NodeCombine.cpp

Go to the documentation of this file.

#include "NodeCombine.hpp"
 
#include "MayaFlux/Nodes/NodeGraphManager.hpp"
 
#include "MayaFlux/Journal/Archivist.hpp"
 
namespace MayaFlux::Nodes {
 
BinaryOpContext::BinaryOpContext(double value, double lhs_value, double rhs_value)
    : NodeContext(value, typeid(BinaryOpContext).name())
    , lhs_value(lhs_value)
    , rhs_value(rhs_value)
{
}
 
BinaryOpContextGpu::BinaryOpContextGpu(double value, double lhs_value, double rhs_value, std::span<const float> gpu_data)
    : BinaryOpContext(value, lhs_value, rhs_value)
    , GpuVectorData(gpu_data)
{
    type_id = typeid(BinaryOpContextGpu).name();
}
 
CompositeOpContext::CompositeOpContext(double value, std::vector<double> input_values)
    : NodeContext(value, typeid(CompositeOpContext).name())
    , input_values(std::move(input_values))
{
}
 
CompositeOpContextGpu::CompositeOpContextGpu(double value, std::vector<double> input_values, std::span<const float> gpu_data)
    : CompositeOpContext(value, std::move(input_values))
    , GpuVectorData(gpu_data)
{
    type_id = typeid(CompositeOpContextGpu).name();
}
 
BinaryOpNode::BinaryOpNode(const std::shared_ptr<Node>& lhs, const std::shared_ptr<Node>& rhs, CombineFunc func)
    : m_lhs(lhs)
    , m_rhs(rhs)
    , m_func(std::move(func))
    , m_context(0.0, 0.0, 0.0)
    , m_context_gpu(0.0, 0.0, 0.0, get_gpu_data_buffer())
{
    if (!m_lhs || !m_rhs) {
        error<std::invalid_argument>(
            Journal::Component::Nodes, Journal::Context::Init,
            std::source_location::current(),
            "BinaryOpNode requires both lhs and rhs nodes to be non-null");
    }
}
 
BinaryOpNode::BinaryOpNode(
    const std::shared_ptr<Node>& lhs,
    const std::shared_ptr<Node>& rhs,
    CombineFunc func,
    NodeGraphManager& manager,
    ProcessingToken token)
    : BinaryOpNode(lhs, rhs, std::move(func))
{
    m_manager = &manager;
    m_token = token;
}
 
void BinaryOpNode::initialize()
{
    if (!m_is_initialized) {
 
        if (m_manager) {
            auto self = shared_from_this();
            uint32_t lhs_mask = m_lhs ? m_lhs->get_channel_mask().load() : 0;
            uint32_t rhs_mask = m_rhs ? m_rhs->get_channel_mask().load() : 0;
            uint32_t combined_mask = lhs_mask | rhs_mask;
 
            if (combined_mask != 0) {
                for (auto& channel : get_active_channels(combined_mask, 0)) {
                    m_manager->add_to_root(self, m_token, channel);
                }
            } else {
                m_manager->add_to_root(self, m_token, 0);
            }
            m_is_initialized = true;
        }
    }
 
    if (m_manager) {
        auto semantics = m_manager->get_node_config().binary_op_semantics;
        switch (semantics) {
        case NodeBinaryOpSemantics::REPLACE:
            if (m_lhs) {
                for (auto& channel : get_active_channels(m_lhs, 0)) {
                    m_manager->remove_from_root(m_lhs, m_token, channel);
                }
            }
            if (m_rhs) {
                for (auto& channel : get_active_channels(m_rhs, 0)) {
                    m_manager->remove_from_root(m_rhs, m_token, channel);
                }
            }
            break;
        case NodeBinaryOpSemantics::KEEP:
        default:
            break;
        }
    }
}
 
double BinaryOpNode::process_sample(double input)
{
    if (!m_lhs || !m_rhs) {
        return input;
    }
 
    if (!is_initialized())
        initialize();
 
    atomic_inc_modulator_count(m_lhs->m_modulator_count, 1);
    atomic_inc_modulator_count(m_rhs->m_modulator_count, 1);
 
    uint32_t lstate = m_lhs->m_state.load();
    if (lstate & NodeState::PROCESSED) {
        m_last_lhs_value = input + m_lhs->get_last_output();
    } else {
        m_last_lhs_value = m_lhs->process_sample(input);
        atomic_add_flag(m_lhs->m_state, NodeState::PROCESSED);
    }
 
    uint32_t rstate = m_rhs->m_state.load();
    if (rstate & NodeState::PROCESSED) {
        m_last_rhs_value = input + m_rhs->get_last_output();
    } else {
        m_last_rhs_value = m_rhs->process_sample(input);
        atomic_add_flag(m_rhs->m_state, NodeState::PROCESSED);
    }
 
    m_last_output = m_func(m_last_lhs_value, m_last_rhs_value);
 
    if (!m_state_saved || (m_state_saved && m_fire_events_during_snapshot))
        notify_tick(m_last_output);
 
    atomic_dec_modulator_count(m_lhs->m_modulator_count, 1);
    atomic_dec_modulator_count(m_rhs->m_modulator_count, 1);
 
    try_reset_processed_state(m_lhs);
    try_reset_processed_state(m_rhs);
 
    return m_last_output;
}
 
std::vector<double> BinaryOpNode::process_batch(unsigned int num_samples)
{
    std::vector<double> output(num_samples);
    for (unsigned int i = 0; i < num_samples; ++i) {
        output[i] = process_sample(0.0);
    }
    return output;
}
 
void BinaryOpNode::notify_tick(double value)
{
    update_context(value);
    auto& ctx = get_last_context();
 
    for (const auto& callback : m_callbacks) {
        callback(ctx);
    }
 
    for (const auto& [callback, condition] : m_conditional_callbacks) {
        if (condition(ctx)) {
            callback(ctx);
        }
    }
}
 
void BinaryOpNode::reset_processed_state()
{
    atomic_remove_flag(m_state, NodeState::PROCESSED);
    if (m_lhs)
        m_lhs->reset_processed_state();
    if (m_rhs)
        m_rhs->reset_processed_state();
}
 
void BinaryOpNode::save_state()
{
    m_saved_last_lhs_value = m_last_lhs_value;
    m_saved_last_rhs_value = m_last_rhs_value;
 
    if (m_lhs)
        m_lhs->save_state();
    if (m_rhs)
        m_rhs->save_state();
 
    m_state_saved = true;
}
 
void BinaryOpNode::restore_state()
{
    m_last_lhs_value = m_saved_last_lhs_value;
    m_last_rhs_value = m_saved_last_rhs_value;
 
    if (m_lhs)
        m_lhs->restore_state();
    if (m_rhs)
        m_rhs->restore_state();
 
    m_state_saved = false;
}
 
void BinaryOpNode::update_context(double /*value*/)
{
    if (m_gpu_compatible) {
        m_context_gpu.value = m_last_output;
        m_context_gpu.lhs_value = m_last_lhs_value;
        m_context_gpu.rhs_value = m_last_rhs_value;
    } else {
        m_context.value = m_last_output;
        m_context.lhs_value = m_last_lhs_value;
        m_context.rhs_value = m_last_rhs_value;
    }
}
 
NodeContext& BinaryOpNode::get_last_context()
{
    if (m_gpu_compatible) {
        return m_context_gpu;
    }
    return m_context;
}
 
bool BinaryOpNode::is_initialized() const
{
    auto lstate = m_lhs->m_state.load();
    auto rstate = m_rhs->m_state.load();
    bool is_lhs_registered = m_lhs ? (lstate & NodeState::ACTIVE) : false;
    bool is_rhs_registered = m_rhs ? (rstate & NodeState::ACTIVE) : false;
    return !is_lhs_registered && !is_rhs_registered;
}
 
namespace detail {
 
    void composite_initialize(
        const std::vector<std::shared_ptr<Node>>& inputs,
        NodeGraphManager& manager,
        ProcessingToken token,
        const std::shared_ptr<Node>& self)
    {
        uint32_t combined_mask = 0;
        for (const auto& input : inputs)
            combined_mask |= input->get_channel_mask().load();
 
        if (combined_mask != 0) {
            for (auto ch : get_active_channels(combined_mask, 0))
                manager.add_to_root(self, token, ch);
        } else {
            manager.add_to_root(self, token, 0);
        }
    }
 
    void composite_apply_semantics(
        const std::vector<std::shared_ptr<Node>>& inputs,
        NodeGraphManager& manager,
        ProcessingToken token)
    {
        if (manager.get_node_config().binary_op_semantics != NodeBinaryOpSemantics::REPLACE)
            return;
 
        for (const auto& input : inputs) {
            for (auto ch : get_active_channels(input, 0))
                manager.remove_from_root(input, token, ch);
        }
    }
 
    void composite_validate(const std::vector<std::shared_ptr<Node>>& inputs, size_t N)
    {
        if (N > 0 && inputs.size() != N) {
            error<std::invalid_argument>(
                Journal::Component::Nodes, Journal::Context::Init,
                std::source_location::current(),
                "CompositeOpNode<{}> requires exactly {} inputs, got {}",
                N, N, inputs.size());
        }
 
        if (inputs.size() < 2) {
            error<std::invalid_argument>(
                Journal::Component::Nodes, Journal::Context::Init,
                std::source_location::current(),
                "CompositeOpNode requires at least 2 inputs, got {}",
                inputs.size());
        }
 
        for (size_t i = 0; i < inputs.size(); ++i) {
            if (!inputs[i]) {
                error<std::invalid_argument>(
                    Journal::Component::Nodes, Journal::Context::Init,
                    std::source_location::current(),
                    "CompositeOpNode input at index {} is null", i);
            }
        }
    }
}
 
} // namespace MayaFlux::Nodes