DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://nagoya.apache.org/bugzilla/show_bug.cgi?id=26509>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=26509

DataSourceRealm for local contexts

           Summary: DataSourceRealm for local contexts
           Product: Tomcat 5
           Version: 5.0.16
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: Enhancement
          Priority: Other
         Component: Catalina
        AssignedTo: [EMAIL PROTECTED]
        ReportedBy: [EMAIL PROTECTED]


DataSourceRealms only validate against DataSources defined
in the global context.  Several folks were confused by this,
or asked to have it changed:

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=16316
http://nagoya.apache.org/bugzilla/show_bug.cgi?id=24545
http://nagoya.apache.org/bugzilla/show_bug.cgi?id=24723
http://nagoya.apache.org/bugzilla/show_bug.cgi?id=25805

As well, there has been comments on the tomcat-dev mailing list about this:

http://www.servlets.com/archive/servlet/BrowseList?listName=tomcat-dev&by=subject&from=148510

I was surprised by this this week -- the fact that DataSourceRealm only
works against the global context is undocumented (except in mailing lists).

I'm going to guess that if I just asked for things to be changed, the request
would be rejected, as all the other requests were.  Besides, changing the
existing DataSourceRealm's behaviour could break existing code.  I have
a different proposal:

I'ld like to introduce a new class: LocalDataSourceRealm.  It would be
identical to DataSourceRealm, except that it queries the container's JNDI
context, instead of the global one.  I've coded up a subclass of DataSourceRealm
to do this; it works for me.

I did have to make one change to DataSourceRealm to avoid copying the
whole file: I changed the open() method to be protected instead of private.
I changed close() to protected also for symmetry.

Remy, if you find this more acceptable I'll attach a full patch, including
documentation updates.

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to