Problem with mixed case in suffix

[Endi Sukma Dewata <en...@vergenet.com>]

Hi,

 

I'm Endi Dewata from Verge Archemedia Labs in Austin, TX. I'm working with
Adison Wongkar trying to hook up our Virtual Directory product with
ApacheDS. So far so good, but it seems that we just found a minor problem
with suffixes that contain capital letters (e.g. dc=VergeNet,dc=com). It's
not a blocking problem for us, but I just want to know if this is a known
issue or has been addressed before. Here's the description of the problem:

 

 

In ServerContextFactory.java line #630, JdbmDatabase is initialized with
un-normalized suffix.

 

            Database db = new JdbmDatabase( upSuffix, wkdir );

 

But in JdbmDatabase.java line #673, when adding an entry to the database,
the suffix is compared with the normalized dn of the new entry.

 

        if ( dn.equals( suffix ) )

 

This is causing the add operation to fail.

 

 

A similar problem also happens here. In RootNexus.java line #203, the suffix
is being normalized during registration.

 

        backends.put( backend.getSuffix( true ).toString(), backend );

 

However, in RootNexus.java line #556, the dn used to look up the backend is
not normalized.

 

                return ( ContextPartition ) backends.get(
clonedDn.toString() );

 

This is causing the search operation to fail.

 

 

Should I file a bug? Thank you very much.

 

--

Endi S. Dewata

 

Re: Problem with mixed case in suffix

[Alex Karasulu <ao...@bellsouth.net>]

Also if you could (not necessary though) please submit a simple JUnit 
test case or two to demonstrate the bug.  If you attach it to your JIRA 
issue, I can commit the test contribution to svn and it would make 
fixing this a snap for me.

Thanks again,
Alex

Alex Karasulu wrote:

> Yep Endi you caught a bug - good job! Could you go through the motions 
> of filing it via JIRA?
>
> Thanks much,
> Alex
>
> Endi Sukma Dewata wrote:
>
>> Hi,
>>
>> I’m Endi Dewata from Verge Archemedia Labs in Austin, TX. I’m working 
>> with Adison Wongkar trying to hook up our Virtual Directory product 
>> with ApacheDS. So far so good, but it seems that we just found a 
>> minor problem with suffixes that contain capital letters (e.g. 
>> dc=VergeNet,dc=com). It’s not a blocking problem for us, but I just 
>> want to know if this is a known issue or has been addressed before. 
>> Here’s the description of the problem:
>>
>> In ServerContextFactory.java line #630, JdbmDatabase is initialized 
>> with un-normalized suffix.
>>
>> Database db = new JdbmDatabase( upSuffix, wkdir );
>>
>> But in JdbmDatabase.java line #673, when adding an entry to the 
>> database, the suffix is compared with the normalized dn of the new 
>> entry.
>>
>> if ( dn.equals( suffix ) )
>>
>> This is causing the add operation to fail.
>>
>> A similar problem also happens here. In RootNexus.java line #203, the 
>> suffix is being normalized during registration.
>>
>> backends.put( backend.getSuffix( true ).toString(), backend );
>>
>> However, in RootNexus.java line #556, the dn used to look up the 
>> backend is not normalized.
>>
>> return ( ContextPartition ) backends.get( clonedDn.toString() );
>>
>> This is causing the search operation to fail.
>>
>> Should I file a bug? Thank you very much.
>>
>> -- 
>>
>> Endi S. Dewata
>>
>
>

Re: Problem with mixed case in suffix

[Alex Karasulu <ao...@bellsouth.net>]

Yep Endi you caught a bug - good job! Could you go through the motions 
of filing it via JIRA?

Thanks much,
Alex

Endi Sukma Dewata wrote:

> Hi,
>
> I’m Endi Dewata from Verge Archemedia Labs in Austin, TX. I’m working 
> with Adison Wongkar trying to hook up our Virtual Directory product 
> with ApacheDS. So far so good, but it seems that we just found a minor 
> problem with suffixes that contain capital letters (e.g. 
> dc=VergeNet,dc=com). It’s not a blocking problem for us, but I just 
> want to know if this is a known issue or has been addressed before. 
> Here’s the description of the problem:
>
> In ServerContextFactory.java line #630, JdbmDatabase is initialized 
> with un-normalized suffix.
>
> Database db = new JdbmDatabase( upSuffix, wkdir );
>
> But in JdbmDatabase.java line #673, when adding an entry to the 
> database, the suffix is compared with the normalized dn of the new entry.
>
> if ( dn.equals( suffix ) )
>
> This is causing the add operation to fail.
>
> A similar problem also happens here. In RootNexus.java line #203, the 
> suffix is being normalized during registration.
>
> backends.put( backend.getSuffix( true ).toString(), backend );
>
> However, in RootNexus.java line #556, the dn used to look up the 
> backend is not normalized.
>
> return ( ContextPartition ) backends.get( clonedDn.toString() );
>
> This is causing the search operation to fail.
>
> Should I file a bug? Thank you very much.
>
> --
>
> Endi S. Dewata
>