[jira] [Commented] (JENA-512) SSE Tags are used inconsistently

["ASF subversion and git services (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/JENA-512?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13740381#comment-13740381 ] 

ASF subversion and git services commented on JENA-512:
------------------------------------------------------

Commit 1514086 from [~rvesse] in branch 'jena/trunk'
[ https://svn.apache.org/r1514086 ]

Add Tags.tagIsNumeric and associated SSE builder (JENA-512)
                
> SSE Tags are used inconsistently
> --------------------------------
>
>                 Key: JENA-512
>                 URL: https://issues.apache.org/jira/browse/JENA-512
>             Project: Apache Jena
>          Issue Type: Bug
>          Components: ARQ
>    Affects Versions: Jena 2.10.1
>            Reporter: Rob Vesse
>            Assignee: Rob Vesse
>            Priority: Minor
>              Labels: algebra, sse
>
> In some of our more recent code we are trying to map functions into a messaging data structure by inspecting the function symbol.  However we have found that there are some inconsistencies in the symbols.
> Namely that the capitalization and punctuation are not consistent, likely this cannot change as it would mean breaking many existing algebra examples/tests and systems like ours that use algebra strings internally.
> The more concerning inconsistencies are that some functions report symbols without using the tag constants and so some queries written report symbols that don't match their tag constants.
> BuilderExpr appears to get around this by doing case insensitive key lookup which seems very hacky
> There is also at least one function (isNumeric) which has no Tag constant and no SSE builder defined for it so queries containing this cannot be decoded from algebra.
> I plan to do two things:
> - Make expression classes return their Tags constant where they don't already and particularly in the cases where the two values aren't exact matches
> - Fix the isNumeric case (and any others I discover) where there is no registered builder for the symbol

--
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