var ip = req.headers['x-forwarded-for'] ||
     req.socket.remoteAddress ||
     null;

请注意，有时您可以在req.headers['x-forwarded-for']中获得多个IP地址。此外，并不总是设置x-forward -for报头，这可能会抛出错误。

该字段的一般格式为:

x-forward -for: client, proxy1, proxy2, proxy3

其中的值是一个逗号+空格分隔的IP地址列表，最左边是原始客户端，每个传递请求的后续代理添加接收请求的IP地址。在本例中，请求通过proxy1、proxy2和proxy3传递。Proxy3显示为请求的远程地址。

这是Arnav Gupta提出的解决方案，Martin在下面的评论中为未设置x-forward -for的情况提出了修复建议:

var ip = (req.headers['x-forwarded-for'] || '').split(',').pop().trim() || 
         req.socket.remoteAddress

使用现代JS的建议:

仅在设置时处理x-forward -for，如果设置了，则取第一个地址 其他参数使用可选链接(?.)

const parseIp = (req) =>
    req.headers['x-forwarded-for']?.split(',').shift()
    || req.socket?.remoteAddress

console.log(parseIp(req))
// => 127.0.0.1

请求。headers['x-forwarded-for'] || request.connection.remoteAddress . headers['x-forwarded-for'

如果有x-forward -for报头，则使用它，否则使用. remoteaddress属性。

The x-forwarded-for header is added to requests that pass through load balancers (or other types of proxy) set up for HTTP or HTTPS (it's also possible to add this header to requests when balancing at a TCP level using proxy protocol). This is because the request.connection.remoteAddress the property will contain the private IP address of the load balancer rather than the public IP address of the client. By using an OR statement, in the order above, you check for the existence of an x-forwarded-for header and use it if it exists otherwise use the request.connection.remoteAddress.

在Typescript中使用ValidatorJS。下面是NodeJS中间件:

// Extract Client IP Address
app.use((req, res, next) => {
    let ipAddress = (req.headers['x-forwarded-for'] as string || '').split(',')[0]
    if (!validator.isIP(ipAddress))
        ipAddress = req.socket.remoteAddress?.toString().split(':').pop() || ''
    if (!validator.isIP(ipAddress))
        return res.status(400).json({errorMessage: 'Bad Request'})

    req.headers['x-forwarded-for'] = ipAddress
    next()
})

在这里，我假设所有请求都应该有一个有效的IP地址，因此如果没有找到有效的IP地址，则返回一个代码为400的响应。

获取ip地址有两种方式:

让IP = req.ip 让ip = req.connection.remoteAddress;

但上述方法存在一个问题。

如果你在Nginx或任何代理程序后面运行你的应用程序，每个IP地址将是127.0.0.1。

因此，获取user的ip地址的最佳方案是:-

let ip = req.header('x-forwarded-for') || req.connection.remoteAddress;

如果你正在使用Graphql-Yoga，你可以使用以下函数:

const getRequestIpAddress = (request) => { const requestIpAddress = request.request。headers['X-Forwarded-For'] || request.request.connection.remoteAddress . headers['X-Forwarded-For' if (!requestIpAddress)返回null const ipv4 = new RegExp(“(?:(?:25(0 - 5)| 2[0 - 9][0 - 4]|[01]?[0 - 9][0 - 9]?)\){3}(?:25(0 - 5)| 2[0 - 9][0 - 4]|[01]?[0 - 9][0 - 9]?)”) const [ipAddress] = requesttipaddress .match(ipv4) 返回ipAddress ｝

这里有很多很棒的观点，但没有一个是全面的，所以这里是我最终使用的:

function getIP(req) {
  // req.connection is deprecated
  const conRemoteAddress = req.connection?.remoteAddress
  // req.socket is said to replace req.connection
  const sockRemoteAddress = req.socket?.remoteAddress
  // some platforms use x-real-ip
  const xRealIP = req.headers['x-real-ip']
  // most proxies use x-forwarded-for
  const xForwardedForIP = (() => {
    const xForwardedFor = req.headers['x-forwarded-for']
    if (xForwardedFor) {
      // The x-forwarded-for header can contain a comma-separated list of
      // IP's. Further, some are comma separated with spaces, so whitespace is trimmed.
      const ips = xForwardedFor.split(',').map(ip => ip.trim())
      return ips[0]
    }
  })()
  // prefer x-forwarded-for and fallback to the others
  return xForwardedForIP || xRealIP || sockRemoteAddress || conRemoteAddress
}