In general wolifecycle implements some kind of classifier support, but setting a classifier on a woapplication project generates incomplete and broken applications.
<build>
<plugins>
<plugin>
<groupId>org.wocommunity</groupId>
<artifactId>wolifecycle-maven-plugin</artifactId>
<extensions>true</extensions>
<configuration>
<skipAppleProvidedFrameworks>${skip.apple.frameworks}</skipAppleProvidedFrameworks>
<classifier>testclassifier</classifier>
</configuration>
</plugin>
</plugins>
</build>
The content of the WOA build is split between two directories, one with the classifier embedded in the name (e.g. ERXTest-7.2-SNAPSHOT-testclassifier.woa), and one without the classifier (ERXTest-7.2-SNAPSHOT.woa). The classifier part contains all application resources (components, webserverresources, etc.), while the non-classifier part contains the java classes and WOA infrastructure (classpath files, start scripts).
The tgz-archives are packed and named from the WOA without the classifier, but installed and deployed under the classifier name.
In my opinion the build should generate only the non-classifier WOA directory (named after the finalName property) with all content, and only then generate the tgz-archives with the classifier name. If it is necessary to embed the classifier into the app directory name, then finalName could be modified accordingly.
In general wolifecycle implements some kind of classifier support, but setting a classifier on a woapplication project generates incomplete and broken applications.
The content of the WOA build is split between two directories, one with the classifier embedded in the name (e.g.
ERXTest-7.2-SNAPSHOT-testclassifier.woa), and one without the classifier (ERXTest-7.2-SNAPSHOT.woa). The classifier part contains all application resources (components, webserverresources, etc.), while the non-classifier part contains the java classes and WOA infrastructure (classpath files, start scripts).The tgz-archives are packed and named from the WOA without the classifier, but installed and deployed under the classifier name.
In my opinion the build should generate only the non-classifier WOA directory (named after the
finalNameproperty) with all content, and only then generate the tgz-archives with the classifier name. If it is necessary to embed the classifier into the app directory name, then finalName could be modified accordingly.