06.2_c/c++开源库boost_coroutine2 协程库

1.安装与说明

安装

sudo apt install libboost-coroutine1.71-dev

编译链接

libboost-coroutine不支持.pc格式查看, 支持.cmake导入
cat /usr/lib/x86_64-linux-gnu/cmake/boost_coroutine-1.71.0/boost_coroutine-config.cmake
cat /usr/lib/x86_64-linux-gnu/cmake/boost_coroutine-1.71.0/libboost_coroutine-variant-shared.cmake

list(APPEND _BOOST_COROUTINE_DEPS atomic chrono context thread headers)
1

可知依赖: atomic chrono context thread

从安装路径可知
头文件路径: /usr/include/boost/coroutine2/coroutine.hpp
/usr/include/boost/coroutine2/all.hpp
库文件路径: /usr/lib/x86_64-linux-gnu/libboost_coroutine.so

故从路径可知
编译选项:
链接选项: -lboost_coroutine -lboost_context

2.实例

1.代码

1_boost.coroutine2协程实例.cc

#include <iostream>
#include <boost/coroutine2/all.hpp>

// 协程函数，接收一个 sink 对象作为参数，用于输出数据
void fibonacci(boost::coroutines2::coroutine<int>::push_type& sink) {
    int a = 0, b = 1;
    while (true) 
    {
        sink(a);
        std::tie(a, b) = std::make_tuple(b, a + b);
        if (a > 100) break; // 限制输出前10个斐波那契数
    }
}

int main() {
    // 创建一个 pull 类型的协程对象，它将调用 fibonacci 函数
    boost::coroutines2::coroutine<int>::pull_type source(fibonacci);

    // 从协程中拉取并打印斐波那契数
    while (source) 
    {
        std::cout << "Main: " << source.get() << std::endl;
        source();
    }

    std::cout << std::endl;

    return 0;
}
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29

2_boost.coroutine2_协程.cc

#include<iostream>
#include<boost/coroutine2/coroutine.hpp>

using namespace boost::coroutines2;

void coro_function(coroutine<int>::push_type& sink) {
    for (int i = 0; i < 5; ++i)
    {
        std::cout << "Coroutine: " << i << std::endl;
        sink(i);
    }
}

int main() {
    coroutine<int>::pull_type source(coro_function);

    while (source)
    {
        std::cout << "Main: " << source.get() << std::endl;
        source();
    }

    return 0;
}
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

2.scons构建

SConstruct

## 模板1
import os
env = Environment()
env["PROGSUFFIX"] = ".out"              # 可执行文件后缀.out
env["CCFLAGS"] = " -g3 -O0 -Wall"       # gdb 调试
env["LIBS"] = ["boost_coroutine","boost_context"]

sources=0
ccflags=1
linkflags=2
target_list = [
    "1_boost.coroutine2协程实例.cc",
    "2_boost.coroutine2_协程.cc",
]
for target in target_list:
    if isinstance(target,(str)):
        env.Program(Split(target))
    elif isinstance(target,(list)):
        env.Program(Split(target[sources]),CCFLAGS=target[ccflags]+env["CCFLAGS"],LINKFLAGS=target[linkflags]+env["LINKFLAGS"])

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20

scons

scons: Reading SConscript files …
scons: done reading SConscript files.
scons: Building targets …
g++ -o 1_boost.coroutine2协程实例.o -c -g3 -O0 -Wall 1_boost.coroutine2协程实例.cc
g++ -o 1_boost.coroutine2协程实例.out 1_boost.coroutine2协程实例.o -lboost_coroutine -lboost_context
g++ -o 2_boost.coroutine2_协程.o -c -g3 -O0 -Wall 2_boost.coroutine2_协程.cc
g++ -o 2_boost.coroutine2_协程.out 2_boost.coroutine2_协程.o -lboost_coroutine -lboost_context
scons: done building targets.

3.运行

gitee在线代码

协程看打印结果,意义不大;
建议gdb调试, 跟踪查看协程是如何跳转, 流程处理顺序;
这样才能掌握的更深一些, 写出有价值的协程处理.

思考

1. 2个例子, 只做了2个协程相互交替运行
2. 如何像python asyncio, 多个协议同时交替运行
3. 对应系统调用阻塞API,read/select/sleep, 是否能够切换到另一个协程

---