink()->multi_input_socket_index;
return index_a > index_b;
});
}
}
this->create_socket_identifier_maps();
this->create_linked_socket_caches();
for (NodeRef *node : nodes_by_id_) {
const bNodeType *nodetype = node->bnode_->typeinfo;
nodes_by_type_.add(nodetype, node);
}
const Span<const NodeRef *> group_output_nodes = this->nodes_by_type("NodeGroupOutput");
if (group_output_nodes.is_empty()) {
group_output_node_ = nullptr;
}
else if (group_output_nodes.size() == 1) {
group_output_node_ = group_output_nodes.first();
}
else {
for (const NodeRef *group_output : group_output_nodes) {
if (group_output->bnode_->flag & NODE_DO_OUTPUT) {
group_output_node_ = group_output;
break;
}
}
}
}
NodeTreeRef::~NodeTreeRef()
{
/* The destructor has to be called manually, because these types are allocated in a linear
* allocator. */
for (NodeRef *node : nodes_by_id_) {
node->~NodeRef();
}
for (InputSocketRef *socket : input_sockets_) {
socket->~InputSocketRef();
}
for (OutputSocketRef *socket : output_sockets_) {
socket->~OutputSocketRef();
}
for (LinkRef *link : links_) {
link->~LinkRef();
}
}
InputSocketRef &NodeTreeRef::find_input_socket(Map<bNode *, NodeRef *> &node_mapping,
bNode *bnode,
bNodeSocket *bsocket)
{
NodeRef *node = node_mapping.lookup(bnode);
for (InputSocketRef *socket : node->inputs_) {
if (socket->bsocket_ == bsocket) {
return *socket;
}
}
BLI_assert_unreachable();
return *node->inputs_[0];
}
OutputSocketRef &NodeTreeRef::find_output_socket(Map<bNode *, NodeRef *> &node_mapping,
bNode *bnode,
bNodeSocket *bsocket)
{
NodeRef *node = node_mapping.lookup(bnode);
for (OutputSocketRef *socket : node->outputs_) {
if (socket->bsocket_ == bsocket) {
return *socket;
}
}
BLI_assert_unreachable();
return *node->outputs_[0];
}
void NodeTreeRef::create_linked_socket_caches()
{
for (InputSocketRef *socket : input_sockets_) {
/* Find directly linked socket based on incident links. */
Vector<const SocketRef *> directly_linked_sockets;
for (LinkRef *link : socket->directly_linked_links_) {
directly_linked_sockets.append(link->from_);
}
socket->directly_linked_sockets_ = allocator_.construct_array_copy(
directly_linked_sockets.as_span());
/* Find logically linked sockets. */
Vector<const SocketRef *> logically_linked_sockets;
Vector<const SocketRef *> logically_linked_skipped_sockets;
Vector<const InputSocketRef *> seen_sockets_stack;
socket->foreach_logical_origin(
[&](const OutputSocketRef &origin) { logically_linked_sockets.append(&origin); },
[&](const SocketRef &socket) { logically_linked_skipped_sockets.append(&socket); },
false,
seen_sockets_stack);
if (logically_linked_sockets == directly_linked_sockets) {
socket->logically_linked_sockets_ = socket->directly_linked_sockets_;
}
else {
socket->logically_linked_sockets_ = allocator_.construct_array_copy(
logically_linked_sockets.as_span());
}
socket->logically_linked_skipped_sockets_ = allocator_.construct_array_copy(
logically_linked_skipped_sockets.as_span());
}
for (OutputSocketRef *socket : output_sockets_) {
/* Find directly linked socket based on incident links. */
Vector<const SocketRef *> directly_linked_sockets;
for (LinkRef *link : socket->directly_linked_links_) {
directly_linked_sockets.append(link->to_);
}
socket->directly_linked_sockets_ = allocator_.construct_array_copy(
directly_linked_sockets.as_span());
/* Find logically linked sockets. */
Vector<const SocketRef *> logically_linked_sockets;
Vector<const SocketRef *> logically_linked_skipped_sockets;
Vector<const OutputSocketRef *> handled_sockets;
socket->foreach_logical_target(
[&](const InputSocketRef &target) { logically_linked_sockets.append(&target); },
[&](const SocketRef &socket) { logically_linked_skipped_sockets.append(&socket); },
handled_sockets);
if (logically_linked_sockets == directly_linked_sockets) {
socket->logically_linked_sockets_ = socket->directly_linked_sockets_;
}
else {
socket->logically_linked_sockets_ = allocator_.construct_array_copy(
logically_linked_sockets.as_span());
}
socket->logically_linked_skipped_sockets_ = allocator_.construct_array_copy(
logically_linked_skipped_sockets.as_span());
}
}
void InputSocketRef::foreach_logical_origin(
FunctionRef<void(const OutputSocketRef &)> origin_fn,
FunctionRef<void(const SocketRef &)> skipped_fn,
bool only_follow_first_input_link,
Vector<const InputSocketRef *> &seen_sockets_stack) const
{
/* Protect against loops. */
if (seen_sockets_stack.contains(this)) {
return;
}
seen_sockets_stack.append(this);
Span<const LinkRef *> links_to_check = this->directly_linked_links();
if (only_follow_first_input_link) {
links_to_check = links_to_check.take_front(1);
}
for (const LinkRef *link : links_to_check) {
if (link->is_muted()) {
continue;
}
const OutputSocketRef &origin = link->from();
const NodeRef &origin_node = origin.node();
if (!origin.is_available()) {
/* Non available sockets are ignored. */
}
else if (origin_node.is_reroute_node()) {
const InputSocketRef &reroute_input = origin_node.input(0);
const OutputSocketRef &reroute_output = origin_node.output(0);
skipped_fn.call_safe(reroute_input);
skipped_fn.call_safe(reroute_output);
reroute_input.foreach_logical_origin(origin_fn, skipped_fn, false, seen_sockets_stack);
}
else if (origin_node.is_muted()) {
for (const InternalLinkRef *internal_link : origin_node.internal_links()) {
if (&internal_link->to() == &origin) {
const InputSocketRef &mute_input = internal_link->from();
skipped_fn.call_safe(origin);
skipped_fn.call_safe(mute_input);
mute_input.foreach_logical_origin(origin_fn, skipped_fn, true, seen_sockets_stack);
}
}
}
else {
origin_fn(origin);
}
}
seen_sockets_stack.pop_last();
}
void OutputSocketRef::foreach_logical_target(
FunctionRef<void(const InputSocketRef &)> target_fn,
FunctionRef<void(const SocketRef &)> skipped_fn,
Vector<const OutputSocketRef *> &seen_sockets_stack) const
{
/* Protect against loops. */
if (seen_sockets_stack.contains(this)) {
return;
}
seen_sockets_stack.append(this);
for (const LinkRef *link : this->directly_linked_links()) {
if (link->is_muted()) {
continue;
}
const InputSocketRef &target = link->to();
const NodeRef &target_node = target.node();
if (!target.is_available()) {
/* Non available sockets are ignored. */
}
else if (target_node.is_reroute_node()) {
const OutputSocketRef &reroute_output = target_node.output(0);
skipped_fn.call_safe(target);
skipped_fn.call_safe(reroute_output);
reroute_output.foreach_logical_target(target_fn, skipped_fn, seen_sockets_stack);
}
else if (target_node.is_muted()) {
skipped_fn.call_safe(target);
for (const InternalLinkRef *internal_link : target_node.internal_links()) {
if (&internal_link->from() == &target) {
/* The internal link only forwards the first incoming link. */
if (target.is_multi_input_socket()) {
if (target.directly_linked_links()[0] != link) {
continue;
}
}
const OutputSocketRef &mute_output = internal_link->to();
skipped_fn.call_safe(target);
skipped_fn.call_safe(mute_output);
mute_output.foreach_logical_target(target_fn, skipped_fn, seen_sockets_stack);
}
}
}
else {
target_fn(target);
}
}
seen_sockets_stack.pop_last();
}
namespace {
struct SocketByIdentifierMap {
SocketIndexByIdentifierMap *map = nullptr;
std::unique_ptr<SocketIndexByIdentifierMap> owned_map;
};
} // namespace
static std::unique_ptr<SocketIndexByIdentifierMap> create_identifier_map(const ListBase &sockets)
{
std::unique_ptr<SocketIndexByIdentifierMap> map = std::make_unique<SocketIndexByIdentifierMap>();
int index;
LISTBASE_FOREACH_INDEX (bNodeSocket *, socket, &sockets, index) {
map->add_new(socket->identifier, index);
}
return map;
}
/* This function is not threadsafe. */
static SocketByIdentifierMap get_or_create_identifier_map(
const bNode &node, const ListBase &sockets, const bNodeSocketTemplate *sockets_template)
{
SocketByIdentifierMap map;
if (sockets_template == nullptr) {
if (BLI_listbase_is_empty(&sockets)) {
static SocketIndexByIdentifierMap empty_map;
map.map = &empty_map;
}
else if (node.type == NODE_REROUTE) {
if (&node.inputs == &sockets) {
static SocketIndexByIdentifierMap reroute_input_map = [] {
SocketIndexByIdentifierMap map;
map.add_new("Input", 0);
return map;
}();
map.map = &reroute_input_map;
}
else {
static SocketIndexByIdentifierMap reroute_output_map = [] {
SocketIndexByIdentifierMap map;
map.add_new("Output", 0);
return map;
}();
map.map = &reroute_output_map;
}
}
else {
/* The node has a dynamic amount of sockets. Therefore we need to create a new map. */
map.owned_map = create_identifier_map(sockets);
map.map = &*map.owned_map;
}
}
else {
/* Cache only one map for nodes that have the same sockets. */
static Map<const bNodeSocketTemplate *, std::unique_ptr<SocketIndexByIdentifierMap>> maps;
map.map = &*maps.lookup_or_add_cb(sockets_template,
[&]() { return create_identifier_map(sockets); });
}
return map;
}
void NodeTreeRef::create_socket_identifier_maps()
{
/* `get_or_create_identifier_map` is not threadsafe, therefore we have to hold a lock here. */
static std::mutex mutex;
std::lock_guard lock{mutex};
for (NodeRef *node : nodes_by_id_) {
bNode &bnode = *node->bnode_;
SocketByIdentifierMap inputs_map = get_or_create_identifier_map(
bnode, bnode.inputs, bnode.typeinfo->inputs);
SocketByIdentifierMap outputs_map = get_or_create_identifier_map(
bnode, bnode.outputs, bnode.typeinfo->outputs);
node->input_index_by_identifier_ = inputs_map.map;
node->output_index_by_identifier_ = outputs_map.map;
if (inputs_map.owned_map) {
owned_identifier_maps_.append(std::move(inputs_map.owned_map));
}
if (outputs_map.owned_map) {
owned_identifier_maps_.append(std::move(outputs_map.owned_map));
}
}
}
Geometry Nodes: initial scattering and geometry nodes This is the initial merge from the geometry-nodes branch. Nodes: * Attribute Math * Boolean * Edge Split * Float Compare * Object Info * Point Distribute * Point Instance * Random Attribute * Random Float * Subdivision Surface * Transform * Triangulate It includes the initial evaluation of geometry node groups in the Geometry Nodes modifier. Notes on the Generic attribute access API The API adds an indirection for attribute access. That has the following benefits: * Most code does not have to care about how an attribute is stored internally. This is mainly necessary, because we have to deal with "legacy" attributes such as vertex weights and attributes that are embedded into other structs such as vertex positions. * When reading from an attribute, we generally don't care what domain the attribute is stored on. So we want to abstract away the interpolation that that adapts attributes from one domain to another domain (this is not actually implemented yet). Other possible improvements for later iterations include: * Actually implement interpolation between domains. * Don't use inheritance for the different attribute types. A single class for read access and one for write access might be enough, because we know all the ways in which attributes are stored internally. We don't want more different internal structures in the future. On the contrary, ideally we can consolidate the different storage formats in the future to reduce the need for this indirection. * Remove the need for heap allocations when creating attribute accessors. It includes commits from: * Dalai Felinto * Hans Goudey * Jacques Lucke * Léo Depoix
2020-12-02 13:25:25 +01:00
static bool has_link_cycles_recursive(const NodeRef &node,
MutableSpan<bool> visited,
MutableSpan<bool> is_in_stack)
{
const int node_id = node.id();
if (is_in_stack[node_id]) {
return true;
}
if (visited[node_id]) {
return false;
}
visited[node_id] = true;
is_in_stack[node_id] = true;
for (const OutputSocketRef *from_socket : node.outputs()) {
if (!from_socket->is_available()) {
continue;
}
Geometry Nodes: initial scattering and geometry nodes This is the initial merge from the geometry-nodes branch. Nodes: * Attribute Math * Boolean * Edge Split * Float Compare * Object Info * Point Distribute * Point Instance * Random Attribute * Random Float * Subdivision Surface * Transform * Triangulate It includes the initial evaluation of geometry node groups in the Geometry Nodes modifier. Notes on the Generic attribute access API The API adds an indirection for attribute access. That has the following benefits: * Most code does not have to care about how an attribute is stored internally. This is mainly necessary, because we have to deal with "legacy" attributes such as vertex weights and attributes that are embedded into other structs such as vertex positions. * When reading from an attribute, we generally don't care what domain the attribute is stored on. So we want to abstract away the interpolation that that adapts attributes from one domain to another domain (this is not actually implemented yet). Other possible improvements for later iterations include: * Actually implement interpolation between domains. * Don't use inheritance for the different attribute types. A single class for read access and one for write access might be enough, because we know all the ways in which attributes are stored internally. We don't want more different internal structures in the future. On the contrary, ideally we can consolidate the different storage formats in the future to reduce the need for this indirection. * Remove the need for heap allocations when creating attribute accessors. It includes commits from: * Dalai Felinto * Hans Goudey * Jacques Lucke * Léo Depoix
2020-12-02 13:25:25 +01:00
for (const InputSocketRef *to_socket : from_socket->directly_linked_sockets()) {
if (!to_socket->is_available()) {
continue;
}
Geometry Nodes: initial scattering and geometry nodes This is the initial merge from the geometry-nodes branch. Nodes: * Attribute Math * Boolean * Edge Split * Float Compare * Object Info * Point Distribute * Point Instance * Random Attribute * Random Float * Subdivision Surface * Transform * Triangulate It includes the initial evaluation of geometry node groups in the Geometry Nodes modifier. Notes on the Generic attribute access API The API adds an indirection for attribute access. That has the following benefits: * Most code does not have to care about how an attribute is stored internally. This is mainly necessary, because we have to deal with "legacy" attributes such as vertex weights and attributes that are embedded into other structs such as vertex positions. * When reading from an attribute, we generally don't care what domain the attribute is stored on. So we want to abstract away the interpolation that that adapts attributes from one domain to another domain (this is not actually implemented yet). Other possible improvements for later iterations include: * Actually implement interpolation between domains. * Don't use inheritance for the different attribute types. A single class for read access and one for write access might be enough, because we know all the ways in which attributes are stored internally. We don't want more different internal structures in the future. On the contrary, ideally we can consolidate the different storage formats in the future to reduce the need for this indirection. * Remove the need for heap allocations when creating attribute accessors. It includes commits from: * Dalai Felinto * Hans Goudey * Jacques Lucke * Léo Depoix
2020-12-02 13:25:25 +01:00
const NodeRef &to_node = to_socket->node();
if (has_link_cycles_recursive(to_node, visited, is_in_stack)) {
return true;
}
}
}
is_in_stack[node_id] = false;
return false;
}
bool NodeTreeRef::has_link_cycles() const
{
const int node_amount = nodes_by_id_.size();
Array<bool> visited(node_amount, false);
Array<bool> is_in_stack(node_amount, false);
for (const NodeRef *node : nodes_by_id_) {
if (has_link_cycles_recursive(*node, visited, is_in_stack)) {
return true;
}
}
return false;
}
bool NodeTreeRef::has_undefined_nodes_or_sockets() const
{
for (const NodeRef *node : nodes_by_id_) {
if (node->is_undefined()) {
return true;
}
}
for (const SocketRef *socket : sockets_by_id_) {
if (socket->is_undefined()) {
return true;
}
}
return false;
}
bool NodeRef::any_input_is_directly_linked() const
{
for (const SocketRef *socket : inputs_) {
if (!socket->directly_linked_sockets().is_empty()) {
return true;
}
}
return false;
}
bool NodeRef::any_output_is_directly_linked() const
{
for (const SocketRef *socket : outputs_) {
if (!socket->directly_linked_sockets().is_empty()) {
return true;
}
}
return false;
}
bool NodeRef::any_socket_is_directly_linked(eNodeSocketInOut in_out) const
{
if (in_out == SOCK_IN) {
return this->any_input_is_directly_linked();
}
return this->any_output_is_directly_linked();
}
struct ToposortNodeState {
bool is_done = false;
bool is_in_stack = false;
};
static void toposort_from_start_node(const NodeTreeRef::ToposortDirection direction,
const NodeRef &start_node,
MutableSpan<ToposortNodeState> node_states,
NodeTreeRef::ToposortResult &result)
{
struct Item {
const NodeRef *node;
/* Index of the next socket that is checked in the depth-first search. */
int socket_index = 0;
/* Link index in the next socket that is checked in the depth-first search. */
int link_index = 0;
};
/* Do a depth-first search to sort nodes topologically. */
Stack<Item, 64> nodes_to_check;
nodes_to_check.push({&start_node});
while (!nodes_to_check.is_empty()) {
Item &item = nodes_to_check.peek();
const NodeRef &node = *item.node;
const Span<const SocketRef *> sockets = node.sockets(
direction == NodeTreeRef::ToposortDirection::LeftToRight ? SOCK_IN : SOCK_OUT);
while (true) {
if (item.socket_index == sockets.size()) {
/* All sockets have already been visited. */
break;
}
const SocketRef &socket = *sockets[item.socket_index];
const Span<const SocketRef *> linked_sockets = socket.directly_linked_sockets();
if (item.link_index == linked_sockets.size()) {
/* All links connected to this socket have already been visited. */
item.socket_index++;
item.link_index = 0;
continue;
}
const SocketRef &linked_socket = *linked_sockets[item.link_index];
const NodeRef &linked_node = linked_socket.node();
ToposortNodeState &linked_node_state = node_states[linked_node.id()];
if (linked_node_state.is_done) {
/* The linked node has already been visited. */
item.link_index++;
continue;
}
if (linked_node_state.is_in_stack) {
result.has_cycle = true;
}
else {
nodes_to_check.push({&linked_node});
linked_node_state.is_in_stack = true;
}
break;
}
/* If no other element has been pushed, the current node can be pushed to the sorted list. */
if (&item == &nodes_to_check.peek()) {
ToposortNodeState &node_state = node_states[node.id()];
node_state.is_done = true;
node_state.is_in_stack = false;
result.sorted_nodes.append(&node);
nodes_to_check.pop();
}
}
}
NodeTreeRef::ToposortResult NodeTreeRef::toposort(const ToposortDirection direction) const
{
ToposortResult result;
result.sorted_nodes.reserve(nodes_by_id_.size());
Array<ToposortNodeState> node_states(nodes_by_id_.size());
for (const NodeRef *node : nodes_by_id_) {
if (node_states[node->id()].is_done) {
/* Ignore nodes that are done already. */
continue;
}
if (node->any_socket_is_directly_linked(
direction == ToposortDirection::LeftToRight ? SOCK_OUT : SOCK_IN)) {
/* Ignore non-start nodes. */
continue;
}
toposort_from_start_node(direction, *node, node_states, result);
}
/* Check if the loop above forgot some nodes because there is a cycle. */
if (result.sorted_nodes.size() < nodes_by_id_.size()) {
result.has_cycle = true;
for (const NodeRef *node : nodes_by_id_) {
if (node_states[node->id()].is_done) {
/* Ignore nodes that are done already. */
continue;
}
/* Start toposort at this node which is somewhere in the middle of a loop. */
toposort_from_start_node(direction, *node, node_states, result);
}
}
BLI_assert(result.sorted_nodes.size() == nodes_by_id_.size());
return result;
}
const NodeRef *NodeTreeRef::find_node(const bNode &bnode) const
{
for (const NodeRef *node : this->nodes_by_type(bnode.typeinfo)) {
if (node->bnode_ == &bnode) {
return node;
}
}
return nullptr;
}
std::string NodeTreeRef::to_dot() const
{
dot::DirectedGraph digraph;
digraph.set_rankdir(dot::Attr_rankdir::LeftToRight);
Map<const NodeRef *, dot::NodeWithSocketsRef> dot_nodes;
for (const NodeRef *node : nodes_by_id_) {
dot::Node &dot_node = digraph.new_node("");
dot_node.set_background_color("white");
Vector<std::string> input_names;
Vector<std::string> output_names;
for (const InputSocketRef *socket : node->inputs()) {
input_names.append(socket->name());
}
for (const OutputSocketRef *socket : node->outputs()) {
output_names.append(socket->name());
}
dot_nodes.add_new(node,
dot::NodeWithSocketsRef(dot_node, node->name(), input_names, output_names));
}
for (const OutputSocketRef *from_socket : output_sockets_) {
for (const InputSocketRef *to_socket : from_socket->directly_linked_sockets()) {
dot::NodeWithSocketsRef &from_dot_node = dot_nodes.lookup(&from_socket->node());
dot::NodeWithSocketsRef &to_dot_node = dot_nodes.lookup(&to_socket->node());
digraph.new_edge(from_dot_node.output(from_socket->index()),
to_dot_node.input(to_socket->index()));
}
}
return digraph.to_dot_string();
}
const NodeTreeRef &get_tree_ref_from_map(NodeTreeRefMap &node_tree_refs, bNodeTree &btree)
{
return *node_tree_refs.lookup_or_add_cb(&btree,
[&]() { return std::make_unique<NodeTreeRef>(&btree); });
}
PointerRNA NodeRef::rna() const
{
PointerRNA rna;
RNA_pointer_create(&tree_->btree()->id, &RNA_Node, bnode_, &rna);
return rna;
}
PointerRNA SocketRef::rna() const
{
PointerRNA rna;
RNA_pointer_create(&this->tree().btree()->id, &RNA_NodeSocket, bsocket_, &rna);
return rna;
}
} // namespace blender::nodes