Mailing-List: contact jackrabbit-dev-help@incubator.apache.org; run by ezmlm
Precedence: bulk
Reply-To: jackrabbit-dev@incubator.apache.org
Message-ID: <2023581270.1115906469135.JavaMail.jira@ajax.apache.org>
Date: Thu, 12 May 2005 16:01:09 +0200 (CEST)
From: "fabrizio giustina (JIRA)" <jira@apache.org>
To: jackrabbit-dev@incubator.apache.org
Subject: [jira] Commented: (JCR-116) JavaCC grammar generation to
 ${maven.build.dir}/generated-src
In-Reply-To: <17663890.1114587746728.JavaMail.jira@ajax.apache.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

     [ http://issues.apache.org/jira/browse/JCR-116?page=comments#action_65126 ]
     
fabrizio giustina commented on JCR-116:
---------------------------------------

> Keep custom maven scripts to generate the parsers. Unless we find a way how to specify two grammar files for jackrabbit? If you are still watching this issue fabrizio, do you know if this is possible? 

Mh, at the moment the javacc maven plugin doesn't support more than one grammar file, so there is no a clean-maven-way to do it.
Anyway, I think that you could get rid of ant with a little hack in maven.xml: since the problem is that the maven.javacc.javacc.grammar and maven.javacc.jjtree.grammar can hold a single value you can simply change the property on the fly in maven.xml and repeat the javacc goal twice... not really clean but it should work

You should add an enhancement request to the maven javacc jira at http://jira.codehaus.org/browse/MPJAVACC. A better solution would be adding the multiple file support to the plugin directly and submit a patch (the maven javacc plugin is just a jelly file which calls ant, just like the custom maven.xml in jackrabbit, so it should not be so difficult)

> JavaCC grammar generation to ${maven.build.dir}/generated-src
> -------------------------------------------------------------
>
>          Key: JCR-116
>          URL: http://issues.apache.org/jira/browse/JCR-116
>      Project: Jackrabbit
>         Type: Wish
>   Components: maven, xpath, sql
>     Reporter: Jukka Zitting
>     Priority: Minor
>  Attachments: jackrabbit-javacc-target-patch.txt
>
> Currently the JavaCC grammars in src/grammar/{xpath,sql} are processed into Java source files in src/java/org/apache/jacrabbit/core/query/{xpath,sql} where we also have normal version controlled source files. This leads to the need to maintain special svn:ignore properties and also the more general issue of mixing manually written and automatically generated source files. Because of this the "maven clean" command does not (at the moment) truly restore your source tree to a "fresh checkout" state.
> I'm proposing (as a wish, you are free to disagree) that the JavaCC grammars be generated into Java files within the Maven build directory. The attached patch modifies the javacc maven goals to generate files into ${maven.build.dir}/generated-src. The modified prepare-filesystem goal also adds the generated source path ${maven.build.dir}/generated-src/main/java into the maven compile set so that the generated sources are included in the normal builds.
> PS. There are a couple of JavaCC generated files that have been intentionally modified for Jackrabbit. The ant:delete commands at the end of the jacrabbit:generate-*-parser goals specifically remove these generated files. It would however be nicer if custom modifications would not be needed.

-- 
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