[
https://issues.apache.org/jira/browse/OPENJPA-2558?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14289841#comment-14289841
]
ASF subversion and git services commented on OPENJPA-2558:
----------------------------------------------------------
Commit 1654352 from [~struberg] in branch 'openjpa/trunk'
[ https://svn.apache.org/r1654352 ]
OPENJPA-2558 add getRepresentation to allow BooleanRepresentation be used in
SQLBuffer
> Implement a way to select the db representation of Boolean values
> -----------------------------------------------------------------
>
> Key: OPENJPA-2558
> URL: https://issues.apache.org/jira/browse/OPENJPA-2558
> Project: OpenJPA
> Issue Type: New Feature
> Components: jdbc
> Affects Versions: 2.3.0, 2.4.0
> Reporter: Mark Struberg
> Assignee: Mark Struberg
> Attachments: OPENJPA-2558.patch
>
>
> Many Databases do not natively support the SQL-99 BOOLEAN datatype. Boolean
> and boolean fields in JPA entities will thus get represented as another type
> in the database. OpenJPA currently has a hardcoded mapping to int 1/0 in the
> base DBDictionary.
> But many existing databases don't use NUMBER(1) as column type but e.g.
> CHAR(1) (because it uses less space in the index).
> There are also all sorts of possible representations where one of those is
> usually consistently used throughout a customer application:
> * Native Boolean -> PreparedStatement#setBoolean
> * Int 0/1 -> PreparedStatement#setInt
> * String "Y"/"N" -> PreparedStatement#setString
> * String "T"/"F" -> PreparedStatement#setString
> * String "y"/"n" -> PreparedStatement#setString
> * String "t"/"f" -> PreparedStatement#setString
> And I've also seen translated ones: "J"/"N" in german apps for "Ja"/ "Nein"
> (yes/no).
> So it should be possible to even add your own custom BooleanRepresentation.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
