[ 
https://issues.apache.org/jira/browse/HIVE-2693?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13535252#comment-13535252
 ] 

Gunther Hagleitner commented on HIVE-2693:
------------------------------------------

Mark, thanks for the additional tests, I'll take a closer look this afternoon.

To answer the questions:

1: I introduced +/-2 when I was at a point in the debugging stage where 
paranoia took over. I can remove that, it'll make the code more readable.
2: It is only implicitly tested in all the queries that use a reduce stage. I 
agree that a test of just that code would be good. Is there a place in the 
current unit tests that already does that/that I could use as a model?

Sign bit: I introduced a value for zero to avoid a "factor" of negative 
infinity. If you lump 0 into either the positive or negative bucket it would 
become the number that has an infinite number of zeros before the first 
non-zero digit (after the decimal point). MIN_INT might have been an option, 
but it seems cleaner to just make the sign have three states (-1,0,1). 
BigDecimal class in Java itself for instance more or less randomly defines 
precision of 0 (i.e.: number of unscaled digits) as 1. 

Non-deterministic order: 3.14 and 3.140 are indeed equal. Their representation 
should be exactly the same (<1>,<1>,<314>). Given that, I'm not sure how to 
enforce a deterministic order or even what that would be. Are you suggesting 
3.14 should always appear before 3.140?

I am worried about your comments about the where clause. I'll take a look at 
the tests. But you say it's not working right?

                
> Add DECIMAL data type
> ---------------------
>
>                 Key: HIVE-2693
>                 URL: https://issues.apache.org/jira/browse/HIVE-2693
>             Project: Hive
>          Issue Type: New Feature
>          Components: Query Processor, Types
>    Affects Versions: 0.10.0
>            Reporter: Carl Steinbach
>            Assignee: Prasad Mujumdar
>         Attachments: 2693_7.patch, 2693_8.patch, 2693_fix_all_tests1.patch, 
> HIVE-2693-10.patch, HIVE-2693-11.patch, HIVE-2693-12-SortableSerDe.patch, 
> HIVE-2693-13.patch, HIVE-2693-1.patch.txt, HIVE-2693-all.patch, 
> HIVE-2693-fix.patch, HIVE-2693.patch, HIVE-2693-take3.patch, 
> HIVE-2693-take4.patch
>
>
> Add support for the DECIMAL data type. HIVE-2272 (TIMESTAMP) provides a nice 
> template for how to do this.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Reply via email to