以下是我的数据文件，我想创建一个表将是什么分隔符？

Question

我对hive中的数据使用以下查询。

CREATE EXTERNAL TABLE IF NOT EXISTS aircel1 (subscriberID INT, towerID STRING, dataDownloaded STRING)
ROW FORMAT DELIMITED
FIELDS TERMINATED BY ''
STORED AS TEXTFILE
LOCATION '/user/username/name';

数据如下所示的分隔符是什么。

subId=00001111911128052627towerid=11232w34532543456345623453456984756894756bytes=122112212212212218.4621702216543667E17
    subId=00001111911128052639towerid=11232w34532543456345623453456984756894756bytes=122112212212212219.6726312167218586E17
    subId=00001111911128052615towerid=11232w34532543456345623453456984756894756bytes=122112212212212216.9431647633139046E17
    subId=00001111911128052615towerid=11232w34532543456345623453456984756894756bytes=122112212212212214.7836041833447418E17
    subId=00001111911128052639towerid=11232w34532543456345623453456984756894756bytes=122112212212212219.0366596827240525E17
    subId=00001111911128052619towerid=11232w34532543456345623453456984756894756bytes=122112212212212218.0686280014540467E17
    subId=00001111911128052658towerid=11232w34532543456345623453456984756894756bytes=122112212212212216.9860890496178944E17
    subId=00001111911128052652towerid=11232w34532543456345623453456984756894756bytes=122112212212212218.303981333116041E17

Answer 1

您可以尝试使用正则表达式（我还没有对此进行测试）

CREATE EXTERNAL TABLE IF NOT EXISTS aircel1 (
  subscriberID STRING, towerID STRING, dataDownloaded STRING
) 
ROW FORMAT SERDE 'org.apache.hadoop.hive.serde2.RegexSerDe' 
WITH SERDEPROPERTIES ('input.regex'='subId=(.*)towerid=(.*)bytes=(.*)')
LOCATION '/user/username/dirname';

Answer 2

使用等号作为分隔符，您可以分两步构建表格。

首先，创建一个包含所有字符串列的临时表。

例如，第一列是字符串00001111911128052627towerid。

然后使用实际数据类型创建“真实表”，然后您可以将字符串"towerid"子串出来，例如，从第一列

Answer 3

我们可以使用以下代码完成任务，

create external table table1 (del string, subid string, towerid string, bytes double)

row format delimited
fields terminated by '='
location '/user/murali/';

create table table2 (subid string, towerid string, bytes double);

insert table table2 select 
    substring(subid,1,20),substring(towerid,1,41),bytes from table1;

select * from table2;