Hive导入csv文件

时间:2022-06-28 鸭梨山大哎人气:0

正文

现有文件为csv格式,需要导入hive中,设csv内容如下

1001,zs,23
1002,lis,24

首先创建表

create table if not exists csv2(
    uid int,
    uname string,
    age int
)
row format serde 'org.apache.hadoop.hive.serde2.OpenCSVSerde'
stored as textfile ;

导入数据及查询

load data local inpath '/data/csv2.csv' into table csv2;
select * from csv2;

其他注意事项

如果建表是parquet格式可否load导入csv文件?

drop table csv2;
create table if not exists csv2(
    uid int,
    uname string,
    age int
)
row format serde 'org.apache.hadoop.hive.serde2.OpenCSVSerde'
stored as parquet ;

load data local inpath '/data/csv2.csv' into table csv2;
select * from csv2;

使用时会报错

Failed with exception java.io.IOException:java.lang.RuntimeException: hdfs://192.168.10.101:8020/user/hive/warehouse/csv2/csv2.csv is not a Parquet file. expected magic number at tail [80, 65, 82, 49] but found [44, 50, 52, 10]

**不可以,需要先导入成textfile,之后再从临时表导入成parquet,**如下

drop table csv2;
create table if not exists csv2
(
    uid   int,
    uname string,
    age   int
)
    row format serde 'org.apache.hadoop.hive.serde2.OpenCSVSerde'
    stored as textfile;
-- 先导入csv文件到表格csv2,保存格式是textfile
load data local inpath '/data/csv2.csv' into table csv2;


drop table csv3;
-- 创建csv3,保存格式parquet
create table if not exists csv3
(
    uid   int,
    uname string,
    age   int
)
    row format delimited
        fields terminated by ','
    stored as parquet;
-- 提取csv2的数据插入到csv3
insert overwrite table csv3 select * from csv2;

总结

关键是要引入org.apache.hadoop.hive.serde2.OpenCSVSerde
csv要保存到hive的parquet,需要先保存成textfile

加载全部内容