Notes on using Talend Open Studio

To save a job, e.g., to share it with others

This saves the "item" file, properties, etc. It does not save the generated classes that are installed in CollectionSpace to invoke the export.

Once a job has been defined, right click on the job and select "export items" to save the job. When the dialog appears, check the "Export Dependencies" box, and ensure that the connections and contexts are selected, as well as the Job design. You can then share the generated files (e.g., by zipping and mailing). The person you are sharing with will run an "Import" and specify the root of the tree of output files you generated.

Integration Notes

See this thread for notes on integrating with Java code. Can also call run rather than main.

Need to figure out:

  1. How to code the tJava component so that we can provide an output stream. It basically needs to call some static method, which will be awkward, but we could make it happen somehow. 
    1. One alternative would be to edit the Java when it is done to add a method to set the output stream. 
    2. Can set class names into context, and refer to them in tJava, but this still does not get us to an instance.
  2. How to get progress indicators back - see notes on wrappers others have written.
  3. Whether the shared libs they produce are custom per job, or can be linked once for simplicity.
  4.  

Additional notes pages of interest as I dig through their code for integration: