創建從不同的記錄格式爲csv文件

Question

文件的外部HIVE表我有一個CSV與由第一列值來定義不同的記錄格式文件： 樣本數據：

"EL","XXXXXXX", 2017-07-17 
"EH","XXXXXXX",1,2017-07-17,"AAA" 
"BI","XXXXXXX","AAA","BBBB" 

在這種情況下，我正在獲取3個已定義記錄類型的文件。 有沒有辦法將它加載到不同的配置單元表？

Answer 1

演示

create table el (s1 string,d1 date); 
create table eh (s1 string,i1 int,dt1 date,s2 string); 
create table bi (s1 string,s2 string,s3 string); 

---

create external table myfile 
(
    c1 string 
    ,c2 string 
    ,c3 string 
    ,c4 string 
    ,c5 string 
) 

row format serde 'org.apache.hadoop.hive.serde2.OpenCSVSerde' 
with serdeproperties 
(
    'separatorChar' = ',' 
    ,'quoteChar'  = '"' 
    ,'escapeChar' = '\\' 
) 
stored as textfile 
; 

---

select * from myfile; 

+-----+----------+--------------+-------------+-------+ 
| c1 | c2 |  c3  |  c4  | c5 | 
+-----+----------+--------------+-------------+-------+ 
| EL | XXXXXXX | 2017-07-17 | NULL  | NULL | 
| EH | XXXXXXX | 1   | 2017-07-17 | AAA | 
| BI | XXXXXXX | AAA   | BBBB  | NULL | 
+-----+----------+--------------+-------------+-------+ 

---

from myfile 
insert into el select c2,c3  where c1='EL' 
insert into eh select c2,c3,c4,c5 where c1='EH' 
insert into bi select c2,c3,c4 where c1='BI' 
; 

---

select * from el; 

+----------+-------------+ 
| s1 |  d1  | 
+----------+-------------+ 
| XXXXXXX | 2017-07-17 | 
+----------+-------------+ 

---

select * from eh; 

+----------+-----+-------------+------+ 
| s1 | i1 |  dt1  | s2 | 
+----------+-----+-------------+------+ 
| XXXXXXX | 1 | 2017-07-17 | AAA | 
+----------+-----+-------------+------+ 

---

select * from bi; 

+----------+------+-------+ 
| s1 | s2 | s3 | 
+----------+------+-------+ 
| XXXXXXX | AAA | BBBB | 
+----------+------+-------+