[DISCUSS]Separator of Hive query results

[rita <ri...@163.com>]

Dear: 

Wechat group "Apache Linkis community Development group "chat records are as follows:  微信群"Apache Linkis 社区开发群"的聊天记录如下:

—————  2022-8-25  —————

hcl  12:08

 



 

这里的分隔符是写死的吗@Enjoyyin_尹强@WDS @peacewong@WDS

Is the delimiter here written dead @enjoyyin_Yoon Qiang@wds @peacewong@WDS

 

 

peacewong@WDS  12:28

这个是0.x的吧？

This is 0. X, right?

 

peacewong@WDS  12:29

hive查询的结果就是制表符分割的

Hive query results are separated by tabs

 hcl  13:57

如果字段内容中包含\t呢

What if the contents of the field contain \t

 r@FY2  13:58

下面抛了异常

I threw an exception down here

 

 hcl  14:08

不能正常处理吗。。

Can't handle it properly..

 

 hcl  14:10

jdbc查询是ok的

The JDBC query is OK

 r@FY2  14:10

自己手动改一下吧[捂脸]

Change it yourself manually. [Facepalm]

 

 hcl  14:13

外部表分隔符，用户自定义的，千奇百怪了。

External table delimiters, user - defined, all kinds of strange.

 

 r@FY2  14:15

这个只是对result的切分吧，linkis默认是制表符，split之前把制表符去掉，sendResult不就正常了吗

Linkis is TAB by default. If TAB is removed before split, sendResult will be normal, right

 

 peacewong@WDS  14:40

是的，这个和表分隔符没关系的。 Yes, this has nothing to do with table separators.是列数据里面如果有制表符会导致数据列异常的

  Is column data if there is a TAB character will cause data column exception

 

 hcl  15:15

ok,懂了

Ok, got it

 

 hcl  15:37

通过修改 hive.fetch.output.serde 是不是就可以修改序列化的方式。

You can modify the serialization mode by modifying Hive.fetch.output. serde.

 

 peacewong@WDS  15:41

是的，1.0也加了这个功能

Yes, 1.0 also added this feature

 hcl  15:48

如果设置成多分隔符的话，他应该会获取表的信息，根据表的分隔符进行序列化吧。 If set to multiple delimiters, it should fetch the table's information and serialize it according to the table's delimiter.