English | 中文
libhv
是一个类似于libevent、libev、libuv
的跨平台网络库,提供了更易用的接口和更丰富的协议。
- libhv QQ群:
739352073
,欢迎加群交流 - libhv 源码剖析: https://hewei.blog.csdn.net/article/details/123295998
- libhv 接口手册: https://hewei.blog.csdn.net/article/details/103976875
- libhv 教程目录: https://hewei.blog.csdn.net/article/details/113733758
- libhv教程01--介绍与体验
- libhv教程02--编译与安装
- libhv教程03--链库与使用
- libhv教程04--编写一个完整的命令行程序
- libhv教程05--事件循环以及定时器的简单使用
- libhv教程06--创建一个简单的TCP服务端
- libhv教程07--创建一个简单的TCP客户端
- libhv教程08--创建一个简单的UDP服务端
- libhv教程09--创建一个简单的UDP客户端
- libhv教程10--创建一个简单的HTTP服务端
- libhv教程11--创建一个简单的HTTP客户端
- libhv教程12--创建一个简单的WebSocket服务端
- libhv教程13--创建一个简单的WebSocket客户端
- libhv教程14--200行实现一个纯C版jsonrpc框架
- libhv教程15--200行实现一个C++版protorpc框架
- libhv教程16--多线程/多进程服务端编程
- libhv教程17--Qt中使用libhv
- libhv教程18--动手写一个tinyhttpd
- libhv教程19--MQTT的实现与使用
- 跨平台(Linux, Windows, MacOS, BSD, Solaris, Android, iOS)
- 高性能事件循环(网络IO事件、定时器事件、空闲事件、自定义事件)
- TCP/UDP服务端/客户端/代理
- TCP支持心跳、重连、转发、多线程安全write和close等特性
- 内置常见的拆包模式(固定包长、分界符、头部长度字段)
- 可靠UDP支持: WITH_KCP
- SSL/TLS加密通信(可选WITH_OPENSSL、WITH_GNUTLS、WITH_MBEDTLS)
- HTTP服务端/客户端(支持https http1/x http2 grpc)
- HTTP支持静态文件服务、目录服务、代理服务、同步/异步API处理函数
- HTTP支持RESTful风格、URI路由、keep-alive长连接、chunked分块等特性
- WebSocket服务端/客户端
- MQTT客户端
libhv提供了以下构建方式:
1、通过Makefile:
./configure
make
sudo make install
2、通过cmake:
mkdir build
cd build
cmake ..
cmake --build .
3、通过vcpkg:
vcpkg install libhv
4、通过xmake:
xrepo install libhv
运行脚本./getting_started.sh
:
# 下载编译
git clone https://github.com/ithewei/libhv.git
cd libhv
./configure
make
# 运行httpd服务
bin/httpd -h
bin/httpd -d
#bin/httpd -c etc/httpd.conf -s restart -d
ps aux | grep httpd
# 文件服务
bin/curl -v localhost:8080
# 目录服务
bin/curl -v localhost:8080/downloads/
# API服务
bin/curl -v localhost:8080/ping
bin/curl -v localhost:8080/echo -d "hello,world!"
bin/curl -v localhost:8080/query?page_no=1\&page_size=10
bin/curl -v localhost:8080/kv -H "Content-Type:application/x-www-form-urlencoded" -d 'user=admin&pswd=123456'
bin/curl -v localhost:8080/json -H "Content-Type:application/json" -d '{"user":"admin","pswd":"123456"}'
bin/curl -v localhost:8080/form -F 'user=admin' -F 'pswd=123456'
bin/curl -v localhost:8080/upload -d "@LICENSE"
bin/curl -v localhost:8080/upload -F "file=@LICENSE"
bin/curl -v localhost:8080/test -H "Content-Type:application/x-www-form-urlencoded" -d 'bool=1&int=123&float=3.14&string=hello'
bin/curl -v localhost:8080/test -H "Content-Type:application/json" -d '{"bool":true,"int":123,"float":3.14,"string":"hello"}'
bin/curl -v localhost:8080/test -F 'bool=1' -F 'int=123' -F 'float=3.14' -F 'string=hello'
# RESTful API: /group/:group_name/user/:user_id
bin/curl -v -X DELETE localhost:8080/group/test/user/123
# 压力测试
bin/wrk -c 1000 -d 10 -t 4 http://127.0.0.1:8080/
c版本: examples/tcp_echo_server.c
c++版本: evpp/TcpServer_test.cpp
#include "TcpServer.h"
using namespace hv;
int main() {
int port = 1234;
TcpServer srv;
int listenfd = srv.createsocket(port);
if (listenfd < 0) {
return -1;
}
printf("server listen on port %d, listenfd=%d ...\n", port, listenfd);
srv.onConnection = [](const SocketChannelPtr& channel) {
std::string peeraddr = channel->peeraddr();
if (channel->isConnected()) {
printf("%s connected! connfd=%d\n", peeraddr.c_str(), channel->fd());
} else {
printf("%s disconnected! connfd=%d\n", peeraddr.c_str(), channel->fd());
}
};
srv.onMessage = [](const SocketChannelPtr& channel, Buffer* buf) {
// echo
channel->write(buf);
};
srv.setThreadNum(4);
srv.start();
// press Enter to stop
while (getchar() != '\n');
return 0;
}
注意:
以上示例只是简单的echo
服务,TCP是流式协议,实际应用中请务必添加边界进行拆包。
文本协议建议加上\0
或者\r\n
分隔符,可参考 examples/jsonrpc;
二进制协议建议加上自定义协议头,通过头部长度字段表明负载长度,可参考 examples/protorpc;
通过setUnpack
(c接口即hio_set_unpack
)设置拆包规则,支持固定包长、分隔符、头部长度字段三种常见的拆包方式,
内部根据拆包规则处理粘包与分包,保证onMessage
回调上来的是完整的一包数据,大大节省了上层处理粘包与分包的成本。
不想自定义协议和拆包组包的可直接使用现成的HTTP/WebSocket
协议。
channel->write
(c接口即hio_write
)是非阻塞的(事件循环异步编程里所有的一切都要求是非阻塞的),且多线程安全的。
发送大数据时应该做流控,通过onWriteComplete
监听写完成事件,在可写时再发送下一帧数据。
具体示例代码可参考 examples/tinyhttpd.c 中的 http_serve_file
。
channel->close
(c接口即hio_close
) 也是多线程安全的,这可以让网络IO事件循环线程里接收数据、拆包组包、反序列化后放入队列,
消费者线程/线程池从队列里取出数据、处理后发送响应和关闭连接,变得更加简单。
c版本: examples/tcp_client_test.c
c++版本: evpp/TcpClient_test.cpp
#include <iostream>
#include "TcpClient.h"
using namespace hv;
int main() {
int port = 1234;
TcpClient cli;
int connfd = cli.createsocket(port);
if (connfd < 0) {
return -1;
}
cli.onConnection = [](const SocketChannelPtr& channel) {
std::string peeraddr = channel->peeraddr();
if (channel->isConnected()) {
printf("connected to %s! connfd=%d\n", peeraddr.c_str(), channel->fd());
} else {
printf("disconnected to %s! connfd=%d\n", peeraddr.c_str(), channel->fd());
}
};
cli.onMessage = [](const SocketChannelPtr& channel, Buffer* buf) {
printf("< %.*s\n", (int)buf->size(), (char*)buf->data());
};
cli.start();
std::string str;
while (std::getline(std::cin, str)) {
if (str == "close") {
cli.closesocket();
} else if (str == "start") {
cli.start();
} else if (str == "stop") {
cli.stop();
break;
} else {
if (!cli.isConnected()) break;
cli.send(str);
}
}
return 0;
}
见examples/http_server_test.cpp
golang gin 风格
#include "HttpServer.h"
using namespace hv;
int main() {
HttpService router;
router.GET("/ping", [](HttpRequest* req, HttpResponse* resp) {
return resp->String("pong");
});
router.GET("/data", [](HttpRequest* req, HttpResponse* resp) {
static char data[] = "0123456789";
return resp->Data(data, 10);
});
router.GET("/paths", [&router](HttpRequest* req, HttpResponse* resp) {
return resp->Json(router.Paths());
});
router.GET("/get", [](HttpRequest* req, HttpResponse* resp) {
resp->json["origin"] = req->client_addr.ip;
resp->json["url"] = req->url;
resp->json["args"] = req->query_params;
resp->json["headers"] = req->headers;
return 200;
});
router.POST("/echo", [](const HttpContextPtr& ctx) {
return ctx->send(ctx->body(), ctx->type());
});
HttpServer server;
server.registerHttpService(&router);
server.setPort(8080);
server.setThreadNum(4);
server.run();
return 0;
}
注意:
上面示例直接运行在main
主线程,server.run()
会阻塞当前线程运行,所以router
和server
对象不会被析构,
如使用server.start()
内部会另起线程运行,不会阻塞当前线程,但需要注意router
和server
的生命周期,
不要定义为局部变量被析构了,可定义为类成员变量或者全局变量,下面的WebSocket
服务同理。
见examples/http_client_test.cpp
python requests 风格
#include "requests.h"
int main() {
auto resp = requests::get("http://www.example.com");
if (resp == NULL) {
printf("request failed!\n");
} else {
printf("%s\n", resp->body.c_str());
}
resp = requests::post("127.0.0.1:8080/echo", "hello,world!");
if (resp == NULL) {
printf("request failed!\n");
} else {
printf("%s\n", resp->body.c_str());
}
return 0;
}
见examples/websocket_server_test.cpp
#include "WebSocketServer.h"
using namespace hv;
int main(int argc, char** argv) {
WebSocketService ws;
ws.onopen = [](const WebSocketChannelPtr& channel, const HttpRequestPtr& req) {
printf("onopen: GET %s\n", req->Path().c_str());
};
ws.onmessage = [](const WebSocketChannelPtr& channel, const std::string& msg) {
printf("onmessage: %.*s\n", (int)msg.size(), msg.data());
};
ws.onclose = [](const WebSocketChannelPtr& channel) {
printf("onclose\n");
};
WebSocketServer server;
server.registerWebSocketService(&ws);
server.setPort(9999);
server.setThreadNum(4);
server.run();
return 0;
}
见examples/websocket_client_test.cpp
#include "WebSocketClient.h"
using namespace hv;
int main(int argc, char** argv) {
WebSocketClient ws;
ws.onopen = []() {
printf("onopen\n");
};
ws.onmessage = [](const std::string& msg) {
printf("onmessage: %.*s\n", (int)msg.size(), msg.data());
};
ws.onclose = []() {
printf("onclose\n");
};
// reconnect: 1,2,4,8,10,10,10...
reconn_setting_t reconn;
reconn_setting_init(&reconn);
reconn.min_delay = 1000;
reconn.max_delay = 10000;
reconn.delay_policy = 2;
ws.setReconnect(&reconn);
ws.open("ws://127.0.0.1:9999/test");
std::string str;
while (std::getline(std::cin, str)) {
if (!ws.isConnected()) break;
if (str == "quit") {
ws.close();
break;
}
ws.send(str);
}
return 0;
}
- 事件循环: examples/hloop_test.c
- TCP回显服务: examples/tcp_echo_server.c
- TCP聊天服务: examples/tcp_chat_server.c
- TCP代理服务: examples/tcp_proxy_server.c
- UDP回显服务: examples/udp_echo_server.c
- UDP代理服务: examples/udp_proxy_server.c
- SOCKS5代理服务: examples/socks5_proxy_server.c
- HTTP服务: examples/tinyhttpd.c
- HTTP代理服务: examples/tinyproxyd.c
- jsonRPC示例: examples/jsonrpc
- MQTT示例: examples/mqtt
- 多accept进程模式: examples/multi-thread/multi-acceptor-processes.c
- 多accept线程模式: examples/multi-thread/multi-acceptor-threads.c
- 一个accept线程+多worker线程: examples/multi-thread/one-acceptor-multi-workers.c
- 事件循环: evpp/EventLoop_test.cpp
- 事件循环线程: evpp/EventLoopThread_test.cpp
- 事件循环线程池: evpp/EventLoopThreadPool_test.cpp
- TCP服务端: evpp/TcpServer_test.cpp
- TCP客户端: evpp/TcpClient_test.cpp
- UDP服务端: evpp/UdpServer_test.cpp
- UDP客户端: evpp/UdpClient_test.cpp
- HTTP服务端: examples/http_server_test.cpp
- HTTP客户端: examples/http_client_test.cpp
- WebSocket服务端: examples/websocket_server_test.cpp
- WebSocket客户端: examples/websocket_client_test.cpp
- protobufRPC示例: examples/protorpc
- Qt中使用libhv示例: examples/qt
- 网络连接工具: examples/nc
- 网络扫描工具: examples/nmap
- HTTP服务程序: examples/httpd
- HTTP压测工具: examples/wrk
- URL请求工具: examples/curl
- 文件下载工具: examples/wget
- 服务注册与发现: examples/consul
cd echo-servers
./build.sh
./benchmark.sh
吞吐量:
libevent running on port 2001
libev running on port 2002
libuv running on port 2003
libhv running on port 2004
asio running on port 2005
poco running on port 2006
==============2001=====================================
[127.0.0.1:2001] 4 threads 1000 connections run 10s
total readcount=1616761 readbytes=1655563264
throughput = 157 MB/s
==============2002=====================================
[127.0.0.1:2002] 4 threads 1000 connections run 10s
total readcount=2153171 readbytes=2204847104
throughput = 210 MB/s
==============2003=====================================
[127.0.0.1:2003] 4 threads 1000 connections run 10s
total readcount=1599727 readbytes=1638120448
throughput = 156 MB/s
==============2004=====================================
[127.0.0.1:2004] 4 threads 1000 connections run 10s
total readcount=2202271 readbytes=2255125504
throughput = 215 MB/s
==============2005=====================================
[127.0.0.1:2005] 4 threads 1000 connections run 10s
total readcount=1354230 readbytes=1386731520
throughput = 132 MB/s
==============2006=====================================
[127.0.0.1:2006] 4 threads 1000 connections run 10s
total readcount=1699652 readbytes=1740443648
throughput = 165 MB/s
# sudo apt install iperf
iperf -s -p 5001 > /dev/null &
bin/tcp_proxy_server 1212 127.0.0.1:5001 &
iperf -c 127.0.0.1 -p 5001 -l 8K
iperf -c 127.0.0.1 -p 1212 -l 8K
带宽:
------------------------------------------------------------
[ 3] local 127.0.0.1 port 52560 connected with 127.0.0.1 port 5001
[ ID] Interval Transfer Bandwidth
[ 3] 0.0-10.0 sec 20.8 GBytes 17.9 Gbits/sec
------------------------------------------------------------
[ 3] local 127.0.0.1 port 48142 connected with 127.0.0.1 port 1212
[ ID] Interval Transfer Bandwidth
[ 3] 0.0-10.0 sec 11.9 GBytes 10.2 Gbits/sec
# sudo apt install wrk
wrk -c 100 -t 4 -d 10s http://127.0.0.1:8080/
# sudo apt install apache2-utils
ab -c 100 -n 100000 http://127.0.0.1:8080/
libhv(port:8080) vs nginx(port:80)
以上测试结果可以在 Github Actions 中查看。
如果您在使用libhv
,欢迎通过PR将信息提交至此列表,让更多的用户了解libhv
的实际使用场景,以建立更好的网络生态。
用户 (公司名/项目名/个人联系方式) | 案例 (项目简介/业务场景) |
---|---|
阅面科技 | 猎户AIoT平台设备管理、人脸检测HTTP服务、人脸搜索HTTP服务 |
socks5-libhv | socks5代理 |
hvloop | 类似uvloop的python异步IO事件循环 |
tsproxyd-android | 一个基于libhv实现的android端web代理服务 |
玄舟智维 | C100K设备连接网关服务 |