Hadoop Pig、Hive 自定义输入输出分隔符

01

add

jar
/home/june/hadoop/hive-0.8.1-bin/lib/hive_contrib.jar;

02

CREATE

TABLE

b(

03

c0
string,

04

c1
string,

05

c2
string)

06

ROW
FORMAT

07

SERDE

'org.apache.hadoop.hive.contrib.serde2.RegexSerDe'

08

WITH

SERDEPROPERTIES

09

(

'input.regex'

=

'([^,]*),,,,([^,]*),,,,([^,]*)'

,

10

'output.format.string'

=

'%1$s
%2$s %3$s'

)

11

STORED

AS

TEXTFILE;

12

13

cat
b.txt

14

1,,,,2,,,,3

15

a,,,,b,,,,c

16

9,,,,5,,,,7

17

load

data

local

inpath

'b.txt'

overwrite

into

table

b;

18

select

*

from

b

01

hive>

CREATE

TABLE

sunwg02
(id

int

,

name

STRING)

02

ROW
FORMAT SERDE

'org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe'

03

WITH

SERDEPROPERTIES
(

04

'field.delim'

=

'\t'

,

05

'escape.delim'

=

'\\'

,

06

'serialization.null.format'

='

07

)
STORED

AS

TEXTFILE;

08

OK

09

Time

taken:
0.046 seconds

10

11

hive>

insert

overwrite

table

sunwg02

select

*

from

sunwg00;

12

Loading
data

to

table

sunwg02

13

2

Rows

loaded

to

sunwg02

14

OK

15

Time

taken:
18.756 seconds

16

17

查看sunwg02在hdfs的文件

18

[hjl@sunwg
src]$ hadoop fs -cat /hjl/sunwg02/attempt_201105020924_0013_m_000000_0

19

mary

20

101
tom

21

22

NULL

值没有被转写成’\N’