[teiid-issues] [JBoss JIRA] (TEIID-1834) Add support for a comparable binary type

Steven Hawkins (JIRA) jira-events at lists.jboss.org
Mon Jan 23 13:56:18 EST 2012


    [ https://issues.jboss.org/browse/TEIID-1834?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12661257#comment-12661257 ] 

Steven Hawkins commented on TEIID-1834:
---------------------------------------

Added the varbinary type logic.  The runtime type is a byte[] wrapper called BinaryType.  Our batches and any values passed to translators will have binary values wrapped with a BinaryType.  However UDF functions will just use a type signature of byte[] (or object).  We are also taking ownership of built-in type metadata.  For now the first step was to remove unused information and to move the datatype info out of the System.vdb datatypes index and into a types.dat sv file in the metadata project.  The only modification was adding the varbinary type.  We'll need further work to correct/consolidate/refine type info.
                
> Add support for a comparable binary type
> ----------------------------------------
>
>                 Key: TEIID-1834
>                 URL: https://issues.jboss.org/browse/TEIID-1834
>             Project: Teiid
>          Issue Type: Feature Request
>          Components: Query Engine
>            Reporter: Steven Hawkins
>            Assignee: Steven Hawkins
>             Fix For: 8.0
>
>         Attachments: varbinary.patch
>
>
> Teiid currently lacks a binary/varbinary type.  Since we have both blob/clob, it would be natural to also introduce a varbinary analog to string.  It would be a comparable type and need an escape syntax for entering literals.  Translators would need to be modified to account for binary literals.
> See the forum post for current workarounds.  We also as a one off have previously allowed clob types to be comparable, see also TEIID-1248.  Another approach would by to allow clob/blob to comparable in Teiid, but possibly not at the source as per the searchability metadata.  We would still need to support an escape syntax and modify translators to use native type information and the appropriate literal format.  The advantage of this latter approach is that there is nothing more that designer must do to support this case.  The most practical benefit from introducing a varbinary type would be in having a better size estimate (it would be limited to 8000 bytes) for buffering.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.jboss.org/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira

        


More information about the teiid-issues mailing list