[jira] [Updated] (PHOENIX-973) Sequences with null names created

["James Taylor (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/PHOENIX-973?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

James Taylor updated PHOENIX-973:
---------------------------------
    Attachment: PHOENIX-973.patch

The root cause of this issue is that the lexer was eating characters without giving an error. In this case, the table name was prefixed with an underscore which is invalid unless surrounded by double quotes. I've fixed the lexer, surround these with double quotes, and ensure that our generated sequence names have both a schema and sequence/table name.

> Sequences with null names created
> ---------------------------------
>
>                 Key: PHOENIX-973
>                 URL: https://issues.apache.org/jira/browse/PHOENIX-973
>             Project: Phoenix
>          Issue Type: Bug
>            Reporter: Samarth Jain
>         Attachments: PHOENIX-973.patch
>
>
> While working on refactoring Phoenix tests, I see sequences created with null names like _SEQ_T.null. Of course, parser doesn't like when it sees the sequence name as null causing DROP SEQUENCE _SEQ_T.null to fail. 
> [~jamestaylor] mentioned we use sequence names with _SEQ_T for creating ids for indexes on views. So that is likely the area where this bug resides. Needs further investigation.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)