[ 
https://issues.apache.org/jira/browse/MCOMPILER-203?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15094924#comment-15094924
 ] 

Andreas Gudian commented on MCOMPILER-203:
------------------------------------------

[~dmlloyd]: my advice would be make sure the processors don't bring any 
dependencies - either by suggesting that to the developers of them, or by 
creating your own repackaged version using the maven-shade-plugin. It's not too 
complicated and you'll have less trouble when trying to integrate it into 
further tooling (e.g. in IDEs - working with annotation processors with the 
Eclipse incremental compiler is really cool).
Ooor (but that may not work in every case), you add multiple executions of the 
maven-compiler-plugin to your pom, each of them using a different processor and 
with {{proc:only}}... But I get that it's not what you're looking for :-/

> Allow compiler-plugin to specify annotation processor dependencies
> ------------------------------------------------------------------
>
>                 Key: MCOMPILER-203
>                 URL: https://issues.apache.org/jira/browse/MCOMPILER-203
>             Project: Maven Compiler Plugin
>          Issue Type: New Feature
>    Affects Versions: 2.3.2, 3.1
>         Environment: Java 6+
>            Reporter: David M. Lloyd
>            Assignee: Andreas Gudian
>             Fix For: 3.5
>
>
> Right now the status quo for annotation processor artifacts requires one of 
> two actions:
> # Use an external plugin for annotation processing
> # Put the annotation processor in as a dependency with {{provided}} scope
> The former is suboptimal because the external plugins are clunky and 
> ill-supported, and inflexible/hard to use.  The latter is suboptimal because 
> it is often the case that you do not want to leak annotation processor 
> classes on to the application class path.
> It should be possible to add annotation processor dependency artifacts to the 
> compiler plugin configuration such that they are recognized by the annotation 
> processing search algorithm of the compiler, but they do not actually appear 
> on the compilation class path.  Ideally they would also be isolated from one 
> another (dependency graphs and all), but that's more of a "nice to have".



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to