hive中存Array类型的数据的案例，将字符串的数组格式转成数组的字符串，自定义函数方式处理‘[12,23,23,34]‘字符串格式的数据为array＜int＞格式的数据。

1、创建表带有Array的表：

create table t_afan_test
(
info1 array<int>,
info2 array<string>
)
ROW FORMAT DELIMITED
FIELDS TERMINATED BY '\t'
COLLECTION ITEMS TERMINATED BY ',';

2、插入数据：

insert into t_afan_test values (array(12,23,23,34),array("what","are","this"));
insert into t_afan_test values (array(12,23,23,34,56,32),array("what","are","this","aaa"));

3、查询出的结果如下：

4、再如以下案例：

drop table if exists t_afan_test;

create table t_afan_test
(
info1 string
)
ROW FORMAT DELIMITED
FIELDS TERMINATED BY '\t'
COLLECTION ITEMS TERMINATED BY ',';

insert into t_afan_test values ('[12,23,23,34]');
insert into t_afan_test values ('[22,33,43,54]');

select * from t_afan_test;

+--------------------+
| t_afan_test.info1  |
+--------------------+
| [12,23,23,34]      |
| [22,33,43,54]      |
+--------------------+

drop table if exists tmp_xxx;

create table tmp_xxx as
select 
split(regexp_extract(`info1`,'^\\[(.*)]$',1),',') as key_word_label
from t_afan_test 
where `info1` is not null limit 10;

select * from tmp_xxx;

+-------------------------+
| tmp_xxx.key_word_label  |
+-------------------------+
| ["22","33","43","54"]   |
| ["12","23","23","34"]   |
+-------------------------+

select collect_list(cast(array_element as int)) int_array 
from (select explode(key_word_label) array_element from tmp_xxx) s;

+----------------------------+
|         int_array          |
+----------------------------+
| [22,33,43,54,12,23,23,34]  |
+----------------------------+

select explode(key_word_label) array_element from tmp_xxx;

+----------------+
| array_element  |
+----------------+
| 22             |
| 33             |
| 43             |
| 54             |
| 12             |
| 23             |
| 23             |
| 34             |
+----------------+

5、将’[12,23,23,34]'格式数据转成struct<> , array 的方式：

转成array编写UDF，代码如下：

package com.xxx.stringtoarray;

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

import java.util.ArrayList;
import java.util.List;

/**
 * 将数组格式的字符串 转成 整型数组
 * @author tzq
 */
public final class StringToArray extends UDF {

    private static final String NULL_STRING = "null";

    /**
     * 如果想最后hive的数据格式是struct<>, 返回值是Integer[]的。
     * @param sourceText     :源字符串
     * @return
     */
    public List<Integer> evaluate(String sourceText) {
        if (isBlank(sourceText)) {
            return null;
        }
        if (NULL_STRING.equalsIgnoreCase(sourceText)) {
            return null;
        }

        String[] arr1 = sourceText.replace("[","").replace("]","").split(",");
        //Integer[] arr2 = new Integer[arr1.length];
        List<Integer> list = new ArrayList<>();

        for(int i = 0; i < arr1.length; i++) {
            list.add(Integer.parseInt(arr1[i]));
        }

        return list;
    }

    public static boolean isBlank(String str) {
        int strLen;
        if (str != null && (strLen = str.length()) != 0) {
            for(int i = 0; i < strLen; ++i) {
                if (!Character.isWhitespace(str.charAt(i))) {
                    return false;
                }
            }

            return true;
        } else {
            return true;
        }
    }

}

hive中创建临时函数：

hive> add jar /xxx/xxx/xxx/xx.jar;
hive> create temporary function stringToArray as 'com.xxx.stringtoarray.StringToArray';

使用的时候，类似如下：

drop table if exists t_afan_test;
create table t_afan_test
(
info1 string
)
ROW FORMAT DELIMITED
FIELDS TERMINATED BY '\t'
COLLECTION ITEMS TERMINATED BY ',';

insert into t_afan_test values ('[12,23,23,34]');
insert into t_afan_test values ('[22,33,43,54]');

drop table if exists tmp_xxx;

create table tmp_xxx as
select 
 stringToArray(`info1`)  as key_word_label
from t_afan_test;

查看结果：