一文教会你使用Nginx访问日志统计PV与UV

// index.js
import Cookies from 'js-cookie';
import {v4 as uuidv4} from 'uuid';

try {
  if (!Cookies.get('uuid')) {
    Cookies.set('uuid', uuidv4().replace(/-/g, ''), {expires: 30});
  }
  // 上报访问
  axios.get(`https://www.example.com/report.gif?t=${Date.now()}`);
}
catch (e) {}

location =/report.gif {
  return 204;
}

http {
  log_format main '$remote_addr - [$time_local] "$request" '
    ' - $status "uuid:$uuid" ';
  server {
    access_log /path/to/log/access443.log main;
    if ( $http_cookie ~* "uuid=([A-Z0-9]*)"){
        set $uuid $1;
    }
  }
}

#!/bin/bash
# 定位到脚本所在目录（注意我这里也是Nginx写访问日志的目录，当然这不是必须的）
log_base=$(cd `dirname $0`; pwd)
# 根据前一天的时间生成日志所在目录名
log_path=${log_base}/$(date -d yesterday +%Y)/$(date -d yesterday +%m)
# 创建日志目录
mkdir -p $log_path
# 将当前Nginx的日志移动到指定存放目录
mv $log_base/access443.log $log_path/$(date -d yesterday +%d).log
# 重新创建日志文件，给Nginx写日志用
touch $log_base/access443.log
# 给Nginx发送信号，注意你的Nginx目录可能不同
kill -USR1 `cat /www/server/nginx/logs/nginx.pid`

// stats.js
const fs = require('fs');
const path = require('path');
const args = process.argv.slice(2);
const [year] = args;

// 打印统计结果
function echo() {
  yearDir = year || '2022';
  const stats = statsYearLog(yearDir);
  Object.entries(stats)
    .sort(([a], [b]) => a - b)
    .forEach(([month, dateStats]) => {
    console.log(`${month}月`);
    Object.entries(dateStats)
      .sort(([a], [b]) => a - b)
      .forEach(([date, {pv, uv}]) => {
      console.log('  ', `${date}日`, `pv: ${pv}`, `uv: ${uv}`);
    });
    console.log('\n');
  });
}

// 统计某一年的数据
function statsYearLog(year) {
  // 读取目录下的文件夹名字
  const dir = path.resolve(__dirname, year);
  const monthDirList = fs.readdirSync(dir);
  
  const logMap = monthDirList.reduce((result, monthDir) => {
    const monthStats = statsMonthLog(year, monthDir);
    result[monthDir] = monthStats;
    return result;
  }, {});
  
  return logMap;
}

// 统计每个月的数据
function statsMonthLog(year, month) {
  const dir = path.resolve(__dirname, year, month);
  const dateLogList = fs.readdirSync(dir);
  
  const monthLogMap = dateLogList.reduce((result, dateLogFileName) => {
    const dateStats = statsDateLog(year, month, dateLogFileName);
    result[dateLogFileName.replace('.log', '')] = dateStats;
    return result;
  }, {});
  
  return monthLogMap;
}

// 统计某天的数据
function statsDateLog(year, month, dateFile) {
  const logPath = path.resolve(__dirname, year, month, dateFile);
  const logText = fs.readFileSync(logPath, 'utf-8');
  const logList = logText.split('\n');
  const pvLogList = logList.filter((line) => {
    return /report.gif/.test(line)
  });
  const uvLogMap = pvLogList.reduce((result, line) => {
    const match = line.match(/uuid:(\S+)"/);
    if (match && match[1]) {
      result[match[1]] = 1;
    }
    return result;
  }, {});
  
  return {pv: pvLogList.length, uv: Object.keys(uvLogMap).length};
}

// 执行打印统计结果
echo();