hive建表指定列分隔符为多字符分隔符实战(默认只支持单字符)_hive row formate ###

‘hdfs://CMBHHA/apps/hive/datahouse/test/table_tset’

### 4、几经搜索，问题终于解决，附上完整示例


#### 4.1、找到和hive版本对应的hive-contrib包，下载jar上传到hdfs上


[org/apache/hive/hive-contrib所有版本Jar文件及下载 -时代Java]( )


[hive-contrib包下载链接]( )


#### 4.2、以三个等号作为列分隔符 构建hive表的完整demo

add jar hdfs://team/work/libs/xxx/hive-contrib-2.0.1.jar;

drop table if exists mb_tmp.yz_1013;
create table if not exists mb_tmp.yz_1013(
test_id int,
work_place string,
other string
)
partitioned by(dt string)
ROW FORMAT SERDE ‘org.apache.hadoop.hive.contrib.serde2.MultiDelimitSerDe’ WITH SERDEPROPERTIES (“field.delim”=“===”)

load data local inpath ‘/data/home/xxxx/client/ct_data_1010.txt’ overwrite into table mb_tmp.yz_1013 partition(dt=‘2022-10-10’);

其中 以三个等号分割进行作为列分割的数据如下：

[work@xxxx client]$ cat ct_data_1010.txt
10011=beijing=chushi
10012=tianjin=changchun
10013=zhengzhou=siji
100014=zhangda=job

#### 4.3、结果验证

hive> desc mb_tmp.yz_1013;
OK
test_id int from deserializer
work_place string from deserializer
other string from deserializer
dt string

Partition Information

col_name data_type comment

dt string
Time taken: 0.332 seconds, Fetched: 9 row(s)

hive> set hive.cli.print.header=true;

hive> select * from mb_tmp.yz_1013 limit 100;
OK
yz_1013.test_id yz_1013.work_place yz_1013