[
http://opensource.atlassian.com/projects/hibernate/browse/HHH-2235?page=c...
]
Ulf gohde commented on HHH-2235:
--------------------------------
You're definitely right about oracle having crazy ideas about null. It is almost a
religious topic for them. And the fact that a query for SOMETHINIG != 'bla' does
not give you rows with a null in that column back is aways driving me mad....
Just tested on SQLServer. You are right. As soon as I have one entry with null in the
compound key I can not insert a second one with null even if the other columns are
different. That's pretty weird behaviour.
I guess that invalidates my request....
Thanks for your time.
Support for null map-keys in <map> mappings
-------------------------------------------
Key: HHH-2235
URL:
http://opensource.atlassian.com/projects/hibernate/browse/HHH-2235
Project: Hibernate3
Issue Type: Bug
Components: core
Affects Versions: 3.2.0.ga
Reporter: Igor Lobanov
Priority: Minor
I'm not sure if I should file it as a bug.
Hibernate requires index column for <map> mappings to be not null. This is required
by the code in AbstractCollectionPersister#readIndex(...), which raises exception if index
column is null. However, when I override collection persister and force readIndex method
to return null, everything seems to work fine. Hibernate allows me to work with maps using
null value as a key. Though I havn't tested it extensivly yet.
What is the reason for disallowing null values as a keys? If there are possible
complicated error conditions, then this issue is a feature request for null-values support
for <map> mappings. If there are no errors possible, then constraining such behavior
is probably wrong.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
http://opensource.atlassian.com/projects/hibernate/secure/Administrators....
-
For more information on JIRA, see:
http://www.atlassian.com/software/jira