Mailing-List: contact dev-help@directory.apache.org; run by ezmlm
Precedence: bulk
Reply-To: "Apache Directory Developers List" <dev@directory.apache.org>
Message-ID: <14937780.1160664159102.JavaMail.jira@brutus>
Date: Thu, 12 Oct 2006 07:42:39 -0700 (PDT)
From: "Norval Hope (JIRA)" <jira@apache.org>
To: dev@directory.apache.org
Subject: [jira] Commented: (DIRSERVER-760) reading .schema files at server
 start-up
In-Reply-To: <17758845.1160663915139.JavaMail.jira@brutus>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit

    [ http://issues.apache.org/jira/browse/DIRSERVER-760?page=comments#action_12441759 ] 
            
Norval Hope commented on DIRSERVER-760:
---------------------------------------

I should also mention from the testing I've done the difference in performance between loading the schemas dynamically at startup versus running the Maven generated classes is pretty minimal, I've loaded eight schemas where one is large (689 attribute types and 58 object classes) in less then a second. The overhead is really just the cost of parsing the .schema files as the generated code ultimately creates the same classes that my convertors do.

> reading .schema files at server start-up
> ----------------------------------------
>
>                 Key: DIRSERVER-760
>                 URL: http://issues.apache.org/jira/browse/DIRSERVER-760
>             Project: Directory ApacheDS
>          Issue Type: New Feature
>          Components: core
>    Affects Versions: 1.1.1
>         Environment: N/A
>            Reporter: Norval Hope
>         Attachments: schema_loader.patch
>
>
> I am submitting a patch for a feature I required, and which I've seen a number of queries about on the DEV list. Currently the only way to get new .schema information into ApacheDS is to use the Maven2 plugin, requiring a rebuild of the server (not to mention working around problems like methods being generated which are too long for Java to handle).
> Instead this patch adds support for reading of  specified .schema files from server.xml at server startup, via a new interface SchemaLoader and a default implementation FileSystemSchemaLoader. The latter supports reading all files matching a specified regex in a specified directory (in which case users must be careful to ensure that files appear lexicographically before their dependant files, ala init.rc in Linux) or reading an ordered list of file names under a specified directory. An example server.xml snippet is as follows:
>     <property name="bootstrapSchemas">
>       <set>
>         <bean class="org.apache.directory.server.core.schema.bootstrap.AutofsSchema"/>
>         <bean class="org.apache.directory.server.core.schema.bootstrap.CorbaSchema"/>
>         <bean class="org.apache.directory.server.core.schema.bootstrap.CoreSchema"/>
>         <bean class="org.apache.directory.server.core.schema.bootstrap.CosineSchema"/>
>         <bean class="org.apache.directory.server.core.schema.bootstrap.ApacheSchema"/>
>         <bean class="org.apache.directory.server.core.schema.bootstrap.CollectiveSchema"/>
>         <bean class="org.apache.directory.server.core.schema.bootstrap.InetorgpersonSchema"/>
>         <bean class="org.apache.directory.server.core.schema.bootstrap.JavaSchema"/>
>         <bean class="org.apache.directory.server.core.schema.bootstrap.Krb5kdcSchema"/>
>         <bean class="org.apache.directory.server.core.schema.bootstrap.NisSchema"/>
>         <bean class="org.apache.directory.server.core.schema.bootstrap.SystemSchema"/>
>         <bean class="org.apache.directory.server.core.schema.bootstrap.ApachednsSchema"/>
>       </set>
>     </property>
>     <property name="schemaLoaders">
>         <list>
>             <bean class="org.apache.directory.server.core.schema.FileSystemSchemaLoader">
>                 <constructor-arg><value>./</value></constructor-arg>
>                 <!--<constructor-arg><value>.*_openldap.schema</value></constructor-arg>-->
>                 <constructor-arg>
>                     <list>
>                         <value>my1_openldap.schema</value>
>                         <value>my2_openldap.schema</value>
>                     </list>
>                 </constructor-arg>
>             </bean>
>         </list>
>     </property>
> noting that the Maven2 style approach is of course still supported where desired. A list is used so that multiple SchemaLoader implementations can be appended if required (which I do). 
> I have also included SchemaFromSearchConverter which can read a schema from another directory via conventional query, which may come in useful for people using ApacheDS as a virtual directory container of sorts. It is not required by the core patch and can be excluded if no generic use is seen for it.
> A few issues are still outstanding:
>    1. Is the patch seen as worthwhile by the DEV list and/or are there any issues with my implementation of it?
>    2. At it's core the patch uses the OpenLdapSchemaParser to read the .schema files, which is currently only available in the Maven2 plugin artifact     
> <groupId>org.apache.directory.server</groupId>
> <artifactId>apacheds-core-plugin</artifactId>
>    
>     which is a bit clunky as it means people need to include this jar in their classpath (or in the UberJar). Perhaps this parsing component should be separated from the rest of the maven plugin to tidy this up (I don't know maven well enough to do this myself).
>    
>    3. If the feature and implementation are ok, what further steps should I take re preparing the patch for inclusion?
> Thanks (by the way I was previously known as Norbert Reilly)

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira