Hive学习笔记——HQL用法及UDF,Transform

Hive中没有定义专门的数据格式，数据格式可以由用户指定，用户定义数据格式需要指定三个属性：列分隔符（通常为空格、” ”、”x001″）、行分隔符 （” ”）以及读取文件数据的方法（Hive 中默认有三个文件格式 TextFile，SequenceFile 以及 RCFile）。由于在加载数据的过程中，不需要从用户数据格式到 Hive 定义的数据格式的转换，因此，Hive 在加载的过程中不会对数据本身进行任何修改，而只是将数据内容复制或者移动到相应的 HDFS 目录中。而在数据库中，不同的数据库有不同的存储引擎，定义了自己的数据格式。所有数据都会按照一定的组织存储，因此，数据库加载数据的过程会比较耗时。

基本数据类型
tinyint/smallint/int/bigint
float/double
boolean
string
复杂数据类型
Array/Map/Struct

 external:

//external
CREATE EXTERNAL TABLE tab_ip_ext(id int, name string,
     ip STRING,
     country STRING)
 ROW FORMAT DELIMITED FIELDS TERMINATED BY ','
 STORED AS TEXTFILE
 LOCATION '/external/hive';

hdfs上的数据存在/external/hive上面。这样删除tab_ip_ext表，/external/hive里面的内容不会删除，而且/external/hive里面的内容也不会被剪切。

CTAS:

// CTAS  用于创建一些临时表存储中间结果
CREATE TABLE tab_ip_ctas
   AS
SELECT id new_id, name new_name, ip new_ip,country new_country
FROM tab_ip_ext
SORT BY new_id;

insert:

//insert from select   用于向临时表中追加中间结果数据
create table tab_ip_like like tab_ip;

insert overwrite table tab_ip_like
    select * from tab_ip;

Hive中不能一条一条insert，但是可以成组的insert。不加overwrite会报错。

insert into table t_cost_like select * from t_cost_ctas; 

这种方式可以追加。

partition:

 create table t_cost_pt(   
     id int, 
     name string,
     capacity string,
     price double)
     partitioned by (data string)
     row format delimited
     fields terminated by '	';

 load data local inpath '/home/hadoop/hivedata/cost.data' into table t_cost_pt partition(data='001');
 load data local inpath '/home/hadoop/hivedata/cost.data' into table t_cost_pt partition(data='002');
 load data local inpath '/home/hadoop/hivedata/cost.data' into table t_cost_pt partition(data='003');

我们可以看一下partition实际的存储结构，其实就是文件夹。

show partitions t_cost_pt;

select * from t_cost_pt where data='001'

 array:

//array 
create table tab_array(a array<int>,b array<string>)
row format delimited
fields terminated by '	'
collection items terminated by ',';

示例数据
tobenbrone,laihama,woshishui     13866987898,13287654321
abc,iloveyou,itcast     13866987898,13287654321


select a[0] from tab_array;
select * from tab_array where array_contains(b,'word');
insert into table tab_array select array(0),array(name,ip) from tab_ext t; 

map：

//map
create table tab_map(name string,info map<string,string>)
row format delimited
fields terminated by '	'
collection items terminated by ';'
map keys terminated by ':';

示例数据：
fengjie            age:18;size:36A;addr:usa
furong        age:28;size:39C;addr:beijing;weight:180KG


load data local inpath '/home/hadoop/hivetemp/tab_map.txt' overwrite into table tab_map;
insert into table tab_map select name,map('name',name,'ip',ip) from tab_ext; 

struct：

//struct
create table tab_struct(name string,info struct<age:int,tel:string,addr:string>)
row format delimited
fields terminated by '	'
collection items terminated by ','

load data local inpath '/home/hadoop/hivetemp/tab_st.txt' overwrite into table tab_struct;
insert into table tab_struct select name,named_struct('age',id,'tel',name,'addr',country) from tab_ext;

select:

select * from tab_ext sort by id desc limit 5;

select a.ip,b.book from tab_ext a join tab_ip_book b on(a.name=b.name);

在shell下执行hive语句：

hive -S -e 'select country,count(*) from chenchi.tab_ext' > /home/hadoop/hivetemp/e.txt

chenchi是库名。

有了这种执行机制，就使得我们可以利用脚本语言（bash shell,python）进行hql语句的批量执行。

hive的udf(自定义函数)：

0.要继承org.apache.hadoop.hive.ql.exec.UDF类实现evaluate

自定义函数调用过程：
1.添加jar包（在hive命令行里面执行）
hive> add jar /root/NUDF.jar;

2.创建临时函数
hive> create temporary function getNation as 'cn.itcast.hive.udf.NationUDF';

3.调用
hive> select id, name, getNation(nation) from beauty;

4.将查询结果保存到HDFS中
hive> create table result row format delimited fields terminated by '	' as select * from beauty order by id desc;    
hive> select id, getAreaName(id) as name from tel_rec;


create table result row format delimited fields terminated by '	' as select id, getNation(nation) from beauties;

实战：自定义函数UDF

需求：

13884243554 234 450
13664243554 242 440
13994243554 211 430
13444243554 222 420

自定义一个识别手机号所在地区的函数getarea，然后通过该函数进行hive的查询。

写一个Java类，定义相关的函数逻辑

打成jar包

上传到hive的lib下

在hive中创建一个函数getarea，跟jar包中的自定义java类建立关联。

代码如下：要把hive的lib下面的jar包全部导入。

package com.darrenchan.bigdata;

import java.util.HashMap;

import org.apache.hadoop.hive.ql.exec.UDF;

public class PhoneNumToArea extends UDF {
    private static HashMap<String, String> areaMap = new HashMap<>();
    static {
        areaMap.put("1388", "beijing");
        areaMap.put("1399", "tianjin");
        areaMap.put("1366", "nanjing");
    }

    // 一定要用public修饰才能被hive调用
    // 返回值以及参数个数及类型都可以随意指定
    public String evaluate(String pnb) {
        String result = areaMap.get(pnb.substring(0, 4)) == null ? (pnb + " huoxing")
                : (pnb + " " + areaMap.get(pnb.substring(0, 4)));
        return result;
    }
}

将程序打成jar包，并加入到hive中：

hive>add jar /home/hadoop/phonenum-to-area.jar;

创建该函数：

hive>create temporary function getarea as 'com.darrenchan.bigdata.PhoneNumToArea';

创建表：

hive>create table t_flow(phonenum string, upflow int, downflow int)
> row format delimited
> fields terminated by ' ';

为表导入数据：

hive>load data local inpath '/home/hadoop/hivedata/flow.txt' into table t_flow;

进行查询验证：

hive>select getarea(phonenum),upflow,downflow from t_flow;

结果如下：

13884243554 beijing 234 450
13664243554 nanjing 242 440
13994243554 tianjin 211 430
13444243554 huoxing 222 420

实战：自定义脚本Transform

Hive的 TRANSFORM 关键字提供了在SQL中调用自写脚本的功能

适合实现Hive中没有的功能又不想写UDF的情况。

1、先加载rating.json文件到hive的一个原始表 rat_json

create table rat_json(line string) row format delimited;
load data local inpath '/home/hadoop/rating.json' into table rat_json;

如下：

+----------------------------------------------------------------+--+
| rat_json.line |
+----------------------------------------------------------------+--+
| {"movie":"1193","rate":"5","timeStamp":"978300760","uid":"1"} |
| {"movie":"661","rate":"3","timeStamp":"978302109","uid":"1"} |
| {"movie":"914","rate":"3","timeStamp":"978301968","uid":"1"} |
| {"movie":"3408","rate":"4","timeStamp":"978300275","uid":"1"} |
| {"movie":"2355","rate":"5","timeStamp":"978824291","uid":"1"} |
| {"movie":"1197","rate":"3","timeStamp":"978302268","uid":"1"} |
| {"movie":"1287","rate":"5","timeStamp":"978302039","uid":"1"} |
| {"movie":"2804","rate":"5","timeStamp":"978300719","uid":"1"} |
| {"movie":"594","rate":"4","timeStamp":"978302268","uid":"1"} |
| {"movie":"919","rate":"4","timeStamp":"978301368","uid":"1"} |
+----------------------------------------------------------------+--+

2、需要解析json数据成四个字段，插入一张新的表 t_rating

create table t_rating as
select get_json_object(line,'$.movie') movie,get_json_object(line,'$.rate') rate,
get_json_object(line,'$.timeStamp') timestring,get_json_object(line,'$.uid') uid from rat_json;

3、使用transform+python的方式去转换unixtime为weekday

先编辑一个python脚本文件
########python######代码
vi weekday_mapper.py

#!/bin/python
import sys
import datetime

for line in sys.stdin:
  line = line.strip()
  movieid, rating, unixtime,userid = line.split('	')
  weekday = datetime.datetime.fromtimestamp(float(unixtime)).isoweekday()
  print '	'.join([movieid, rating, str(weekday),userid])

保存文件
然后，将文件加入hive的classpath：

hive>add FILE /home/hadoop/weekday_mapper.py;

hive>create TABLE u_data_new as
SELECT
  TRANSFORM (movie, rate, timestring,uid)
  USING 'python weekday_mapper.py'
  AS (movie, rate, weekday,uid)
FROM t_rating;

检验：