select userid, from_unixtime(createtime,'yyyy-MM-dd') as dateid, regexp_extract(browser,'^([^\\(]*).*$',1) as browser, operationsystem, device, row_number() over (partition by userid order by createtime) as rn from ods_log_full where module='user' and action='login_success' and from_unixtime(createtime,'yyyy-MM-dd')='2023-07-26'
如果这样的话,那么中文会出现乱码。
暂时的解决办法时,使用hive的regexp_replace函数。这个函数没有上述的问题。
标签:number,Hive,乱码,regexp,extract,createtime,row From: https://www.cnblogs.com/lhk20213937/p/18108118