[ 
https://issues.apache.org/jira/browse/TAJO-644?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13922196#comment-13922196
 ] 

Jaehwa Jung edited comment on TAJO-644 at 3/6/14 9:26 AM:
----------------------------------------------------------

+1 for the patch.

Hi [~hyunsik].

Thank you for your contribution,

But we need to consider upper case and lower case.
I tested on my test cluster as follows:
{code:xml}
tajo> create table "AAA" (id int);
tajo> \d AAA;

table name: aaa
table path: hdfs://localhost:9010/tajo/warehouse/aaa
store type: CSV
number of rows: 0
volume: 0 B
Options: 

schema: 
id      INT4

tajo> create table "AAA" (id int);
relation "AAA" already exists
{code}

As you see, tajo low-cases table name by force.
But I think if users write double quotation marks, they will want to assort low 
case and upper case.
So, I suggest handling it at another issue because it will affects lots of 
classes.

Cheers.


was (Author: blrunner):
+1 for the patch.

Hi [~hyunsik].

Thank you for your contribution,

But we need to consider upper case and lower case.
I tested on my test cluster as follows:
{code:xml}
tajo> create table "AAA";
tajo> \d AAA;

table name: aaa
table path: hdfs://localhost:9010/tajo/warehouse/aaa
store type: CSV
number of rows: 0
volume: 0 B
Options: 

schema: 
id      INT4

tajo> create table "AAA" (id int);
relation "AAA" already exists
{code}

As you see, tajo low-cases table name by force.
But I think if users write double quotation marks, they will want to assort low 
case and upper case.
So, I suggest handling it at another issue because it will affects lots of 
classes.

Cheers.

> Support quoted identifiers
> --------------------------
>
>                 Key: TAJO-644
>                 URL: https://issues.apache.org/jira/browse/TAJO-644
>             Project: Tajo
>          Issue Type: Improvement
>    Affects Versions: 0.8-incubating, 1.0-incubating
>            Reporter: Hyunsik Choi
>            Assignee: Hyunsik Choi
>         Attachments: TAJO-644.patch
>
>
> In SQL standards, non-ascii identifiers are supported by using double 
> quotation as follows:
> {code}
> SELECT a, b, "씨", d from table1
> {code}
> We should support quoted identifiers.



--
This message was sent by Atlassian JIRA
(v6.2#6252)

Reply via email to