JarCastingjruby-mains
vision
- install your gem and jar dependencies into some directory /tmp/myapp
- pack the gems + jars into gems.jar
- pack the whole application as it into application.jar
- run the application from these jars or put them along jruby.jar and jruby-rack.jar into WEB-INF/lib
current situation
NOTE: at the time of writing you need to use jruby-complete.jar from https://oss.sonatype.org/content/repositories/snapshots/org/jruby/jruby-complete/1.7.20-SNAPSHOT/ for the examples. jruby-mains itself work with old jruby versions.
first pack all the gem and jar dependencies into one jar:
export JARS_HOME=/tmp/app/jars
export GEM_HOME=/tmp/app
export GEM_PATH=/tmp/app
gem install bundler
bundle install
rmdir /tmp/app/doc /tmp/app/extensions /tmp/app/build_info
rm -rf /tmp/app/cache
jar -cvf gems.jar -C /tmp/app .
now pack the application:
jar -cvf application.jar .
now run these jars with
java -cp application.jar:gems.jar:jruby-complete.jar org.jruby.Main -C uri:classloader:/ -S rackup
this is equivalent to (here the classloader finds the application on the classpath ".")
java -cp .:gems.jar:jruby-complete.jar org.jruby.Main -C uri:classloader:/ -S rackup
or very close to (the current directory is on the file system and not inside the classloader)
java -cp .:gems.jar:jruby-complete.jar org.jruby.Main -C . -S rackup
the last can be reduced to (since GEM_PATH and JARS_HOME needs to be set)
java -jar jruby-complete.jar $GEM_PATH/bin/rackup
the only difference is from where the application is loaded and where the current directory (-C) is pointing to.
note: if GEM_PATH and JARS_HOME are not set then JRuby is looking for gems and jars in uri:classloader:/.
ideally bundler should be only development dependency and then can be excluded.
what about war files ?
just put the jruby-complete.jar, jruby-rack.jar, application.jar and the gems.jar into WEB-INF/lib and configure your WEB-INF/web.xml to use the classpath-layout
<web-app>
<context-param>
<param-name>jruby.rack.layout_class</param-name>
<param-value>JRuby::Rack::ClassPathLayout</param-value>
</context-param>
<filter>
<filter-name>RackFilter</filter-name>
<filter-class>org.jruby.rack.RackFilter</filter-class>
</filter>
<filter-mapping>
<filter-name>RackFilter</filter-name>
<url-pattern>/*</url-pattern>
</filter-mapping>
<listener>
<listener-class>org.jruby.rack.RackServletContextListener</listener-class>
</listener>
</web-app>
then the web-application uses the same loading semantic as the standalone execution.
of course you can just unpack the gems.jar and application.jar into WEB-INF/classes which does not make a difference since from both location all the ruby sources are loaded via the jruby-classloader.
using jruby-mains
this jruby-mains artifact expect the gems and jar dependencies be vendored at the root of the application.jar
JARS_HOME=./jars GEMS_HOME=. GEMS_PATH=. bundle install
jar -cvf application.jar .
and run it with
java -cp application.jar:jruby-complete.jar:jruby-mains.jar org.jruby.Main bin/rackup
or via the -S switch
java -cp application.jar:jruby-complete.jar:jruby-mains.jar org.jruby.mains.JarMain -S rackup
finally you can merge those three jar files into one and set the entry-point to org.jruby.mains.JarMain. this reduces the execution to
java -jar application-uber.jar -S rackup
some executables spawn a new ruby process which is not working from a jruby-complete exectution. but in such cases you can use the entry-point org.jruby.mains.ExtractingMain which unpacks the jar into a temporary directory and then executes the application.
development
use maven 3.3.x or the supplied mvnw wrappers. all integration tests in src/it are using the ruby pom DSL the pom.xml is just a dummy for the invoker plugin to find the test.
license
Almost all code is under the LGPL-3 license.
contributing
- Fork it
- Create your feature branch (
git checkout -b my-new-feature) - Commit your changes (
git commit -am 'Added some feature') - Push to the branch (
git push origin my-new-feature) - Create new Pull Request
meta-fu
report issues and enjoy :)
