Skip to content
master
Switch branches/tags
Code

Latest commit

 

Git stats

Files

Permalink
Failed to load latest commit information.
Type
Name
Latest commit message
Commit time
 
 
src
 
 
 
 
 
 
 
 
 
 

SBT Scala template with auto-generated WSDL

The SBT project kickstarter template that encapsulates all the dirty work around WSDL and raw Java beans generation for WSDL. The goal is to automate and hide this legacy boilerplate code and keep your Scala project pristine in the VCS.

Overview

To use legacy SOAP services, you are often required to generate Java beans based on the WSDL schema. With a template like this, this can be done automatically when you build the project (precisely speaking, with sbt compile), but also could be invoked manually with the bin/wsdl_import.sh script.

The main WSDL generation routine is handled by wsdl subproject, which produces plain Java beans into its own src folder and protects the repository from holding them under VCS. WSDL generation itself employs nothing but native Java's wsimport (also, using JRE's rt.jar for executing SOAP calls) so the client stays clear from unnecessary 3rd parties, having just essential logging & testing ones. A WSDL file could be also placed locally (e.g. /bin/wsdl_local/nl.xml) if it requires a little fixture in the XML schema for wsimport to be able to generate Java beans out of it. Check comments in the sh file for more details.

Typical usage

Start with sbt compile to conveniently set an IDE project, so all the Java bean sources would be produced for you and the IDE would be able to link dependencies. For coding, instantiate a SOAP service as usual and continue using the generated code.

Feel free to fork this repo and start your project with this template. Or just copy-paste what you're interested in.

Building and publishing

All WSDL generation should be done automatically when you build or compile the project. There is a dependency for the main compile task to be executed after subproject's compile task, which in turn involves WSDL generation process itself. Basically, all the work should be done when your Scala code in the main project would start compiling, so nothing much to worry about.

It also pretends to be a bit smart by involving some of SBT's native features like preserving subproject's jars if they exist already, or checking for auto-generated code to avoid running this dirty routine every time. You can notice some code in subproject's SBT config.

Typical sbt commands work fine here. One should note, however, the sbt publish command would expect a file .sbt/nexus_creds as defined in the official docs. Also, it might be required to sbt clean build before publishing e.g. when starting from scratch right after cloning the repo or after deleting all the generated sources.

At the same time, in order to provide tests with some servlet context, SBT has to include JRE's jars like rt.jar so they are linked as unmanaged local jars and therefore require JAVA_HOME to be properly set.

Credits

Thanks oorsprong.org for silently sharing your SOAP endpoints purely for testing & education process. If you are the owner / admin, please let me know if you no longer wish having this service to be used in this repo.

To anyone who plays with it, please have some respect and use this service carefully.

License: Apache 2.0