一、Json文件解析
案例：rating.json文件
{“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″}
1、求出每个人评分最高的3部电影
2、求出被评分次数最多的3部电影
用hive求解过程
step1：上传文件到hdfs
在sftp窗口直接拖动文件到/home/hadoop/soft
step2：创建一张表rate_json：字段jsonline，加载json文件数据
create external table if not exists rate_json(jsonline string) ;
load data local inpath ‘/home/hadoop/soft/rating.json’ into table rate_json;
step3：创建一张表rate：movieId int,rate int,timeStamp bigint,uid int
create external table if not exists rate(movieId int,rate int,ts bigint,uid int) row format delimited fields terminated by ‘\t’;
报错：FAILED: ParseException line 1:63 Failed to recognize predicate ‘timeStamp’. Failed rule: ‘identifier’ in column specification
原因：timeStamp为关键字
step4：解析json对象属性为对应表字段，注$后字段对应json对象属性字段
insert into table rate select get_json_object(jsonline,’$.movie’) as movie,get_json_object(jsonline,’$.rate’) as rate,get_json_object(jsonline,’$.timeStamp’) as ts,get_json_object(jsonline,’$.uid’) as uid from rate_json;
step5：查询5条
select * from Rate limit 5；
表rate中数据
二、清洗数据三种方式
1、创建表的时候使用<正则正反序列化>替换特殊分隔符为单字节分隔符
row format serde ‘org.apache.hadoop.hive.serde2.RegexSerDe’
with serdeproperties(‘input.regex’='(.*)::(.*)::(.*)’,’output.format.string’=’%1$s %2$s %3$s’);
2、写shell脚本replace.sh利用文本处理命令替换双字节分隔符为单字节分隔符
/home/hadoop/hive/shell/replace.sh
#!/bin/bash
sed -i “s/::/:/g” `grep :: -rl /home/hadoop/hive/ratings.dat`
sed -i “s/::/:/g” `grep :: -rl /home/hadoop/hive/movies.dat`
sed -i “s/::/:/g” `grep :: -rl /home/hadoop/hive/users.dat`
3、自定义InputFormat组件
项目：InputFormat_UDF
类名：
cn.serenity.format.MyInputFormat
cn.serenity.format.BiRecordReader
//这里是我们修改的地方：
//加载数据的时候（也就是使用load data [local] inpath）会将所有出现||字符的替换为|
//因此创建表的时候我们制定的是第二个参数’|’
String str = value.toString().replaceAll(“\\|\\|”, “\\|”);
value.set(str);
使用：项目导出为inputformat.jar
上传至/home/hadoop/soft/
转移到/home/hadoop/apps/apache-hive-bin-1.2.1/lib目录下
重启hive
在创建movie表时指定
stored as inputformat ‘cn.serenity.format.MyInputFormat’ outputformat
‘org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat’;
三、用户自定义函数GenericUDTF
案例：影评练习
（8）该影评库中各种类型电影中评价最高的5部电影（类型，电影名，平均影评分）
1、使用UDTF
注意代码中接收到的属性为：
fieldNames.add(“movieid”);
fieldNames.add(“movietitle”);
fieldNames.add(“movietype”);
String[] fields = string.split(“:”);
表中字段和分隔符必须与代码中的字段名和分隔符名一致
create table if not exists small_movie(movieid int,movietitle string,movietype string) row format delimited fields terminated by ‘:’;
load data local inpath ‘/home/hadoop/hive/small_movie’ into table small_movie;
数据样式 1::Toy Story (1995)::Animation|Children’s|Comedy
该需求需要我们将电影类型拆分成n行，就上面这条记录来说，需要拆分成3行
step1：上传movies.dat文件以及jar包到到/home/hadoop/soft,
mv /home/hadoop/soft/movies.dat /home/hadoop/hive/
mv /home/hadoop/soft/InputFormat_UDF.jar /home/hadoop/udf/
并且在hive窗口下使用命令将jar包添加到class path下
hive> add jar /home/hadoop/udf/InputFormat_UDF.jar;
结果：
Added [/home/hadoop/udf/InputFormat_UDF.jar] to class path
Added resources: [/home/hadoop/udf/InputFormat_UDF.jar]
step2：调用hive/replace.sh清洗数据
sh /home/hadoop/hive/replace.sh
step3：hive窗口下创建表,注意分隔字段要个清洗数据时指定的替换字符一致
create table if not exists movie(movieid int,movietitle string,movietype string) row format delimited fields terminated by ‘:’;
step4：加载文件movies.dat数据到表movie中
load data local inpath ‘/home/hadoop/hive/movies.dat’ into table movie;
step5：创建临时函数关联该jar
create temporary function movie_type_split as ‘cn.serenity.udtf.MovieTypeUDTF’;
step5：使用
2、使用lateral view explode
create table if not exists small_movie(movieId int,moviename string,movietype string) row format delimited fields terminated by ‘:’;
load data local inpath ‘/home/hadoop/hive/small_movie’ into table small_movie;
SELECT movieId,moviename,movie2.type FROM small_movie a LATERAL VIEW explode(split(a.movietype, ‘\\|’)) movie2 AS type;
四、复杂类型的json解析
+——————————————————————-+
json
+——————————————————————-+
{“store”:
{“fruit”:[{“weight”:8,”type”:”apple”},{“weight”:9,”type”:”pear”}],
“bicycle”:{“price”:19.95,”color”:”red”}
},
“email”:”amy@only_for_json_udf_test.net”,
“owner”:”amy”
}
+——————————————————————-+
json内容可以用以下查询语句解析
hive> SELECT get_json_object(src_json.json, ‘$.owner’) FROM src_json;
amy
hive> SELECT get_json_object(src_json.json, ‘$.store.fruit[0]’) FROM src_json;
{“weight”:8,”type”:”apple”}
hive> SELECT get_json_object(src_json.json, ‘$.non_exist_key’) FROM src_json;
NULL
