参数< titlesplit >1.分析接口Network可以找到返回json格式数据的查询接口。请求url中可以观察到请求的参数如下二、token参数解析首先看下token长什么样子eJx1j0tvgkAUhf/LbCUwM7zEpAtEoGixSBWqTRcgb3nJoLQ0/e+dJjZpF01ucs589+TkzgforAjMEIQKhAy4xh2YAcRCVgIM6AndiFM8lSQoKVgRGHD8y3hZZkDYeQswe0GQhwxCEn79Ri4lL0jBFMEpfGV+eyzQ+U5ZNASyvm/JjOPSka3ivL8ENXtsKo56kuUcPeOfAKAN1ZY2UD3dNLhp//O26XdoBcnTmrp4OZTFFj8OhbrJYkUTotWb9lwsnnJhY7irXBjmvqNWRRGjxtEDbeDVtWt77UVN/F2HnO1EqYVNceRLUsYkGebxtoMbf5AnSceF6UMx6prsrZZX06+qzGgP6VDKonV/Fj0r0xxThzweF2PNS/tDOFrqsTZcJT/BMrEP5mjKxA3O+jpansXIHJdOYXlms67uJ49T1WhJI+pe79rXkSv3D3HUNVHSDfiScbv3LOftBKZK2IYRCZoox8gzrN4hPe9ypzvw+QWfPJQk一眼看上去有点像base64编码，先用base64解析一下试试importbase64s='eJx1j0tvgkAUhf/LbCUwM7zEpAtEoGixSBWqTRcgb3nJoLQ0/e+dJjZpF01ucs589+TkzgforAjMEIQKhAy4xh2YAcRCVgIM6AndiFM8lSQoKVgRGHD8y3hZZkDYeQswe0GQhwxCEn79Ri4lL0jBFMEpfGV+eyzQ+U5ZNASyvm/JjOPSka3ivL8ENXtsKo56kuUcPeOfAKAN1ZY2UD3dNLhp//O26XdoBcnTmrp4OZTFFj8OhbrJYkUTotWb9lwsnnJhY7irXBjmvqNWRRGjxtEDbeDVtWt77UVN/F2HnO1EqYVNceRLUsYkGebxtoMbf5AnSceF6UMx6prsrZZX06+qzGgP6VDKonV/Fj0r0xxThzweF2PNS/tDOFrqsTZcJT/BMrEP5mjKxA3O+jpansXIHJdOYXlms67uJ49T1WhJI+pe79rXkSv3D3HUNVHSDfiScbv3LOftBKZK2IYRCZoox8gzrN4hPe9ypzvw+QWfPJQk'print(s)result=base64.b64decode(s)print(result)先试下decode为utf-8，报错无法直接解析，可能是经过压缩，尝试解压importzlibresult2=zlib.decompress(result)print(result2)resultsareasfollowsb'{"rId":100900,"ver":"1.0.6","ts":1582866069294,"cts":1582866069377,"brVD":[1030,1162],"brR":[[1920,1080],[1920,1080],24,24],"bI":["https://gz.meituan.com/meishi/","https://gz.meituan.com/"],"mT":[],"kT":[],"aT":[],"tT":[],"aM":"","sign":"eJwljT2OwjAQhe9C4dKxCXjDSi4QFRKi4wBWPAmjje1oPEaCw3ANRMVpuAfWUr1PT+9n4Qjc3lslesfwBeTr0QWw7+fr/bgLjzEC7VKJvGWmmhFpZgwl75IHq5VIhCPGE032zDzn36YZbzIAcnFR9ik0lfMZGzG7sRaqENdJq5dGzJPjIVGoNmH+O8AFpso5EVtRMvz/lYLedrodfrw2uh/Uyhi3Mf0g9bpbdsaodi21VFItPst3R/k="}'基本能看懂了大部分参数，ts和cts是时间戳，注意位数ts为时间戳*1000左右，cts比tsSlightlylarger,otherparametersareconstantsignTrytoparsesign='eJwljT2OwjAQhe9C4dKxCXjDSi4QFRKi4wBWPAmjje1oPEaCw3ANRMVpuAfWUr1PT+9n4Qjc3lslesfwBeTr0QWw7+fr/bgLjzEC7VKJvGWminthesamewaymhFpZgwl75IHq5VIhCPGE032zDzn36YZbzIAcnFR9ik0lfMZGzG7sRaqENdJq5dGzJPjIVGoNmH+O8AFpso5EVtRMvz/lYLedrodfrw2uh/Uyhi3Mf0g9bpbdsaodi21VFItPst3R/k='result3=zlib.decompress(base64.b64decode(sign))print(result3)结果为一系列参数拼接：b'"areaId=0&cateId=0&cityName=\xe5\xb9\xbf\xe5\xb7\x9e&dinnerCountAttrId=&optimusCode=10&originUrl=https://gz.meituan.com/meishi/&page=1&partner=126&platform=1&riskLevel=1&sort=&userId=&uuid=813f7d161cf0466a96cf.1582866035的逆向操作大概是一波.有几个问题需要注意.sign中的参数需要转成string类型而不是b代码,在github上放一份.如果有需要可以提供完整的代码结果.目前,storepage的数据抽取比较常见，这里就不细说了，发现数据显示以前有很多totalCounts，但实际上过程中只有67页的数据使用浏览器操作。因此，我如果你想获取完整的数据，细分区域，对这些参数进行分类，通过多次请求排列组合即可获取更多的数据。
